THE GARDENS’ BULLETIN, SINGAPORE
The Gardens’ Bulletin, Singapore is a peer-reviewed journal publishing original papers and
reviews on a wide range of subjects: plant taxonomy (including revisions), phytogeography,
floristics, morphology, anatomy, as well as horticulture and related fields, such as ecology and
conservation, with emphasis on the plant life of the Southeast Asian-Pacific region.
Dr. K.M. Wong Dr. Jana Leong-Skorniékova
Singapore Botanic Gardens Singapore Botanic Gardens
(Editor) (Assistant Editor)
Dr. P.Y. Tan Christina Soh
National Parks Board, Singapore
(Assistant Editor)
Singapore Botanic Gardens
(Business Manager)
Editorial Advisory Board
Professor Sir Peter Crane Dr. Nigel P. Taylor
Yale University Singapore Botanic Gardens
U.S.A. National Parks Board, Singapore
Dr. Rogier P.J. de Kok Dr. I.M. Turner
Royal Botanic Gardens, Kew, Singapore Botanic Gardens
U.K. National Parks Board, Singapore
Dr. W.J. Kress Dr. J.F. Veldkamp
National Museum of Natural History Nationaal Herbarium Nederland
Smithsonian Institution, U.S.A. The Netherlands
Dr. David J. Middleton Dr. Jun Wen
Royal Botanic Garden Edinburgh National Museum of Natural History
U.K. Smithsonian Institution, U.S.A.
Dr. Kiat W. Tan Professor Nianhe Xia
Gardens By The Bay South China Institute of Botany
National Parks Board, Singapore PR. China
Published twice yearly by the National Parks Board, Singapore, the annual subscription for the journal
is Singapore $100.00 including postage. Overseas subscribers should make payment in the form of
bank draft or international money order in Singapore currency, payable to National Parks Board. Please
forward payment to “Accounts Receivable Section, National Parks Board, Headquarters, Singapore
Botanic Gardens, 1 Cluny Road, Singapore 259569”.
Instructions for contributing authors
(more details in http://www.sbg.org.sg/research/instructions-for-authors.pdf)
Manuscripts should be sent by e-mail to <gardbullsing@gmail.com> or posted to
The Editor, Gardens’ Bulletin Singapore, Singapore Botanic Gardens, 1 Cluny Road,
Singapore 259569.
The language used is English. Prior to submission, it may be advisable to have manuscripts checked by
someone fluent or proficient in that language. Contributions must not have been published or be under
consideration elsewhere. There are no page charges. Submitted material cannot be returned.
Authors currently opt for 50 paper offprints and/or a pdf reproduction of their contribution, gratis.
Cover letter, submission and copyright. In your cover letter or message, state clearly you are submitting
your manuscript (state title) for publication in the Gardens’ Bulletin, Singapore. By submitting the
manuscript(s), authors agree that upon acceptance, copyright of the entire work becomes the property
of the Singapore Botanic Gardens / National Parks Board, Singapore. Authors must ensure all required
permission has been obtained in writing to publish any material they present; proof may be required.
Manuscripts are best submitted in electronic form (Microsoft Word 97-2003 preferred). Use Times New
Roman, 10-point font, 1.5-line spacing. Do not send original artwork unless requested by the Editor.
Number all pages serially including the title and abstract on the first.
(continued on inside back cover)
P; r55, ; PAF re 3 Ae my My LTS v A
2 APR 23 2012 AKOLD ARBORETUM
tO LEAT e eet
4)
/ /
{
The Gardens’ Bulletin
Singapore
VOL. 63(1 & 2) 2011 ISSN 0374-7859
Proceedings of the
8th Flora Malesiana Symposium
Singapore, 23—27 August 2010
Edited by
K.M. Wong, Jana Leong-Skorniékova,
Serena Lee and Y.W. Low
Date of publication: 22 December 2011
Copyright ©
National Parks Board
Singapore Botanic Gardens
1 Cluny Road
Singapore 259569
Printed by Oxford Graphic Printers Pte Ltd
The Gardens’ Bulletin
Singapore
VOL. 63(1 & 2) 2011 ISSN 0374-7859
Proceedings of the 8th Flora Malesiana Symposium
CONTENTS
HISTORICAL PERSPECTIVE
J.F. Veldkamp - Georgius Everhardus Rumphius (1627-1702), the blind seer of Ambon ..... 1
TAXONOMIC STUDIES
H. Rustiami, J.P. Mogea and S.S. Tjitrosoedirdjo - Revision of the rattan genus
Daemonorops (Palmae: Calamoideae) in Sulawesi using a phenetic analysis approach ....... 17
Ary Prihardhyanto Keim, Rugayah and Himmah Rustiami - The Pandanaceae of the
Bukit Baka Bukit Raya National Park and adjacent areas, West and Central Kalimantan,
Indonesia, with notes on their nomenclature and the rediscovery of Pandanus aristatus
VL, SE PEPEL, TESTERS O71 eR as ae eee ere ee ee ee 31
Sri Endarti Rahayu, Alex Hartana, Tatik Chikmawati and Kuswata Kartawinata -
A taxonomic study of the Pandanus furcatus and P. tectorius complexes (Pandanaceae)
(ED ARIAL cericcsBchiccSS S508 yaa ACEO PES BEE CCS SS Sate eee 63
A. Sumadijaya and J.F. Veldkamp - Bothriochloa (Poaceae: Andropogoneae) in Malesia .... 71
Marie Briggs - Saurauia (Actinidiaceae) of New Guinea: current status, future plans ........ 77
Mark Hughes and Deden Girmansyah - Searching for Sumatran Begonia described by
William Jack: following in the footsteps of a 19th century Scottish botanist ...............0..05. 83
S. Syahida-Emiza, G. Staples and N.W. Haron - Materials for a revision of Erycibe
(ConivolvMmlaceac) mn LeninS lat Vial ays toeeeere ene eee waren ese oan eos cek cee sac saneesssoeenees 97
T.M.A. Utteridge and M. Schori - Updating Malesian Icacinaceae .................ceeeeeceeeeeees 105
J.A. Wearn and D.J. Mabberley - Clerodendrum confusion—redefinition of, and new
Poke peeilVesMOl a lalpe: leaiateiOOnUS. 2-512. eee EO ceed. Sache ine ediecsenwopsevestesareetsese 119
I. Nadiah and E. Soepadmo - A synopsis of Coelostegia (Bombacaceae/Malvaceae:
Helicteroideae: Durioneae) and new records from Borneo ................cccccccecescceeeeeeessscceeeeetees 125
H.S. Tan, R.C.K. Chung and E. Soepadmo - A synopsis of Jarandersonia (Malvaceae:
Brownlowio1deae)) s.2iccesvesouresccadaceosseecaa@uvcsetededeseestihec ete Seco eee 137
Peter Wilkie - Towards an account of Sapotaceae for Flora Malesiana ...............0:::cc0c00 145
Barry J. Conn and Julisasi T. Hadiah - Precursor to flora account of Procris
(Urticaceae) in Peninsular Malaysiaies 2:s:sees ese eee ee et 155
G.E. Lee, S. Robbert Gradstein, A. Damanhuri and A. Latiff - Towards a revision of
Lejeunea (ejeuneaceac) in Malaysia «.....2.2...12...ccse.e+-sese-cesteostessceess sk pee eer ee 163
Vincent Demoulin - The study of larger basidiomycetes, especially polypores, in the
Malesian region and the role of the Singapore Botanic Gardens ..............::c:cceceeseeeeeeeereeees WS
FLORAS, VEGETATION, AND FLORISTIC STUDIES
M.C. Roos, W.G. Berendsohn, S. Dessein, T. Hamann, N. Hoffmann, P. Hovenkamp,
T. Janssen, D. Kirkup, R. de Kok, S.E.C. Sierra, E. Smets, C. Webb and P.C. van
Welzen - e-Flora Malesiana: state of the art and perspectives ...............:ceesceessteeeeeceneeeeeeeees 189
K.Y. Chong, Hugh T.W. Tan and Richard T. Corlett - A summary of the total vascular
plant flora Of SimPapore: ...i....:.cecc..scccseoctdecsecaceuds ceesececeneenence coeseecnecces ee eeeeR tec eaee Cen cers meeE 197
A.T.K. Yee, Richard T. Corlett, S.C. Liew and Hugh T.W. Tan - The vegetation of
Singapore—an updated Map ~...<6c-<3<sc.ncendccessevsssseoucatesss erase ore ee eee ee ee 205
Zulhazman Hamzah, Mashhor Mansor and P.C. Boyce - Notes on Araceae of Kuala
Koh; Kelantan; Peninsular Malaysia 2i27.2)..25 icc ccteeseeet rece eee te ee 7 4p\3)
Victor B. Amoroso, Socorro H. Laraga and Bridget V. Calzada - Diversity and
assessment of plants in Mt. Kitanglad Range Natural Park, Bukidnon, Southern
Phatlipyp ines 2st se.cc seeds csceeeccte se Rho eens ae eA BE 219
DIVERSITY STUDIES, MOLECULAR PHYLOGENETICS, AND BIOGEOGRAPHY
S.L. Low, S.Y. Wong, J. Jamliah and P.C. Boyce - Phylogenetic study of the Hottarum
Group (Araceae: Schismatoglottideae) utilising the nuclear ITS region .............:ccccceeeeeee 237
André Schuiteman - Dendrobium (Orchidaceae): To split or not to split? ............cceeeeee 245
Birgit Oelschlagel, Sarah Wagner, Karsten Salomo, Nediyaparambu Sukumaran
Pradeep, Tze Leong Yao, Sandrine Isnard, Nick Rowe, Christoph Neinhuis and
Stefan Wanke - Implications from molecular phylogenetic data for systematics,
biogeography and growth form evolution of Thottea (Aristolochiaceae) ...............0:00:0000000 259
S. Rajbhandary, M. Hughes, T. Phutthai, D.C. Thomas and Krishna K. Shrestha -
Asian Begonia: out of Attica via the Himalayas? ...22 ee eee ee Peg fE |
Gemma L.C. Bramley - Distribution patterns in Malesian Callicarpa (Lamiaceae) ......... 287
Hans P. Nooteboom - How did Magnolias (Magnoliaceae: Magnolioideae) reach
PPUEREE ASSEN? ce ec Bet se Bee aa ne en On eee 299
Pingting Chen, Jun Wen and Longqing Chen - Spatial and temporal diversification of
TERETE EET Se eo oR Ee 307
Peter C. van Welzen and Niels Raes - The floristic position of Java ...........cccceceeeeteeeeees 329
ECOLOGICAL STUDIES
Mohammed Kamrul Huda and Christopher C. Wilcock - Colonisation and diversity
of epiphytic orchids on trees in disturbed and undisturbed forests in the Asian tropics ...... 34]
Didik Widyatmoko - Habitat and ecological preferences of Hydriastele costata (Palmae)
PARED ES EEC eG RLRATO)T A) CORI a ee eS er meee eet a ee ee ee 357
K.M. Wong and Y.W. Low - Hybrid zone characteristics of the intergeneric hybrid
bamboo * Gigantocalamus malpenensis (Poaceae: Bambusoideae) in Peninsular
SN a ARE AAR ae ese eet hn Soe Senn lieanone dh anehoinndscbbavceniapccaiasssavacaeiebevthevedees 375
A. van der Ent - The ecology of ultramafic areas in Sabah: threats and conservation
OEELS cocoocesjecutndedhcdecteceee tc eae een eae 385
CONSERVATION PERSPECTIVES
Teodora D. Balangcod, Virginia C. Cuevas and Ashlyn Kim D. Balangcod -
Cultivation and conservation of Lilium philippinense (Liliaceae), the Philippine endemic
NEADS) LLALE) coaece soe ce cea aie Ae ee el ne eee eee eee Renee 395
Ashlyn Kim D. Balangcod - Predicting distribution of Lilium philippinense (Liliaceae)
over Luzon’s Cordillera Central Range, Philippines, using ArcGIS Geostatistical Analyst .......
cogs gng Sen e227 227 oc020000 gc oS Sena el Ree ee 409
Y.M. Chan, L.S.L. Chua and L.G. Saw - Towards the conservation of Malaysian
Mee REMC SUITANITI(A (EALITIAG)| 21 -25sver-concesectsvee-cedeveodacevet ne eenescteeusceendeh sanarasbaheeaeasodesssedenseowsen 425
R. Kiew, A.R. Ummul-Nazrah & L.S.L. Chua - Conservation status of Paraboea
pee NG CONTAC EAC) Il) WIAIAY SIA) oscu: s-c<n..-c0 es -acceuconeue seasaczsdescanaiesoeeeuctevsssevusbucspoxs<enaccaacesars 433
M.Y. Chew and N.W. Haron - Utricularia (Lentibulariaceae) habitat diversity in
Peninsular Malaysia and its implications for COMSETVAtION .............:esceeeeeeeeeceeeeeeeeeeeeeeneeeeee 451
Julisasi T. Hadiah - Establishment of Enrekang Botanic Garden, South Sulawesi: an
effort to conserve plant diversity in the Wallacea region ..0...........cesceeeseeeenceeeeceeeeeeeeneeeeeees 465
MORPHOLOGY, ANATOMY, AND BIOLOGICAL STUDIES
E.H.S. Chin and A.L. Lim - Comparative pollen morphology of three Al/ternanthera
SPUQCISS (UAMTENTIUINENSERYS) 5 otsootia os codec caesecceceoREGSELe Or coos SEES CCE CEE En aac ceee ee EEE 471
Wiguna Rahman - Flower biology of four epiphytic Malesian Gesneriads ......................- 485
Rogier de Kok - The genus Premna (Lamiaceae) and the presence of ‘pyro-herbs’ in the
Flora Malesiama ated .2...2.5.csece seeks sack So cecte cose ce ceacteate anes sete aeeeteex ae ee ac ieee oe 495
Netty W. Surya and M. Idris - A preliminary study on in vitro seed germination and
rooted callus formation of Tetrastigma rafflesiae (Vitaceae) .......cccccecccesceessceseeeeeseeeeseesseees 499
A.T. Nor-Ezzawanis - Comparative anatomy of Grammitidaceae genera in Peninsular
Moalay sad .3.:.5:csescecscesesazeesateasvecesouttcuschatcen cnetaesssdcateenaceserceeee soneeseccesea ate cee ee S07
ETHNOBOTANY
Vera Budi Lestari Sihotang - Ethnomedicinal study of the Sundanese people at the
Bodogol area, Gede Pangrango Mountain National Park, West Java ...........-.ssccsseeeeeeeeereees S19
Digitized by the Internet Archive
in 2014
https://archive.org/details/gardensbulletins4631unse
Gardens’ Bulletin Singapore 63(1 & 2): 1-15. 2011 ]
Georgius Everhardus Rumphius (1627-1702),
the blind seer of Ambon
J.F. Veldkamp
Netherlands Centre for Biodiversity — Naturalis,
Section National Herbarium of The Netherlands, Leiden University,
P.O. Box 9514, 2300 RA Leiden, The Netherlands
veldkamp@nhn.leidenuniy.nl
ABSTRACT. Georg Eberhard Rumpf, better known as Rumphius (1627-1702) was a Homo
universalis and is the undisputed patriarch of Malesian botany, zoology, geology (including
fossils!), colonial history; pharmaceutical, architectural, juridical (local and Western),
ethnological, linguistic, historical, and religious matters, including astrology and magic. To
botanists he is best known for his Herbarium amboinense (1741—1750), the first account and
sometimes the only one of Malesian plants. This is a 7-volume folio work with extensive
descriptions and discussions in Latin and Dutch of about 1200 species with 811 full-page
illustrations. A brief account of his life and works is given.
Keywords. Ambon, Dutch United East Indian Company, Herbarium amboinense, Malesia,
natural science, Rumpf, Rumphius, VOC
Introduction
Many articles and books have been written about Georg Eberhard Rumpf (Georgius
Everhardus Rumphius in Latin), better known as Rumphius (1627—1702) and his
observations in the Moluccas (Fig. 1, 2), which have given him an everlasting place of
honour in the history of natural science (see Appendix A for a selection). Only a few
of these publications were in English, the last one by Beekman (1999), who translated
d’Amboinschen Rariteitkamer (1705), the Amboinese Curiosity Cabinet. This deals
mainly with animals, mineralogy, and geology, explicitly recognising zoological
fossils. European scientists regarded these as remnants from before the Deluge, or
were created by the Devil to confuse good Christians. Some modern Creationists
still think so. Every taxonomist in Malesia has encountered species named after him:
“rumphii’, “rumphianus”, or based on taxa first described in his works.
Rumphius and Ambon
Ambon island (Pulau Ambon) in the 17th century (Fig. 3) was the centre of the spice
trade: cloves, nutmeg, mace, and pepper. These commodities were very much in
demand in Europe and worth a fortune once there. At one time for a bag of pepper one
could buy an imposing house in a major town in The Netherlands! The organisation
Gard. Bull. Singapore 63(1 & 2) 2011
ute
SENT Agee d OS
FFIGIES ni
GEORGIL EVERHARDI RUMPHIL, HANOVIENSIS MTAT: LXVIIL.
Cae Wes Ret Ocns COUCOS CaM J RANE i ani i
t Tt rmcwroe me COS we Ce SOE edd d any de bare dal
Vs.
thle 2S Wee serl tut ese Ge PRAMAS OFTDOHE, COMLS
Oelga Dshinig Of CLOAIC : AEC decet CPUs,
Or Compere posate Ys Gua donde |
Fig. 1. Portrait of Rumphius at age 68 by his son Paulus.
Georgius Everhardus Rumphius 3
Fig. 2. The title page of the Auctuarium manuscript in the University Library of Leiden depicts
a man, who might well be Rumphius, making notes under a remarkable fig. Below: roads of
Ambon in 1690.
4 Gard. Bull. Singapore 63(1 & 2) 2011
that ran the business was the Dutch Vereenigde Oost Indische Compagnie (VOC) (the
United East Indian Company), a state within a state with its own navy, army, fortresses,
trading posts, and diplomatic treatises with local leaders and political fractions. Its
settlements were the predecessors of the Dutch East Indies, but also of the Republic
of Ghana, and of South Africa, with settlements in many places, of note in the Arabian
coast, S India, Sri Lanka (think of cinnamon!), Malaysia, Japan, and through Taiwan
with China (think of the introduction of China ware, silk and tea in Europe!).
It is a sad thing that the identity of Rumphia amboinensis L. is unknown.
Linnaeus based this on a plate in Rheede’s Hortus malabaricus for a species from
South India, a place where Rumphius had never been and whatever the identity of
the species, it surely is not Amboinese. It just shows how little Linnaeus knew about
geography, sometimes his provenance “India” even refers to the West Indies...
Other generic eponyms in zoology are the gorgonian Rumphella Bayer (1955),
called ‘sea tree” by Rumphius, because at that time everything that moved was an
animal, and what didn’t was a plant. Linnaeus, also, regarded sponges (Spongia L.)
as algae... There are also the sea urchins Rumphia Desor (1846; Miocene to Recent of
the Indo-Pacific), Neorumphia Durham (1954), and Rumphiocrinus Wanner (1924).
However, the butterfly Rumphia Pagenstecher (1909) is a typographical error for
oO 5
Sawer?
vey an
>)
Db anda Mare,
Ambon.
Georgius Everhardus Rumphius 5
Ramphia Guinée (1852). The botanical journal Rumphia appeared between 1836 and
1849. There is even a Rumphius Range in the Lorentz Park in Indonesian New Guinea.
Rumphius was born in at the end of 1627 in Wolfersheim, Hessen, Germany.
His mother came from Cleve (Kleef) near the Dutch border, where at the time Dutch
was spoken, which would explain his impeccable and even innovative command of
that language. His father was an engineer, a builder of fortresses, and passed on this
knowledge to his son. Georg had a good education and finished the Gymnasium, but
did not go the University and did not get a degree in Medicine in nearby Hanau as was
later suggested.
As he said later he wanted to see something of the world and in 1646 enrolled
as a soldier, thinking that he would go the East Mediterranean, but instead he headed
for Barbice in the Guyanas where the West Indian Company (WIC) was involved
in wars with the Portuguese and Indians. There is still a saying in The Netherlands:
“going to the barbiesjes”, that is, to meet certain death. Fortunately for him (and us)
he somehow landed up in Portugal, where he served as a mercenary soldier for three
years after which he was dismissed, possibly because the Catholics there didn’t trust
Protestants like him. He returned to Germany for two years.
He may well have heard about the riches of the East Indies from his Dutch
relatives and wished to see them for himself. And thus at Christmas 1652 he left from
The Netherlands as an ade/borst (midshipman) with the VOC. Note that he would
have had a much higher rank if he had had an academic education and, especially,
when he had been a physician. These were very much sought after by the VOC in view
of the high rate of injuries, diseases and deaths that plagued the fleets and garrisons.
Later authors thought he was a medical doctor, but he denied this and in fact he was an
amateur naturalist and a self-made man in the best sense of the word.
By the end of 1653 he arrived in Pulau Ambon, never to leave the Moluccas
again. His father’s education now proved fruitful in the planning and construction of
fortifications. However, soon after 1657, he switched from the military to the civilian
and was appointed as Junior Merchant (onderkoopman) at Larike, on the West coast
of Hitu. There he married Susanna, a local woman, possibly Chinese, according to
European marriage records. He named the orchid F/os susannae after her, now Pecteilis
susannae (L.) Raf. “in memory of her who during her life was my first companion and
assistant in the finding of herbs and plants, she was the first to show me this flower”
(Fig. 4).
Larike was a back-water dump and because he had so little to do he could
devote a great deal of his time in the studies of the Treasuries of Nature. Rumphius
sent specimens to Europe which are not recorded in the VOC archives. This is not so
strange, as the VOC did its utmost to prevent exports and forbid private mailings. Of
course, everybody circumvented these rules. Rumphius is commended by all for his
honesty, but apparently he had his channels. You might say that he was less corrupt
than the others...
In 1701 he also smuggled out the manuscript of d’Amboinsche Rariteitkamer
(Fig. 5) to the mayor of Delft, Hendrik d’Acquet. It is interesting to note that some
drawings of shells were made by Maria Sybille Merian (1647-1717), who also hand
6 Gard. Bull. Singapore 63(1 & 2) 2011
Fig. 4. Flos susannae, Pecteilis susannae (L.) Raf. (Orchidaceae).
Georgius Everhardus Rumphius 7
coloured a few now priceless copies (Fig. 6). Where Rumphius was the founder of
zoology and botany in the Moluccas, she was that for Surinam.
He was elected as a member of what is now the oldest science society in the
world, the Academia naturae curiosorum of the German Roman Empire (founded in
1657, today still in existence as the Leopoldina). Members were given nicknames,
his was “Plinius”, a most honorific title as it was after the Roman procurator
(‘administrator’) Gaius Plinius Secundus (23-79 A.D.), killed in the eruption of the
Vesuvius which buried Pompeii and Herculaneum, and who was one of the founders
of European natural sciences. His influence lasted for 1500 years until the end of
the Middle Ages. Actually, Rumphius was more than Plinius, as the latter compiled
existing knowledge, while Rumphius did mention his sources, but added personal
experience. Blume, the second Director of the Kebun Raya Bogor, and then the first
Director of the Rijksherbarium, in his turn as a member, was called “Rumphius”, but
although a Medical Doctor from the Leiden University, he was a scientist of an entirely
different kind.
Thirteen of Rumphius’s letters were published in the Society’s journal
Miscellanea Curiosa. In 1682 he sent shells, sea animals, minerals, resins, fossils, and
some parts of plants in a cabinet made from 56 different kinds of wood to the Grand-
duke of Toscane, Cosimo III de’Medici (1639-1723). Unfortunately, the chest and
its contents have been lost, as the original labels have probably been replaced. Some
shells may be present in the State Museum in Vienna, and perhaps some fragments of
palms in Florence.
In 1660 he was promoted to Merchant (koopman) and moved to Hila, a much
more civilised place, where he lived like a prince: daily fresh venison, plenty of fish,
and sheep, geese, ducks, and chickens, horses with superb bridles, some of pure silver.
He had a one-gun vessel with 40 rowers. His gardens yielded cabbages, endives,
lettuce, parsley, Chinese radishes, etc. He even had a small zoo. Life was pretty good!
In 1662 he became Senior Merchant (opperkoopman). He then earned the
extra-ordinary salary of 24 rixdollars (60 guilders, or 1200 Dutch shillings) a month.
In comparison, Jan van Riebeeck, the famous Governor of the Cape, got “only”
21. He stayed there for 10 years, rather exceptional, as VOC employees usually
were translocated after about 5 years to prevent too good connections for graft and
smuggling. This may well be due to the good opinion his superiors had of his honesty.
In August 1663, he wrote a letter to the Board of the VOC in which he asked
for permission to have books sent for his work on the plants and animals of the East
Indies. This was granted, but only in 1666 did he receive them. One of his arguments
was, that God in His wisdom had provided local herbs to cure local diseases, so a better
knowledge of what was available would be beneficial to all. The medicines shipped
from The Netherlands were ineffective and often spoiled. He noted that although many
local recipes might be fables, superstitions, or old wives’ tales, they should be included
as there might be some truth in them. Of course he lacked the occasion and time to try
all medicines, but some he had “tested in mine own house, and with other families”.
; His accounts are not bone dry enumerations as was usual for herbals at the
time, but every species got an extensive description, where it occurred, what it was
8 Gard. Bull. Singapore 63(1 & 2) 2011
used for, a recipe, anecdotes, and sometimes a joke. For instance, that newcomers
(orang baru) were fooled with the resin of the pine Agathis, which they were told were
lumps of sugar, with which they then went into hiding to quietly and privately enjoy
the supposed delicacy. Of course he speaks much of diseases, health, birth, death, and
D AMBOINSCHE
RARITEITKAMER,
Behelzende eene BESCHRYVINGE van allerhande
z00 weeke als harde
SCHAAL ViSsUBEe.
te weeten raare
KRABBEN, KREEFTEN,
en diergelyke Zecdieren ,
als mede allerhande
HOORNTYES en SCHULPEN,
die men in d’Amboinfche Zee vindt:
Daar beneven zommige
MINERAALEN, CESTEENTEN,
en foorten van AARDE, dic in d’ Amboinfche , en zom-
mige omleggende Eilanden gevonden worden.
Verdeelt in drie Boeken,
En met nodige PRINTVERBEELDINGEN,, alle naar ’t leven getckent, voorzicn.
Befchreven door
GEORGIUS EVERHARDUS RUMPHIUS,
van Hanauw , Koopman en Raad in Amboina, mitsgaders Lid in d’ Academia Curioforum Nature ,
in ’t Duitfche Roomfche Ryk opgerccht, onder den naam van
PLINIUS INDICUS
TAMSTERD AM,
ee
Gedukt by FRANCOIS HALMA, Bockverkoper
in Konftantijn den Grooten.
176 5,
Fig. 5. Title page of the Rariteitkamer.
Georgius Everhardus Rumphius 9
all rituals concerned with these. A lot about sex, for that is part of life, but in a quiet,
unsensational way, sometimes with a chuckle on human foolishness.
In 1666, he was temporarily appointed as Secunde (“second man’), a rank
immediately under that of Governor, but was not confirmed in it. In compensation he
was allowed to buy a piece of land. Here he created a Physic Garden, the first western
type of botanical garden in Asia. He also had a forest garden long known as Dusun
Rumphius. His contract with the VOC after having served 16 years, expired in 1668
and he was supposed to retire (at 41!) and go back to Europe. He was quite happy
where he was and with what he was doing, so somehow he managed to extend his stay
with retention of his salary forever. Life was good!
Fig. 6. A rare coloured plate of the Rariteitskabinet by Maria Sybille Merian.
Blindness and personal tragedy
Things changed considerably when in April 1670, at 42, he turned incurably blind
(glaucoma simplex), about the worst thing that can happen to a dedicated naturalist.
With the aid of his son and some assistants provided by the VOC, he continued with
his work. The original manuscripts were in Latin, but because his assistants didn’t
know that language, he had to dictate them in Dutch. Here he showed his linguistic
proficiency. Descriptive botany is a language by itself and in his time it had hardly
10 Gard. Bull. Singapore 63(1 & 2) 2011
evolved, certainly not in Dutch. Consulting literature will have been a problem,
as these were of course in Latin and mainly dealt with the European flora. Works
on Asian plants were by the Portuguese Garcia de Orta (1501-1568) published by
Carolus Clusius (1567, 1582, 1593, 1605) from Goa 1n India, Hendrik Adriaan Rheede
van Drakenstein’s (1633-1691) 12-volume Hortus malabaricus (1678-1692) on
South Indian ones, and a medical textbook by Jacob Bontius (1592—1631) for plants in
Batavia. Later, Burman very carefully translated it all back again into Latin.
On Saturday, 17 February 1674, it was near the end of the Chinese New Year
celebrations. Rumphius didn’t attend, because, as he said, he couldn’t see anything.
Suddenly, there was a huge earthquake followed by tsunamis, killing 2322 people. A
falling wall killed Susanna, two of their daughters, and a maid.
On Saturday, 11 January 1687, in Kota Ambon, where he now lived, he had
another disaster. The town was razed by fire which destroyed his precious library,
collections, and most of his manuscripts. Only parts of the Herbarium amboinense and
about half of its plates were saved. Yet, undaunted by blindness and his awful losses,
Rumphius dictated the lost chapters to his assistants again from memory and they
managed to redraw the lost plates, probably under the supervision of his son Paulus.
By the end of 1690 the first part of his magnum opus, six books of the
Herbarium amboinense were sent to Batavia, where they were copied for safe keeping.
Not without reason, for it was a long and hazardous journey back to Europe. Indeed,
the Waterman that carried them in 1692 was sunk by the French. A second copy was
made including three additional books that had arrived in the meantime, and by August
1697 everything was safely in the Netherlands. A year later, the final three books were
received. Two additional appendices were shipped by 1704.
On May 19, 1702, the Governor of Ambon wrote to the High Government
at Batavia about Rumphius “nothing more was to be expected of that old gentleman,
having lived his years”, and on June 15 he died, 75 years old, very old for a European
in the tropics. No special mention is made in the missives from Ambon, but casually,
under another heading it is noted that he left 4000 rixdollars (10,000 guilders, or
200,000 Dutch shillings), quite a fortune, about two million Euro at present rates, and
various pieces of estate (land, houses), the savings of about 50 years of duty.
Subsequent events
If you think the VOC would have been quick to publish this magnum opus, you are very
much mistaken. Such a publication would be very expensive and would not provide
any profits, so the Board would allow it as long as it would not cost the Company a
penny. But there were no takers. Thus the manuscripts were locked up in their vaults.
Johannes Burman (1707-1779) was allowed by the VOC to prepare the
manuscripts for publication in 1735. He meticulously translated Rumphius’s poetic
and flowery Dutch back into Latin and had etchings made of the drawings and colour
plates. Take, for instance, the comparison Rumphius made of the wine-producing palm
Arenga pinnata Merr.: “and thus it resembles in its ugly and uncouth habit a drunken
Georgius Everhardus Rumphius 11
Fig. 7. Arenga pinnata Merr. as in the Herbarium amboinense (left) and in the manuscript of
the Leiden University (right).
farmer as he jumps up from his sleep with mended rags and tussled hair. Indeed, this
is the most ugly of all trees” (Fig. 7).
It was a very expensive publication, only 500 copies were printed, the set
costing about 100 guilders (2000 Dutch shillings, now Euro 20,000), about a third of
the income of a physician in Amsterdam. The original drawings often are in colour, but
the books were already so expensive, that printing with coloured plates would make
them completely out of reach. Thus it may happen that when Rumphius speaks of
several species it is not quite clear which one has been depicted. An example 1s Corona
ariadnes punicea with red flowers and Corona ariadnes lutea with yellow ones, but
the legend does not state with one is represented. The coloured plate in the Leiden
University Library shows the flowers to be red, so after more than 260 years we now
know that the plate represents the punicea form, a synonym of Hoya coronaria Blume
(Fig. 8). Similar problems may well be solved so easily.
Also, there have been only few attempts to recollect representative material.
Most were only half-heartedly done, or the scientists died before arriving at the island.
The most serious attempt was by Charles Budd Robinson (1871—1913), an American
botanist sent there by Merrill, who made a large collection of Rumphian and non-
12 Gard. Bull. Singapore 63(1 & 2) 2011
Sap
Fig. 8. Hoya coronaria Blume (Asclepiadaceae) as in the Herbarium amboinense (left) and in
the manuscript of the Leiden University (right).
Rumphian species, but was prematurely murdered. Merrill distributed 18 duplicate
series and by these many Rumphian taxa can be identified and the names exclusively
based on them, neo- or epitypified (when the type is the illustration). The series have
been deposited in A, BM, BO, F, K, LMO, NSW, NY, US. The top set with the original
labels are in US, so in typifications these should be designated as the holotypes.
Anyway, these texts after so many years are still the only extensive source
on the flora of Ambon. It is therefore remarkable that although there are various
concordances, there has never been a reprint. Fortunately, the original texts can now be
found on the internet, but unfortunately, very few have sufficient knowledge of Dutch
and Latin to be able to read them. The good news now is that Monty Beekman (1939—
2008) just before his death was able to finalise a translation into English which was
released to the public on 20 June 2011, but was on 4 and 5 February 2011 officially
presented to his widow, Faith Foss, by the Yale University Press at the Fairchild
Botanic Gardens in Florida. This will make this seminal publication available to the
Anglophone public and thus it will be a very valuable addition to the knowledge of
the past, present, and future events of the Flora Malesiana. Of course the work has
been extensively data mined. You will find references in Heyne’s Useful Plants of the
Dutch East Indies, sometimes as the only reference. From there they found their way
to Burkill’s Dictionary of the Economic Products of the Malay Peninsula and, more
recently, into the publications of PROSEA.
oS)
Georgius Everhardus Rumphius ]
Epitaph
Monuments have been raised for Rumphius, the first destroyed by tomb-robbers, the
second (erected in 1824) hit by an Allied bomb in 1944, and a third was built in a
slightly different place (yard of the Xaverius Junior High School, Jl. Pattimura) in
1996 (Buijze 2001: 282, fig.). | was in Ambon in April 2011 and visited this monument
(Fig. 9). His house, however, was burned early 20th century. But no human activity
can destroy the true monuments reminding us of this remarkable man: the works on
plants, animals, and stones that he has left us.
Fig. 9. The Rumphius monument in the yard of the Xaverius Junior High School, Kota Ambon.
A Homo universalis: modest, unprepossessing, lenient to other people’s
views, an unbelievable resilience to disasters (blindness, loss of family, life work),
perseverance under stress, yet with a persistent sense of humour—where this might
easily and understandably have led to bitterness—with a continuous perspective
curiosity, a perpetual surprise. We would do well to make him our example.
ACKNOWLEDGEMENTS. Special thanks are due to Dr. A. van de Beek (Amsterdam),
Dr. P.D. Bostock (BRI), Mr. W. Buijze (The Hague), Ms. J. de Roode, Mssrs J. Cramer, J.
Frankhuizen, and R.B.P Rijkschroeff of the University Library, Leiden, Dr. J. Dransfield (K),
14 Gard. Bull. Singapore 63(1 & 2) 2011
Dr. C.E. Jarvis and Dr. N.K.B. Robson (BM), Dr. D. Mabberley (L), Ms. K. Pocock of Yale
University Press, London, Mr. O.P. van Zandwijk, Leiden, and Dr. G. Zijlstra (U) for various
bits of information and technical help.
References
Beekman, E.M. (1999) The Amboinese Curiosity Cabinet, pp. xxxv—cxi1. New Haven.
Buijze, W. (2001) De Generale Lant-Beschrijvinge van het Ambonse Gouvernement
ofwel de Ambonsche Lant-Beschrijvinge door G.E. Rumphius, Transcriptie,
Noten, Woordenlijst en een Nieuwe Biografie. Amsterdam.
Rheede tot Drakenstein, H.A. van (1678-1703) Hortus Indicus Malabaricus. 12 vols.
Amsterdam.
Appendix A. Other works recommended for further reading.
Baas, P. & Veldkamp, J.F. (2011) Dutch pre-colonial botany and Rumphius's Amboinense
Herbal. Allertonia (in press).
Backer, C.A. (1936) Verklarend Woordenboek, p. 568. Batavia: Groningen (reprinted in 2000,
Amsterdam).
Ballintijn, G. (1944) De Blinde ziener van Ambon. Utrecht.
Beekman, E.M. (1981) The Poison Tree. Selected Writings of Rumphius on the Natural History
of the Indies, pp. 1-40. Amherst.
Bontius, J. (1642) De Medicina Indorum. 4 vols. Leiden.
Buiyze, W. (1998a) The True History of the Terrible Earthquake of ... 1674 ... in and around the
islands of Amboina. The Hague.
Buiyze, W. (1998b) Antwoort en rapport op enige pointen uijt name van seker Heer in ‘t Vaderlant
voorgestelt door d’Edele heer Anthonij Hurt, directeur generaal over Nederlants’ Indien.
The Hague.
Burman, J. (1736) ((1737’). Thesaurus Zeylanicus. Amsterdam.
Burman, J. (1755) Het Auctuarium. Index (unpaged). Amsterdam.
Cribb, P.J. (1999) Orchis Susannae L. In: Cafferty, S. & Jarvis, C.E. (eds) Typification of
Linnaean specific and varietal names in the Orchidaceae. Taxon 48: 49.
Greshoff, M. (ed) (1902) Rumphius Gedenkboek 1702-1902, pp. 57-58. Haarlem.
Hamilton, F. (1824) Commentary on the Herbarium Amboinense. Mem. Wern. Nat. Hist. Soc.
5(2): 307-383.
Hamilton, F. (1832) Commentary on the second book of the Herbarium Amboinense. Mem.
Wern. Nat. Hist. Soc. 6: 286-333.
Hasskarl, J. (1866) Neuer Schlussel zu Rumph’s Herbarium Amboinense. Abh. Naturf. Ges.
Halle 9: 145-389.
Henschel, A.W.E.T. (1833) Clavis Rumphiana Botanica et Zoologica. Breslau.
Hermann, P. (1689) Paradisi Batavi Prodromus. In: Sherard, W. (ed) Schola Botanica Sive
Catalogus Plantarum, p. 358. Amsterdam.
Hermann, P. (1698) Paradisus Batavus, p. 209, t. 209. Leiden.
Hickson, S.J. (1926) The life and work of Georg Everard Rumphius. Manchester Mem. 70(2):
17-28.
Georgius Everhardus Rumphius 15
Kellogg, E.A. (1996) In: Kuijt, J. & Kellogg, E.A., Miscellaneous mistletoe notes, 20-36.
Novon 6: 35.
Leupe, P.A. (1871) Georgius Everardus Rumphius, Ambonsch natuurkundige der 17e eeuw.
Verh. Kon. Akad. Wetensch. 12(3): 1-63.
Linnaeus, C. (1736a) Fundamenta Botanica. Amsterdam.
Linnaeus, C. (1736b) Bibliotheca Botanica. Amsterdam.
Linnaeus, C. (1753) Species Plantarum. Stockholm.
Lotsy, J.P. (1902) Over de in Nederland aanwezige botanische handschriften van Rumphius. In:
Greshoff, M. (ed) Rumphius Gedenkboek 1702-1902, pp. 57-58. Haarlem.
Martelli, U. (1902) Notizie sopra l’Erbario Rumphio. Boll. Soc. Bot. Ital.: 90.
Martelli, U. (1903) Le collezioni di Georgio Everardo Rumpf acquistate dal Granduca Cosimo
IIT de Medici. Florence.
Meeuse, B.J.D. (1965) Straddling two worlds: a biographical sketch of Georg Everhard
Rumphius, Plinius indicus. Biologist 47(3&4): 42-54.
Merrill, E.D. (1914) Charles Budd Robinson, Jr. Philipp. J. Sci., Bot. 9: 194-195.
Mermill, E.D. (1916) Reliquiae robinsoniae. Philipp. J. Sci., Bot. 11: 243-249.
Merrill, E.D. (1917) An interpretation of Rumphius’ Herbarium Amboinense. Bur. Sci. Publ.
9: 1-595.
Nicolson, D.H. & Arculus, D. (2001) Candidates for neotypification of Blanco’s names of
Philippine plants: specimens in the U.S. National Herbarium. Jaxon 50: 947-954.
Nicolson, D.H., Suresh, C.R. & Manilal, K.S. (1988) An interpretation of Van Rheede’s Hortus
Malabaricus. Regn. Veg. 119: 319.
Radermacher, J.C.M. (1780-1782) Naamiijst der planten die gevonden worden op het eiland
Java. 3 vols. Batavia.
Royen, A. van (1740) Florae Leydensis Prodromus. Leiden.
Rumphius, C.E. (Burman, J. (ed.)) (1741-1755) Herbarium Amboinense. 7 vols. Amsterdam,
*s Gravenhage, Utrecht.
Rumphius, C.E. (1686) De ligno Lacca dicto, Funibus sylvaticis, resina Dammar Selan, &
Dammar Batu. Misc. Cur. Ephem. Med.-Phys. German. Acad. Nat. Cur. 2: 76-77.
Rumphius, G.E. (1675) Waerachtigh Verhael van de Schrickelijke Aerdbevinge. Batavia.
Stearn, W.T. (1959) Introduction. In: Linnaeus, C., Species Plantarum, ed. 1, facsimile edition.
Ray Soc. 142: 106. London.
Steenis-Kruseman, M.J. van (1950) Rumphius. Flora Malesiana 1, 1. Jakarta.
Stickman, O. (1754) Herbarium Amboinense. Uppsala.
Uggla, A.H. (1937) Linne och Burmannerna. Sv. Linne-Sallsk. Arsskr. 2: 131, 134.
Valentijn, F. (1724-1726) Oud en Nieuw Oost-Indien. 8 vols. Dordrecht, Amsterdam.
Veldkamp, J.F. & Laubenfels, D.J. de (1984) Proposal to reject Pinus dammara (Araucariaceae).
Taxon 33: 337-347.
Vogel, E. de, Schuiteman, A., Feleus, N. & Vogel, A. (1999) Catalogue, part 1. 1998:
Orchidaceae, p. 10. Leiden.
Warburg, O. (1902) Die botanische Erforschung der Molukken seit Rumpf’s Zeiten. In:
Greshoff, M. (ed) Rumphius Gedenkboek 1702—1902, pp. 63—78. Haarlem.
Wit, H.C.D. de (1952a) In memory of G.E. Rumphius (1702-1952). Taxon 1: 101-110.
Wit, H.C.D. de (1952b) “Rumphius herdacht (1702-1952)'. Madj. IImu Alam untuk Indonesia
108: 161-172. (Emended translation of the preceding.)
Wit, H.C.D. de (ed) (1959a). Rumphius Memorial Volume. Baarn.
Wit, H.C.D. de (1959b) Georgius Everhardus Rumphius. In: Wit, H.C.D. de (1959a): 1-26.
Baarn.
ye
. os Gacoamall)
% fan. aes
no
aeh)-
oo Fe
i
* ca
i+
1? ‘<<
. Ad
Ta ©
hte *
: ew
@
* amma sh
‘ pry Ta)
Par jee @,
iS
re
rio! @
Gardens’ Bulletin Singapore 63(1 & 2): 17-30. 2011 17
Revision of the rattan genus Daemonorops
(Palmae: Calamoideae) in Sulawesi
using a phenetic analysis approach
H. Rustiami '**, J.P. Mogea' and S.S. Tjitrosoedirdjo?
‘Herbarium Bogoriense, Botany Division, Research Center for Biology.
Indonesian Institute of Sciences, Cibinong Science Center (CSC),
Jl. Raya Jakarta-Bogor Km 46, Cibinong, Bogor 16911, Indonesia
> Bogor Agricultural University, Bogor and South East Asian Regional Center for
Tropical Biology (SEAMEO BIOTROP) P.O. Box 116, Bogor, Indonesia
*himmah@hotmail.com (corresponding author)
ABSTRACT. A phenetic analysis based on 27 morphometric characters of seven species of
Daemonorops in Sulawesi recovered two groups with a similarity coefficient value of 0.51.
Group A consists of D. takanensis and D. lamprolepis with a similarity coefficient value of
0.58. Group B is divided into subgroup B1 and subgroup B2, with a similarity coefficient value
of 0.59. Group B1 consists of D. macroptera, D. mogeana and D. robusta. Group B2 consists
of D. riedeliana and D. sarasinorum. An identification key to species and their descriptions are
presented.
Keywords. Calamoideae, Daemonorops, Palmae, phonetic analysis, rattans, Sulawesi
Introduction
The palm flora of Sulawesi is distinctive and combines elements in common with
Sunda, Sahul, the Philippines, and the Papua New Guinea. In the case of Daemonorops,
all seven species recognised are endemic to the island and their affinities are not yet
clear — whether with Sunda, Philippines or East Malesia. The genus Daemonorops
itself is not well collected and poorly represented further east. Until recently, five
species of Daemonorops were recorded for Sulawesi. As a result of recent fieldwork, a
further two species have been recognised and described (Rustiami 2009). The purpose
of this study to investigate morphological variation within Daemonorops in Sulawesi
using a phenetic analysis of morphological data taken from herbarium specimens. This
is done as a precursor to a treatment of the genus for Flora Malesiana.
The genus Daemonorops was described by Blume (1830), based on a single
species which he named Daemonorops melanochaetes Blume. Daemonorops with
more than 120 species is the second largest rattan genus after Ca/amus. It belongs
to the subtribe Calaminae, tribe Calameae of the palm subfamily Calamoideae.
Beccari (1911) divided Daemonorops into two sections based on the structure of
the inflorescence, i.e., section Cymbospatha and section Piptospatha. Basically, the
former have concave boat-shaped bracts which are at anthesis completely enclosed
18 Gard. Bull. Singapore 63(1 & 2) 2011
by the prophyll (the first bract) and splitting longitudinally to expose the flowers. In
contrast, the bracts of the species in the latter section split to the base and only the
lower part is enclosed by the prophyll. Of the 84 species identified, Beccari placed 32
species in the former section and 52 species in the latter. According to Furtado (1953)
the bracts of section Piptospatha usually fall at anthesis and occasionally only the
prophyll remains.
The name Daemonorops is derived from the Greek language combining
two words daemon (devil) and rops (bush or shrub). This name reflects the fearsome
appearance of the plant with its very robust leaf sheaths densely armed with long,
blackish brown or cream-coloured spines (Dransfield et al. 2008).
Material and methods
Field work was carried out in several areas of Sulawesi to collect herbarium material.
Herbarium specimen preparation followed the standard procedure of Dransfield (1986).
Data or information recorded from the field include location; general habitat; altitude;
association with other plant; vernacular name(s); uses; habit (solitary vs. clustered);
stem attributes (height, diameter with/without leaf sheath, internode length, colour);
characteristics of the leaves (length, leaflet arrangement, number of leaflets, length and
width of leaflets); inflorescence characteristics (length, number of rachilla, colour);
flower attributes (colour, scented/not scented); and fruit and seed characteristics
(length and width, colour).
Field work data was combined with herbarium data for each taxon to obtain a
more comprehensive set of morphological data. Morphological studies were carried out
with specimens in several herbaria: Herbarium Bogoriense (BO), the Kew Herbarium
(K) and the Leiden Herbarium (L).
A total of 300 herbarium specimens were studied following Vogel (1987) and
Rifai (1976), using comparative morphological data as the main source of evidence in
developing the species concept (Dransfield 1999).
Twenty seven morphometric characters were chosen (Table 1). These were
scored using two simple states (absent — 0: present — 1). The data processing was carried
out using the NT-Sys program 2.1 (Rohlf 1997). The descriptions of each species
and identification key for Daemonorops were constructed based on the characters and
character-states recorded.
Results and discussion
Fig. | shows the result of the phenetic analysis, in which seven species of Daemonorops
are differentiated morphologically based on twenty seven characters. The specimens
are Clearly divisible into two groups (A and B) with a coefficient similarity value of
0.51. This value means that these two groups only have morphological similarity of
around 51%. Group A consists of D. takanensis and D. lamprolepis where these two
Daemonorops in Sulawesi 19
Table 1. Morphometric characters used in the analysis.
1 Leaf sheath indumentum 15 Leaflets arranged regularly
2 Leaf sheath surface scales 16 Leaflets arranged subdistantly
3 Leaf sheath armed with short easily 17 Leaflets arranged distantly
removed spines
- Leaf sheath armed with brittle 18 Transverse veinlets present
unequal, solitary spines
5 Leaf sheath armed with hair like 19 Corolla same size as the calyx
spines
6 Leaf sheath armed with long, 20 Corolla longer than the calyx
strongly attached spines
aT Leaf sheath armed with large, 21 Fruit spherical
iregularly seriate spines
8 Ocrea present 22 Fruit subglobose
9 Leaf sheath armed with short, 23 Fruit ellipsoid
scattered, seriate spines
10 Direction of spines on leaf sheath 24 Seed surface smooth
horizontal
1] Direction of spines pointing upward 25 Seed surface reticulate
12 Knee armature present 26 Endosperm slightly ruminate
13 Leaf sheath mouth spiny 27 Endosperm deeply ruminate
14 Petiole with indumentum
species have a coefficient similarity value of 0.58. Group B divided into two sub-
groups, B1| and B2, with a coefficient similarity value of 0.59. From the phenogram
we can see that D. macroptera and D. mogeana have a morphological similarity of
around 81%. Those two species are close to D. robusta, with 67.4% similarity. In the
other group we can see that D. riedeliana and D. sarasinorum have a similarity of
around 77.4%.
Daemonorops mogeana, D. macroptera and D. robusta clustered in one group.
This is because they have some similarities in their leaf sheath armature and leaflet
arrangement. However they do differ in their general morphological appearance.
The stem of Daemonorops is covered by tightly sheathing, densely spiny, leaf
sheaths. The diameter of the stem with the leaf sheaths can vary from a few mm to
over 10 cm. Leaves consist of a tubular sheathing base, the leaf sheath, which arises
from the node on the stem; at its upper end, the sheath narrows into the petiole that
continues into the rachis or leaflet-bearing portion of the leaf. Although a petiole
is usually present, it is sometimes very short or absent. In many species, the rachis
is extended beyond the terminal leaflets into a barbed whip (cirrus) which acts as a
20 Gard. Bull. Singapore 63(1 & 2) 2011
A D. lamprolepis
D. takanensis
@! D. macroptera
B I D. mogeana
B D. robusta
B 9} D. riedeliana
D. sarasinorum
0.51 0.59 0.66 0.73 0.8
Coeficient
Fig. 1. Phenogram of morphological similarity among Daemonorops spp. from Sulawesi.
climbing organ (Dransfield & Manokaran 1994). Spine arrangement on the leaf sheath
is remarkably diverse and frequently of diagnostic importance. Just below the petiole
or leaf rachis, there is a marked swelling known as the knee. This character is also of
some diagnostic importance, because some of the Daemonorops species do not have
very obvious knees or the knee 1s only slightly developed.
As with most rattans, Daemonorops are dioecious, that is, female and male
flowers are borne on different plants. The main axis bears a basal bract or prophyll,
which may be short and tubular, or large. Branches are borne in the axils of subsequent
bracts. The branches in turn bear bracts, the lowermost of which is usually empty,
subsequent bracts subtending branches, and so on. The ultimate flower-bearing
branches are termed partial inflorescences. On each female flower, there is a bracteole
which immediately surrounding the flower, known as an involucre and an outer bract
known as the involucrophore. Flowers are borne in dyads with two bracteoles in the
female and in the male flowers are solitary with one bracteole.
Taxonomic treatment
Daemonorops Blume in J.A. & J.H. Schultes, Syst. Veg. 7(2): 1333 (1830).
Solitary or clustering rattans, acaulescent to high-climbing hapaxanthic (then always
very short-stemmed) or pleonanthic, dioecious. Sheaths usually heavily armed with
spines, the spines frequently highly organised. Flagellum absent. Knee frequently
present. Leaves ecirrate in acaulescent species or longly cirrate. Leaflets variously
arranged. Inflorescence male and female superficially similar, but within the genus of
Daemonorops in Sulawesi 21
two basic types: one with all bracts enclosed within the outermost bract or prophyll,
splitting along their length to expose the flowers (section Cymbospatha) or the
other with bracts splitting along their entire length to leave no tubular portion and
frequently falling (section Piptospatha). Bracts variously armed. Partial inflorescences
longer than the subtending bract in section Piptospatha: bracteoles and “involucres”
inconspicuous. Male rachilla bearing male solitary flowers, male flowers small, cup-
shaped; calyx with three small lobes: corolla split to the base into 3 petals: stamen 6,
slightly epipetalous; pistillode minute. Sterile male flower found with each female
flower, as the fertile male, but stamens with empty anthers. Female rachilla bearing
many flowers in dyads consists of one female flower and one sterile male flower. Female
flower with calyx truncate or shallowly 3-lobed: corolla with 3 petals; gynoecium with
3 stigmas and with 3 loculi. Sterile flower smaller or at least more slender than the
female ones, with well-formed calyx and corolla and 6 sterile stamens and an abortive
ovary. Fruit variously shaped, tipped with stigmatic remains and covered with reflexed
scales. Seed only one, covered by thin to thick, sweet or sour sarcotesta. Endosperm
deeply ruminate. Embryo basal.
Distribution. Based on Dransfield et al. (2008), the geographical distribution of
Daemonorops is more restricted than Calamus. The centre of distribution of both are
similar, from China, India to New Guinea, specifically Sumatra, Malaya, Borneo and
Malay Peninsula. Daemonorops does not occur in Africa, Himalaya, Peninsular India,
Sri Lanka and Australia.
Habitat. Rather disturbed primary forest, on alluvial soil near rivers, flat to gently
sloping terrain, ridge tops, lowland forest, and steep hill slopes in primary forest on
volcanic soils.
Uses. One species is recorded to have sweet, edible young shoot (Mogea 1991).
Key to Daemonorops species in Sulawesi
la. Leaf sheath covered with rusty-brown coloured indumentum and armed with short,
up to 10 mm long, easily detached spines ....................eeeeeeeeseeeeeees D. takanensis
1b. Leaf sheath without indumentum and armed with long, more than 15 mm long,
Fan AST RI AER AL ANCSUD: SES INNS eee a ore ee SE SE NN or) hy Sees a tee 2
2a. Leaf sheath armed with brittle, unequal, solitary (or groups of) spines .............. 3
2b. Leaf sheath strongly armed with large, irregularly seriate spines .......................- 5
3a. Leaf sheath armed with solitary, black, brittle spines; ocrea present .....................
<u Sk eee eNO ba ce nee D. lamprolepis
3b. Leaf sheath armed with solitary (or groups of) black spines: ocrea absent ......... -
i)
i)
Gard. Bull. Singapore 63(1 & 2) 2011
4a. Leaf sheath densely armed with very long, solitary, hair-like spines .....................
LE LE Be OE OSE ee D. sarasinorum
4b. Leaf sheath armed with short, scattered, seriate, needle-like spines ...................06.
Sa. Leaf sheath armed with oblique spines joined at their bases; fruit spherical,
endospermdeeplhy mumntmate eei00e esc ceeescase sees tare seen eee cee D. robusta
Sb. Leafsheath armed with upright spines that are solitary or joined at their bases; fruit
subglobose to ellipsoid, endosperm slightly to deeply ruminate .................008 6
6a. Leaf sheath densely armed with solitary, furfuraceous ae fruit ellipsoid,
endosperm deeply ruminate .. cnet seteaosae .. D. macroptera
6b. Leaf sheath densely armed swith Btn 6 spines in groups lot 3 s to 5’s; fruit
subglobose; endosperm slighthy mumiimate Sse). ere eevee eee D. mogeana
1. Daemonorops lamprolepis Becc., Rec. Bot Surv. Ind. 2: 223 (1902). TYPE: South
East Sulawesi, Kendari, July 1874, Beccari s.n. (holo BO).
Clustering rattan. Sheathed stem up to 2 cm in diam., stem without sheath up to 1 cm in
diam. Leaf sheaths green, covered with collar spines with jointed bases, scarcely up to
3 cm long, sheath surface smooth with caducuous reddish-blackish scaly indumentum,
leaf sheath mouth armed as the rest of sheath; knee present, very conspicuous, armed
as the rest of sheath; ocrea present, papery, small, to 5 mm high. Leaves to 3 m long
including petiole 30 cm long, armed adaxially with short, erect, scattered spines
to 2 mm long, abaxially armed with erect, very rarely solitary spines, up to | mm
long; rachis unarmed, or armed only slightly proximally; cirrus up to 80 cm long,
armed with regularly arranged groups of grapnel-like spines, leaflets mostly arranged
regularly, 30 on each side of the rachis, stiff, horizontal; leaflets lanceolate, papery,
acuminate, up to 30 cm long, 2 cm wide, armed with scattered reddish, short bristles
along the main nerve on lower surface, transverse veinlets conspicuous. Female
inflorescences pendulous to 37 cm long, peduncle 10-15 cm long, armed distally with
groups of spines; prophyll papery, erect, 25 cm long, 3 cm wide, ellipsoid oblong,
armed with scattered spines, some spines in groups of 2’s; partial inflorescences up
to 4, each inflorescence bearing up to 8 partial inflorescences; rachilla covered with
chocolate scurf; involucre pendulous, flat, just above the involucrophore, 5 mm long;
involucrophore short, papery, 2 mm long. Female flowers 6 mm long, ovoid, acute;
calyx very short; the corolla several times longer than the calyx, ventricose at the base.
Male inflorescence and male flowers unknown. Young fruits ovoid to ellipsoidal, 15
< 10 mm, covered by 8—9 vertical rows of encrusted scales. Seed ovoid, 10 * 7 mm,
boldly tubercled and coarsely pitted. Endosperm ruminate.
Daemonorops in Sulawes1 23
Distribution. Donggala, Central Sulawesi and Kendari, South East Sulawesi.
Habitat and ecology. Disturbed primary forest.
Vernacular names. Rotan mapis (Donggala language), /asero epe or lita (Tobelo
language).
Notes. This is the only species of Daemonorops from Sulawesi which has an ocrea.
This ocrea is papery, small, to 5 mm high.
Specimens examined: Central Sulawesi: Northern central part, on the coast of South
West of Donggala, 11 May 1975, W. Meijer 10086, fruiting (BO). Mountain Sojo,
November 1913, Rachmat 705, fruiting (BO). South Sulawesi: Maliki, Desoe, 02 Jun
1933, H.N. Reppie 18, sterile (BO). Wadjo, Heyne 2581, Heyne 2587, sterile (BO);
Heyne 2615, fruiting (BO); Bom, Heyne 2599, young fruit (BO); Heyne 2595, Heyne
2604, fruiting (BO); Heyne 12, dead female inflorescence (BO).
2. Daemonorops takanensis Rustiami, Reinwardtia 13(1): 25-30 (2009). TYPE:
Indonesia, South Sulawesi, Kab. Mamuju, District Kaluku, Dusun Roa, Rantai Village,
Kaluak, Bukit Takane-kane, 200 m alt., 06 February 1993, Padmi Kramadibrata 28,
fruiting specimen (holo BO).
Slender, clustering rattan, climbing to 20 m. Sheathed stem 2 cm in diam., without
sheaths 1.5 cm in diam., internodes 20-30 cm long; leaf sheath dark green, covered
with conspicuously rusty brown-coloured indumentum and armed with numerous very
brittle, thinly laminar, unequal, up to | cm long or even shorter, solitary, scattered,
easily to detached, brown spines, with small bulbous bases; leaf sheath mouth densely
armed with similar spines; knee present and conspicuous, 10 mm long, 20 mm wide,
moderately armed. Leaves 3.5 m long including petiole and cirrus; petiole to 20 cm
long, 10 mm wide and 8 mm thick at base, flat adaxially, rounded abaxially, with
acute edges, covered slightly with rusty brown indumentum, as on sheath, armed with
numerous short triangular spines; rachis up to 1.8 m long, armed with very short,
erect, slender, triangular claws, that become ternate near the apex and 5-nate and half-
whorled on the cirrus; cirrus to 150 cm long; leaflets numerous, 55 pairs on each
side of rachis, regularly arranged, linear-lanceolate, acuminate, armed with bristles
to 5 mm long along the midrib of both surfaces; transverse veinlets minute; basal
leaflets 34 cm long and 8 mm broad, middle leaflets 35 cm long and | cm broad, apical
leaflets to 20 cm long and 8 mm broad. Male and female inflorescences not known.
Infructescence pendulous, up to 50 cm long, consisting of 4 partial infructescences, 5
cm apart; peduncle 10 cm long; partial infructescence to 8 cm long bearing 10 partial
inflorescences. Fruit ellipsoid with a short conical beak, pale, covered with 15 vertical
rows of scales, 15 mm long and 10 mm broad. Seed one, ellipsoid. Endosperm deeply
ruminate.
24 Gard. Bull. Singapore 63(1 & 2) 2011
Distribution. Known from the type locality only.
Habitat and ecology. Disturbed primary forest on hill slope.
Uses. Not recorded.
Vernacular name. Rotan api.
Notes. This species can be recognised easily by its dark green leaf sheath, covered
with conspicuously rusty brown-coloured indumentum and armed with numerous very
brittle, thinly laminar, unequal, up to 1 cm long or even shorter, solitary, scattered,
easily to detached, brown spines, with small bulbous bases. So far this species is only
known from the type locality, Bukit Takane-kane.
Specimens examined: South Sulawesi: Kab. Mamuju, District Kaluku, Dusun
Roa, Rantai Village, Kaluak, Bukit Takane-kane, 200 m alt., 06 Feb 1993, Padmi
Kramadibrata 028, fruiting specimen (BO).
3. Daemonorops macroptera (Miq.) Becc., Rec. Bot. Surv. Ind. 2:223 (1902); Calamus
macropterus Mig., Verh. Kon. Akad. Wetensch., Afd. Natuurk. 11 (5): 19 (1868);
Palmijuncus macropterus (Mig.) Kuntze, Revis. Gen. Pl. 2: 733 (1891). TYPE: North
Sulawesi, Manado, Minahasa, Riedel IGF s.n. (holo BO; iso L).
Clustering robust rattan, up to 40 m tall. Leaf with sheaths up to 3 cm in diam., without
sheaths to 2 cm in diam., sheaths covered with basally joined robust spines, up to 5 cm
long, leaf sheath mouth armed as the rest of the sheath; knee present conspicuously,
armed as the rest of the sheath. Leaves up to 6 m long including petiole to 40 cm,
armed with groups of robust spines, to 2 cm long, on both surfaces; rachis armed
with scattered, solitary spines up to | cm long; cirrus more than 2 m long, armed
with regularly arranged groups of very robust grapnel-like spines, blackish at the tip;
leaflets mostly arranged regularly, slightly irregular apically, 70 on each side of the
rachis, stiff, horizontal; leaflets lanceolate, papery, acute, up to 55 cm long, 3 cm wide,
armed with scattered, reddish, short bristles along the main nerve on lower surface, up
to 1 cm long, short bristles along the leaflets margin; transverse veinlets very minute,
and sharp. Male inflorescence pendulous, up to 85 cm long including peduncle 25—
30 cm long, peduncle straight and rigid, flattened, densely armed with flat, irregular,
erect, spreading, 1-2 cm long spines; the outer bract is narrowly lanceolate before
flowering; after flowering it is coriaceous, gradually narrow to acuminate, covered
with furfuraceous indumentum; rachilla about 40 cm long, with 5 small partial
inflorescences. Male flower small, 4-5 mm long; calyx very small, deeply 3-dentate.
Female inflorescence elongate, rather slender, pendulous up to 65 cm long, bearing
6~7 partial inflorescences; secondary spatha short, acute or acuminate, up to 8 cm long,
covered with rusty indumentum. Female flower unknown. Infructescence pendulous,
Daemonorops in Sulawesi a5
to 60 cm long, peduncle up to 15 cm long, armed distally with groups of robust spines;
peduncular bracts leathery, erect 25 cm long, 3 cm wide, ellipsoid oblong, covered by
rusty indumentum, armed with solitary spines up to 2 cm long, partial inflorescences
5 each, bearing up to 9 partial inflorescences; involucre pendulous, flat, just above the
involucrophore, 5 mm long; involucrophore short, papery, 2 mm long. Fruits obovoid,
15 x 15 cm, covered by 7 vertical rows of encrusted scales. Seed ovoid, 10 x 10 mm,
smooth surfaces.
Distribution. North and Central Sulawesi.
Habitat and ecology. Rather disturbed primary forest, on alluvial soil near river, terrain
flat to gently sloping.
Vernacular name. Rotan batang, angah.
Notes. This rattan, based on Vogel’s field record, produces a white gummy exudate
from the cut stem and the immature fruit is green. It is found at low elevations along
river terraces (field note of Musser T10).
Specimens examined: North Sulawesi: Manado, Miquel s.n., fruiting (BO, L). Central
Sulawesi: Sopu Valley, c. 80 km SSE of Palu, 1000 m alt., 2 May 1979, E.F. de Vogel
5171, fruiting (BO); 22 May 1979, E.F. de Vogel 5508, fruiting (BO); 26 Apr 1979,
1000 m asl., E.F de Vogel 5055, fruiting (BO, K); Mountain Rorokatimbu, 13 May
1979, E.F. de Vogel 5326, fruiting (BO, K); Poso, Lore Utara, Gn. Pada Esa, 11 Sep
2010, 1525 masl., Himmah Rustiami, Dewi, M. Amir, Hamzah & Ato HR 447, fruiting
material, Mt. Petulu, Kulawi, 18 Feb 1986, 700 m asl, Anggana & Yusuf Dali 62,
sterile (K); Sungei Tolewonu, 30 km South of Kuala Navusu, between 1974-1976,
G.G. Musser T10 (K). South East Sulawesi: Tongoa, 730 m asl., 4 Mar 1981, J. Th.
Johansson, H. Nybom & S. Riebe 169 (K).
4. Daemonorops mogeana Rustiami, Reinwardtia 13(1): 25-30 (2009). TYPE:
Indonesia, Central Sulawesi, Kab. Poso, District Kulawi, Dusun Moa, Mt. Malemo,
1000 m alt., 21 October 1977, JP Mogea 1356, fruiting specimen (holo BO; iso K, L).
Very large, robust, clustering rattan, climbing to 15 m. Sheathed stem 4 cm in diam.,
stem without sheaths 2 cm in diam.; internodes 20 cm long. Leaf sheaths woody,
creamy-yellow, densely armed with numerous broad spines often with conspicuous
bulbous bases, and arranged in groups of 3’s to 5’s, flat, greyish, irregularly seriate,
1-7 cm long, 5 mm wide, intermixed with smaller and ascendant spines. Leaves very
large, up to 6 m long including petiole and cirrus; petiole very robust, 1 m long, 2
cm wide and | cm thick at base, rounded adaxially and abaxially, densely armed
with, seriate or irregularly, erect, triangular, 1-3 cm long and up to | cm wide spines;
rachis up to 3 m long, with similar triangular spines; leaflets large, 30 pairs on each
26 Gard. Bull. Singapore 63(1 & 2) 2011
side of rachis, regularly arranged, linear-lanceolate, acuminate, armed with small
bristles, 5 mm long along the mid nerve on both surfaces and the apex; transverse
veinlets conspicuous; middle leaflets 40 cm long; 2 cm broad, papyraceous, green
and concolorous; apical leaflets to 20 long, 1.5 cm broad; cirrus to 2 m long, armed
with 4-5 hooked grapnels arranged 3 cm apart. Male and female inflorescences not
known. Infructescence ascending, to about 70 cm long, with 6 erect, very slender,
cupressiform, partial infructescences, 5 cm apart; the main axis cylindrical, 20 cm
long, armed with dense, glaucous, seriate spines, about 1—S cm long, with bulbous
bases, and covered with blackish brown indumentum; partial infructescence about 15
cm long, bearing up to 10 unequal secondary partial infructescences. Fruit subglobose,
covered with 14 vertical rows of glossy yellowish scales, 8 mm long and 4 mm broad.
Seed one, globose. Endosperm slightly ruminate.
Distribution. Known from type locality only.
Habitat and ecology. This species is common in Agathis forest, beside streams on the
slopes of G. Malemo, 1000 m alt.
Distribution. This species 1s only known from the type locality.
Uses. Young shoot 1s edible and good.
Vernacular name. Uwi manis (umbut manis).
Notes. This species has been identified by Maturbongs (in scheda) in 2001 as
Daemonorops macroptera, to which it is morphologically similar. However, after
careful examination, D. mogeana differs from D. macroptera by the leaf-sheath
armature which includes very robust spines, subglobose fruit and slightly ruminate
endosperm. In contrast, D. macroptera has gigantic, fragile, easily broken spines, and
ellipsoidal fruit and deeply ruminate endosperm.
Specimen examined: Central Sulawesi: Kab. Poso, District Kulawi, Dusun Moa, Mt.
Malemo, 1000 m alt., 21 October 1977, J.P. Mogea 1356, fruiting specimen (BO, K,
by
5. Daemonorops robusta Warb. ex Becc., Ann. Roy. Bot. Gard. (Calcutta) 12(1): 101
(1911). TYPE: Bojong, Province Minahasa, North Sulawesi, Warburg s.n. (Herb.
Berlin, n.v., probably destroyed, type specimen pictures seen in Beccari’s book kept
at K).
Solitary to clustering rattan, very robust, 5—7 m tall. Leaf with sheath 7 cm in diam.,
without sheath 2—3 cm in diam.; internodes 15—35 cm long. Leaf sheaths pale yellow-
green, covered with oblique, very large black thorns with joined bases up to 5 cm long,
Daemonorops in Sulawesi 27
sheath surface with caducous glaucous black indumentum, leaf sheath mouth armed
as the rest of sheath; knee present conspicuously, armed as the rest of sheath. Leaves
to 4 m long including petiole to 40 cm long, armed adaxially with densely erect, long
black spines to 5 cm long, abaxially armed with erect, oblique spines in groups up to 5
cm long; rachis armed with erect, solitary spines up to 5 mm long: cirrus up to 150 cm
long, armed with regularly arranged groups of grapnel-like spines, blackish at the tip:
leaflets lanceolate, papery, acute, 60 cm long, 2 cm wide, armed with scattered reddish,
short bristles along the main nerve on the lower surface, leaflet margin armed with
short spinules, reddish. Male inflorescence ascending, to 50 cm long, peduncle up to 22
cm long, armed with groups of 2-8 slightly bulbous based spines 2-20 mm long with
pointed tips, more robust adaxially than abaxially: peduncular bract woody, erect to 48
cm long, 4 cm wide, lanceolate at the tip, covered by rusty brown indumentum. Female
inflorescence not known. Infructescence pendulous, 50—100 cm long, peduncle 20—30
cm long, armed with spines forming rings; peduncular bracts leathery, erect, 30 cm
long, 2 cm wide, ellipsoid oblong, covered by rusty indumentum, armed with scattered
needle like, blackish spines; partial inflorescences 5-6, each bearing up to 8 secondary
partial inflorescences; involucre pendulous, fiat, just above the involucrophore, 5 mm
long; involucrophore short, papery, 2 mm long. Fruits spherical, 15 x 10 cm, covered
by 9 vertical rows of yellowish cream encrusted scales. Seed globular, 8 x 8 mm, seed
surface reticulate, endosperm deeply ruminated.
Distribution. North Sulawesi: Bolaang Mongondow, Manado, Laelumbuan; Central
Sulawesi: Toli-toli.
Habitat and ecology. Primary forest hillslopes, deep valleys, alluvial flats, lowland
riverbank forest.
Vernacular names. Lauro manu (toli-toli), pondas valukan, pondas rasisagan, pondas
kuluwi (Manado), rotan susu (Gorontalo language).
Notes. The mature fruit of this species has a sour sarcotesta; the young fruit is green,
maturing red. Apparently related to D macroptera but the fruit is spherical.
Specimens examined: North Sulawesi: Bolaang Mongondow, Tapak Kulintang,
Dumoga Bone National Park, 220 m alt., 8 Mar 1984, J.P. Mogea JPM 5076, young
fruit (BO, K). Bolaang Mongondow, Pindol, Lolak, 50 m alt., 19 Oct 1973, J. Dransfield
& J.P. Mogea JD 3805, mature fruit (BO); JD 3800, male flower (BO, K). Manado,
Pondok Pingsang, Karoewatoe, 50 m alt., 26 Feb 1895, Koorders KDS 184106, fruiting
(BO, L); Laelumbulan by Paku Ure, 700 m alt., 9 Mar 1895, Koorders KDS 183998,
fruiting (BO, L). Heyne 2510, sterile (BO); Gorontalo, near Marisa, Illoheleuma, 8
Jan 1989, Lynn Clayton 3, fruiting (K). Central Sulawesi: Dako, Mountain Lakatan,
Toli-toli, 750 m alt., 25 Feb 1985, Ramlanto & Z. Fanani 530, fruiting (BO). Malili,
Toli-toli, Kawata, 200 m alt., 13 Apr 1933, J. van Jijll de Jong 1, young fruit (BO);
Inland from Batui and Seseba on Batui river, Sinsing, 16 Oct 1989, 70-100 m asl..
Coode 5967, fruiting (K).
28 Gard. Bull. Singapore 63(1 & 2) 2011
6. Daemonorops riedeliana (Mig.) Becc., Rec. Bot. Surv. Ind. 2:226 (1902);
Calamus riedelianus Miq., Verh. Kon. Akad. Wetensch., Afd. Natuurk. 11: 18 (1868);
Palmijuncus riedelianus (Mig.) Kuntze, Revis. Gen. Pl. 2: 733 (1891). TYPE: North
Sulawesi, Manado, Minahasa, 1895, Riedel s.n. (holo L).
Slender, clustering rattan, up to 10 m tall. Stem with sheaths up to 20 mm in diam.,
without sheaths to 6 mm in diam., internodes 8-10 cm long; leaf sheaths covered
with needle-like, almost uniformly upward-pointing spines that are solitary or in
groups, the spines enormous, up to 6 cm long; sheath surface smooth with corky
creamy indumentum; leaf sheath mouth armed as the rest of the sheath; knee present
conspicuously, armed as the rest of the sheath. Leaves to 2 m long including petiole to
15 cm, armed very densely with 5—10 mm long spines all around, lower side of rachis
armed with ternate claws, upper side slightly prickly; leaflets numerous, arranged
rather distantly, 38—S7 pairs on each side of the rachis, arcuate, somewhat spidery;
leaflets lanceolate, papery, acuminate, 25—30 cm long, | cm wide, armed with scattered,
reddish, bristly spinules along the mid-nerve of the upper surface; young leaf covered
in caducous white indumentum; transverse veinlets slender, visible on both surfaces;
margins armed with rather close ciliate spines,. Male and female inflorescences not
known. Fruits spherical, 13 x 13 mm, covered by 8 vertical rows of yellowish brown,
duil encrusted scales. Seed irregularly globular, 10 x 8 mm, with reticulate surface.
Distribution. North and South Sulawesi.
Habitat and ecology. Disturbed primary forest, and steep hillslopes and ridge top
lowland forest, on volcanic soil, 500 m asl.
Vernacular name. Angah (this name is also applied to D. macroptera).
Notes. It is noted by Dransfield that the seed is both astringent and sweet.
Specimens examined: North Sulawesi: Minahasa, Bitung, Batu Angus Nature Reserve,
500 m alt., 7 Oct 1973, J. Dransfield & J.P. Mogea JD 3737, fruiting (BO); Bolaang
Mongondow, Pindool Lolak, 150 m alt., 18 Oct 1973, JD 3787, male flower (BO,
K); Manado, 2 Mar 1895, Koorders 18389, fruiting (BO, L); A.G. Waturandang 51,
sterile (BO). South Sulawesi: Kabupaten Mamuju, Kec. Kaluku, Dusun Roa, Desa
Dutas Kaluak, Bukit Banga, 300 m alt., 08 Feb 1993, Padmi Kramadibrata 30, sterile
(BO).
7. Daemonorops sarasinorum Watb. ex Becc., Ann. Roy. Bot. Gard. (Calcutta) 12 (1):
100 (1911). TYPE: North Sulawesi, Tomohon, in the province of Minahasa, Sarasin
1082 (Herb. Berol., n.v., probably destroyed, type specimen pictures seen in Beccari’s
book kept in K).
Daemonorops in Sulawesi 29
Robust, clustering rattan, climbing 15—30 m long. Stem with sheath 6 cm in diam.,
without sheath to 4 cm in diam.; internodes generally rather short, 20 cm long. Leaf
sheath very densely covered in reflexed black spines, to 4 cm long, pinkish when
young; sheath surface with buff scurfy sometimes variously grey indumentum, leaf
sheath mouth armed as the rest of sheath; knee present, very conspicuous, armed as
the rest of sheath. Leaves 2.54 m long including petiole 40 cm or more, the petiole
somewhat reddish, armed adaxially with short, erect, scattered spines to 4 mm long,
abaxially armed with erect, solitary spines rarely up to 15 mm long; rachis unarmed or
proximally only slightly armed; cirrus up to 2 m long, armed with regularly arranged
groups of grapnel-like spines, blackish at the tip; leaflets mostly regularly arranged,
densely crowded, 50—75 on each side of the rachis, stiff; leaflets lanceolate, papery,
acuminate, 60-80 cm long, 2 cm wide, armed with scattered, reddish short bristles
along the main nerve on both surfaces. Male inflorescence pendulous, up to 120 cm
long, consisting of 15 rachillae, each rachilla consisting of 8 partial inflorescences;
peduncular bract narrow and elongate, cupressiform with several erect, compact
or appressed flowers; partial inflorescence covered with abundant furfuraceous
indumentum. Male flowers short, 4 mm long, with anthers quite exserted from corolla
and enclosed within spathes; calyx is very small, almost flat, corolla is longer than
calyx. Female inflorescences pendulous, 60 cm long, peduncle 20 cm long, densely
armed with groups of spines; peduncular bract leathery, elongate, 17 cm long, 3 cm
wide, cupressiform, covered with rusty indumentum and innumerable long, narrow,
scattered black spines, tubular before flowering: partial inflorescences 6, each bearing
up to 11 secondary partial inflorescences: involucre pendulous, flat just above the
involucrophore, 5 mm long; involucrophore short, papery, 2 mm long. Female flowers
at the time of anthesis 4 mm long, exclusive of the stigmas which are exserted from the
corolla and are about as long as the whole length of the female flowers (c. 8 mm); calyx
very short, copular, polished (not striate), superficially 3-toothed; corolla ventricose,
urceolate, strongly seriately veined, coriaceous, having 3 broadly triangular, acute
teeth.
Distribution. North Sulawesi: Manado, Bitung, Kotamobagu; Central Sulawesi: Palu,
Mt. Rorokatimbu.
Habitat and ecology. Montane forest, somewhat disturbed lower montane forest on
steep terrain, shallow clayey soil.
Vernacular name. Pondan katunun (Manado).
Notes. Male flowers are unpleasantly ester-scented (from JD 3862).
Specimens examined: North Sulawesi: Manado, Minahasa 2 Mar 1895, Koorders KDS
18391, sterile (BO); 30 Apr 1895, Koorders KDS 18407, sterile (BO). Minahasa,
Bitung, Batu Angus Nature Reserve, 800 m alt., 8 Oct 1973, J. Dransfield & J.P.
Mogea JD 3744, fruiting (BO). Bolaang Mongondow, G. Ambang, Kotamobagu, 1000
30 Gard. Bull. Singapore 63(1 & 2) 2011
malt., 26 Oct 1973, JD 3862, male flower (BO, K); JD 3861, fruiting (BO, K). Manado,
Pondok Simpang, 50 m alt., 2 Mar 1895, Koorders KDS 15388f, sterile (BO). Central
Sulawesi: Mt Rorokatimbu, west slope c. 80 km SSE of Palu, 1700 m alt., E.F: de Vogel
5484, female flower (BO, K); 13 May 1979, c. 1°16’S 120°18’E, 1300 m alt., E.F de
Vogel 5335, sterile (BO, K); Mount Sadaunta, May 1976, G.G. Musser s.n., (K).
ACKNOWLEDGEMENTS. The first author (HR) is very much indebted to LIPI for sponsoring
her study at the Bogor Agricultural University. STORMA, through a BMZ grant, sponsored
HR’s fieldwork at the Lore Lindu National Park, Central Sulawesi. The Royal Botanic Gardens,
Kew through its Bentham-Moxon Trust, sponsored HR’s visit to the Kew and Leiden herbaria.
Dr. John Dransfield and Prof. Mien A. Rifai were very supportive and gave tremendous help
during the preparation of the revision. Special thanks go to Dr. William J. Baker for his kindness
while HR worked in Kew, and to Dr. Lauren Gardiner and Gerard Thijsse for helping HR in
various ways while working at Kew and Leiden.
References
Beccari, O. (1911) Asiatic Palms —Lepidocaryeae Part II. The Species of Daemonorops.
Ann. Roy. Bot. Gard. (Calcutta). 12: 1-241.
Dransfield, J. (1986) A guide to collecting palms. Ann. Missouri Bot. Gard. 73: 166—
176.
Dransfield, J. (1999) Species and species concepts in old world palms. Mem. New York
Bot. Gard. 83: 5—20.
Dransfield, J. & Manokaran, N. (1994) Plant Resources of South East Asia No. 6.
Rattans. Bogor: Prosea Foundation.
Dransfield, J., Uhl, N.W., Asmussen, C.B., Baker, W.J., Harley, M.M. & Lewis, C.E.
(2008) Genera Palmarum: The Evolution and Classification of Palms. Kew:
Royal Botanic Gardens, Kew.
Furtado, C.X. (1953) The species of Daemonorops in Malaya. Gard. Bull. Singapore
14: 49-147.
Mogea, J.P. (1991) Utilization and conservation of Indonesian palms. In: Johnson,
D.V.J. (ed) Palms for Human Needs in Asia, pp 37—73. Rotterdam: A.A. Balkema.
Rifai, M.A. (1976) Sendi-sendi botani sistematika. Bogor: Herbarium Bogoriense.
Roemer, J.J. & Schultes, J.A. (1830) Systema Vegetabilium. 7(2): 1333. Stuttgart.
Rohlf, F.J. (1997) NTSYS-pe Numerical Taxonomy and Multivariate Analysis System
Version 2.01. Setauket, New York: Exeter Software.
Rustiami, H. (2009) Two new species of Daemonorops from Sulawesi. Reinwardtia
13(1): 25-30.
Vogel, E.F. de (1987) Guidelines for the preparation of revisions. In: Vogel, E.F. de (ed)
Manual of Herbarium Taxonomy, Theory and Practice. Jakarta. MAB, UNESCO.
Gardens’ Bulletin Singapore 63(1 & 2): 31-62. 2011 31
The Pandanaceae of the Bukit Baka Bukit Raya National
Park and adjacent areas,
West and Central Kalimantan, Indonesia,
with notes on their nomenclature
and the rediscovery of Pandanus aristatus
and several new records
Ary Prihardhyanto Keim', Rugayah and Himmah Rustiami
Herbarium Bogoriense, Botany Division, Research Center for Biology,
Indonesian Institute of Sciences, Cibinong Science Center (CSC),
Jl. Raya Jakarta-Bogor Km 46, Cibinong, Bogor 16911, Indonesia
‘herbogor@indo.net.id (corresponding author)
ABSTRACT. Eight species of Pandanaceae (3 Freycinetia spp., 5 Pandanus spp.) were recorded
from the Bukit Baka Bukit Raya National Park and adjacent areas in the West and Central
Kalimantan Provinces, Indonesia. Pandanus aristatus was recollected and the description
improved. Pandanus motleyanus has been assigned to synonymy under P. yvanii. Pandanus
yvanii and P. helicopus were found to occupy different niches in the peat swamps. Pandanus
epiphyticus Martelli and P. pachyphyllus Merrill were recorded for the first time in Kalimantan.
The doubtful presence of F’ sumatrana in Java is resolved. Two Eastern Malesian species,
F, amboinensis and F; ceramensis are synonyms of F. sumatrana, thus the species is now an
exceptionally widespread species in both western and eastern Malesia. Full descriptions of
species are provided.
Keywords. Borneo, Bukit Baka Raya, Freycinetia, Kalimantan, Pandanaceae, Pandanus
Introduction
The pandan flora of Borneo has been studied by Stone (1970a, 1993). Most of
the species enumerated were from Sarawak and Sabah and little was known from
Kalimantan, the Indonesian part of Borneo, particularly from the southwestern part,
where the Bukit Baka Bukit Raya National Park 1s located.
Bukit Baka Bukit Raya National Park is a protected area within the Schwaner
Range on the border between the Indonesian Provinces of West and Central Kalimantan,
with an area of approximately 181,090 hectares. Prior to this study, the latest pandan
collecting activity in the vicinity was by Nieuwenhuis in the western and central parts
of Kalimantan (Nieuwenhuis 1898, Steenis 1950). The pandan collections collected
during the Nieuwenhuis expedition were identified by Martelli and later mentioned in
his two enumerations (Martelli 1910, 1913). Merrill (1922) and Martelli (1929) also
cited these expedition specimens in their accounts on the pandan of Borneo, in which
32 Gard. Bull. Singapore 63(1 & 2) 2011
many of the collections had been appointed as types.
Although numerous publications have been produced since (see Kanehira 1938;
St. John 1961, 1965; Stone 1967a, 1967b, 1970a, 1978, 1980, 1982, 1983b, 1993), the
pandan flora of the Bukit Baka Bukit Raya National Park and surrounding areas was
hardly mentioned and still made reference to Martelli (1929). Only Pandanus aristatus
was mentioned for the National Park (Stone 1993). The most recent expedition carried
out within the National Park was the 1982—1983 expedition in Bukit Raya conducted
by Reuler (1987), when no pandan was collected.
This paper describes several species of the pandan flora in the Bukit Baka
Bukit Raya National Park and adjacent areas, including the Gunung Kelam Ecopark
and Baning Protected Forest, based on the most recent expedition in 2006, as well as
observations made with herbarium specimens kept in the BO Herbarium.
Enumeration of species
Freycinetia
1. Freycinetia kartawinatae A.P. Keim, Reinwardtia 13(1): 15 (2009). TYPE: A.P.
Keim 770 (holo BO!), Indonesia, West Kalimantan, Katingan Hulu, Waringin Timur,
on logging road, 2 May 2006. (Fig. 1)
Robust climbing pandan, to 50 m high. Stem c. 9.5 cm girth (c. 3 cm diameter).
Leaves spirally arranged in 3 ranks (tristichous); blade lanceolate-elongate, c. 43.5—44
cm long, c. 3 cm wide, apex acuminate, margin with spines up to distal 1/3 of the
length; adaxial surface green, glabrous; abaxial surface pale green, glabrous; auricle
tapered, margin entire, brownish yellow to creamy brown; leaf sheath yellowish green.
Infructescence terminal, c. 14 cm long, consisting of 3 to 4 spirally arranged cephalia;
pedicel yellow to pale yellowish orange, glabrous; bracts distinctively bright orange,
thick, hard, fleshy, glabrous, apex acuminate, margin with spines, each boat shaped, c.
22.5—23 cm long, caducous. Cephalium cylindric-elongate, c. 9 cm long, c. 3 cm wide,
pale green to dull greyish green; stigma 4—5, mostly 4.
Distribution. Endemic to Kalimantan.
Habitat. Lowland tropical rain forest at about 300 m altitude.
Vernacular names. Kajak rajak, mérajak (Dayak, Belaban dialect).
Uses. Although not used by local people, orang utans and gibbons are said to consume
the bracts.
Notes. In general appearance, F. kartawinatae is very similar to F: insignis, but differs
in its orange infructescence bracts compared to the purplish white bracts in F) insignis.
Pandanaceae in Bukit Baka Bukit Raya National Park 33
‘ > ~ > ~
—
Fig. 1. Freycinetia kartawinatae A.P. Keim. A. A terminal infructescence with four spirally
arranged cephalia and very distinctive bright orange bracts (above). B. Terminal infructescence
with three spirally arranged cylindric-elongate cephalia with yellow glabrous pedicels, the
bracts already fallen away (below). Photos: A.P. Keim, Rugayah & H. Rustiami.
34 Gard. Bull. Singapore 63(1 & 2) 2011
Ap
Among members of the section Polystachya in Borneo that have infructescences
with 3 cephalia or more, FE. kartawinatae differs from F) kinabaluana in its terminal
infructescences with 3-4 cephalia, compared to the lateral infructescences of F
kinabaluana that have 5—6 cephalia per infructescence. Freycinetia kuchinensis also
has reddish orange bracts, but differs in having 2—3 cephalia per infructescence (fide
Martelli 1910) that are globose, compared to the 3-4 cylindric-elongate cephalia per
infructescence in F. kartawinatae. Although it has similarly long (40-60 cm) leaves
and also 3-4 stigmas, F) sarawakensis differs from F. kartawinatae in its lateral (not
terminal) infructescences and scabrid (not glabrous) pedicels.
A taxon from Sarawak would appear to have many similarities with F
kartawinatae, especially in having orange bracts, terminal infructescences, and 3-4
cephalia per infructescence. This taxon was named F. andersoniana by Stone (1967c),
but invalidly published.
Specimens examined: Only known from the type.
2. Freycinetia sumatrana Hemsley, J. Linn. Soc. Bot. 30: 167 (1894). TYPE: Beccari
211 (holo K; iso FI), Indonesia, Sumatra, Mount Singalan (presumably Mount
Singgalang in West Sumatra), June 1878. (Fig. 2—5)
Freycinetia valida Ridley, Mat. Fl. Mal. Penins. 2: 234 (1907). LECTOTYPE: Ridley
3937 (SING!), Malaysia, Malay Peninsula.
Freycinetia auriculata Merrill, Philipp. J. Sci., C. Bot. 3: 312 (1908). TYPE: Bur. Sci.
Foxworthy 876 (holo PNH7; iso A), Philippines, Palawan, Puerto Princesa, May 1906.
Freycinetia loheri Martelli, Webbia 3: 15 (1910). SYNTYPES: Loher 1577 (K; iso
PNH?), Philippines, Luzon, Benguet, 1908-1915; Loher 1578; Loher 5469, Luzon,
Montalban, June 1908-1915.
Freycinetia lucida Martelli, Webbia 3: 168 (1910). TYPE: H. Hallier 3188 (holo BO!;
iso L), Indonesia, Kalimantan, Amai-Ambit, 1893-1894.
Freycinetia ceramensis Martelli, Webbia 3: 169 (1910). TYPE: G.H. de Vriese s.n.
(holo L), Indonesia, Moluccas, Seram, 1857—1861.
Freycinetia amboinensis Martelli, Webbia 3: 170 (1910). TYPE: Zeysman s.n. (holo
BO! iso L), Indonesia, Moluccas, Ambon.
Freycinetia sumatrana Hemsley var. penangiana B.C.Stone, Gard. Bull. Singapore
25(2): 202 (1970). TY PE: Stone5890 (holo KLU), Malay Peninsula, Selangor, Templer
Park, Kanching, June 1965.
Pandanaceae in Bukit Baka Bukit Raya National Park 35
Robust climbing pandan, to 20 m high. Stem 7-8 cm girth. Leaves spirally arranged in
3 ranks (tristichous); blade lanceolate-elongate, apex acuminate, margin with spines
throughout, 60-135 cm long, 2.5—3.5 cm wide; adaxial surface green, glabrous;
abaxial surface pale green, glabrous; auricle lobed, margin entire, pale purplish red
to pale brownish yellow; leaf sheath white, purplish red in young leaf. /nfructescence
terminal, consisting of 3-4 spirally arranged cephalia. Cephalium elongate, 16-19 cm
long, green to yellowish green; stigmas 2-4, mostly 2; pedicel 9-10 cm long, glabrous,
pale yellowish green.
Distribution. Andaman Islands, Malay Peninsula, Sumatra, Java, Borneo, the
Philippines, and Moluccas.
Habitat. Lowland tropical rain forest. In Bukit Baka Bukit Raya National Park
commonly found at elevations lower than 500 m (Fig. 2). One specimen (F’H. Endert
3875, BO!) was collected from Mount Kemoel in East Kalimantan at 1200 m.
Vernacular names. Rajak (Dayak, Balaban dialect).
Uses. Leaves are used for baskets, mats and handicrafts.
Notes. The differences between F) swmatrana and the taxa listed above in the
synonymy are slight and merely in leaf dimensions, cephalia size, colour of bracts,
appearance of auricles, pedicel surface, and stigma number.
The possession of lobed auricles is a more important feature that distinguishes
this species, as acknowledged previously by Stone (1968), who newly proposed the
section, Auriculifoliae based on this character. A year later, Stone (1969) excluded F:
loheri and F; vidalii from the section based on the fact that neither of these two species
actually possessed lobed auricles. In the same publication Ff! auriculata was placed
as a synonym of F) sumatrana. Finally, F) valida was included into synonymy (Stone
1970a).
The present study supports Stone in placing F) auwriculata and F. valida as
synonyms of F. sumatrana. Despite no information concerning bract colour and the
relatively smaller size of the cephalium, there is no significant difference between F:
auriculata and F) sumatrana.
For F- valida, Ridley (1907; 1925) did not mention anything about the bract
colour. Only a brief note on the female bract was given, which was described as leaf-
like and not coloured. We interpret “not coloured” as white, so that is not different
from F. sumatrana. Actually, Ridley (1925) himself noted that F) valida is very near
to F) sumatrana.
Regardless of the immature state of the cephalia, the observations made on the
types indicate that there is no single decisive morphological character that can be used
to distinguish F) amboinensis or F: lucida from F. sumatrana.
The placement of F) amboinensis as synonym has a further consequence that
F. sumatrana 1s now found in the Moluccas, particularly in Ambon and Halmahera
36 Gard. Bull. Singapore 63(1 & 2) 2011
Fig. 2. Freycinetia sumatrana Hemsley from Bukit Baka Bukit Raya National Park. A. Robust
habit with terminal infructescence. B. Infructescence of four elongate-ellipsoidal cephalia.
Photos: A.P. Keim, Rugayah & H. Rustiami.
Pandanaceae in Bukit Baka Bukit Raya National Park 37
Islands; thus new records of FE. swmatrana for the two islands. The presence of F:
sumatrana in the Moluccas 1s also supported by the placement of F) ceramensis into its
synonymy. Although F’ ceramensis possesses auricles with more pronounced spines
(Fig. 3), this is insufficient for distinction from F. swmatrana. Observations made on
specimens recently collected from Seram, an island within the Moluccas Archipelago
just north of Ambon (Keim et al. 2008) indicate that the spines vary from minute to
apparent even in the same branch. Auricles with minute spines are usually observed
on younger leaves (1.e., the terminal part of the branch), whereas auricles with larger
spines are found on older leaves. Prior to this placement F) ceramensis was known
as an endemic of Seram Island; so now F. sumatrana is also newly recorded for the
island.
The placement of two Eastern Malesian species (F- amboinensis and F.
ceramensis) into the synonymy of F) swmatrana has the consequence that the species
is now recognised as one of the most widespread species of the genus in Malesia,
occupying both western and eastern parts. The other such species is F. scandens.
Stone (1970b) described the smaller overall dimensions and “entire auricle” as
two important characters that distinguish F) /ucida from F- sumatrana. That description
of the auricle does not agree with the protologue (Martelli 1910) that clearly described
the auricle as with minute spines (original text: “ad margines crebre et minute fimbriato-
denticulatis”). The present study is in favour of Martelli; thus, the possession of spiny
auricles helps place F. /ucida in the synonymy of F) sumatrana.
Stone (1970a) identified a specimen collected from the top of Mount Beratus
(altitude 1200 m) in South Kalimantan, W. Meijer 904 (BO! with duplicate at L) as F.
lucida. This specimen has now been determined as F- sumatrana, and marks the first
record of F- sumatrana in the Indonesian part of Borneo (Kalimantan). The collections
made in this present study support the identification of the species in Kalimantan.
Previously F. sumatrana was only known from Sarawak and Sabah (Stone 1970a). As
a consequence, Ff. sumatrana is now known to be more widely distributed throughout
Borneo. A specimen (F’H. Endert 3875) was also collected from 1200 m, and in Sumatra
and the Malay Peninsula, the species can still be found at an altitude of approximately
1828 m (Stone 1970b).
The varietal status of F) swmatrana var. penangiana is not accepted as the
slight differences in the shape and colour of the auricles mentioned by Stone (1970b)
are regarded as insufficient distinction, based on observations from the Bukit Baka
Bukit Raya National Park. The specimens clearly show that auricles can vary from
tapered to lobed even in the same shoot (Fig. 4). There is also variation in auricle
colour from bright purplish pink to light brownish yellow. Actually the variation in the
type of auricle in F. swmatrana is not unusual.
An infructescence with 4 cephalia observed in one of the collections made for
this study (4.P. Keim 764, BO!; Fig. 2) raises the possibility that this could be a member
of the section Polystachya, which in Borneo is represented by a single known species,
F. kinabaluana (Stone 1970a). However, the specimen clearly possesses a lobed
auricle, a feature that is completely absent in F: kinabaluana. Freycinetia kinabaluana,
on the other hand, has a lateral infructescence consisting of S—6 cephalia. Frevcinetia
ard. bull. Singapore 63( 2) 20
38 Gard. Bull. Singapore 63(1 & 2) 2011
cu
Fig. 3. Freycinetia ceramensis Martelli, just recently rediscovered and collected from Ceram
Island in the Moluccas (Keim et al. 2008). A. Robust habit and climbing to more than 20 m
high. B. Ternate staminate inflorescences (the bracts are pale yellow and basally tinged reddish
purple) that are obviously simlar to those of F) sumatrana. C. Infructescences, here binate but
mostly ternate. D. The distinctive lobed auricles with spines on the margins. The characters
shown in these pictures are regarded as clearly supporting the placement of F’ ceramensis into
the synonymy of F. sumatrana. Photos: A.P. Keim & S. Susiarti.
Pandanaceae in Bukit Baka Bukit Raya National Park 39
sumatrana always has terminal infructescences and has never been reported having
more than 4 cephalia in one infructescence. Stone (1970c) notes that F. swmatrana in
the Malay Peninsula has 3-4 cephalia per infructecence, and we know of at least one
collection with 4 cephalia, Endert 3875 (BO! duplicate at L), which was collected
from Sumatra.
Fig. 4. Auricle variation in FE’ sumatrana from Bukit Baka Bukit Raya National Park. Tapered
(left arrow) and lobed (right arrow) auricles can be found in the same young shoot (auricles are
purplish pink in fresh specimens, but on mature branches or stems, brownish yellow ones are
more common). Photo: A.P. Keim, Rugayah & H. Rustiami.
Numerous specimens from Java kept in BO that were previously labelled F:
valida had been identified by Stone as F. sumatrana. Based on this, Stone (1972)
suggested the presence of the species in Java. The same specimens have been re-
identified in the present study as belonging to F. insignis based on the fact that none
of those specimens actually has the distinctive lobed auricle and information from
the fieldnotes clearly mentioned deep red or pale purplish red bracts, a feature that is
characteris of F) insignis.
The presence of the true auriculate F) swmatrana in Java was finally proven
based on collections recently made in the Gunung Tukung Gede Nature Reserve in
Banten, western Java (7. Djarwaningsih 1499; Djarwaningsih pers. comm., 2010; Fig.
5). This is expected, as the islands of Java and Sumatra are only separated by the
40 Gard. Bull. Singapore 63(1 & 2) 2011
NY odie
ie ‘A
Fig. 5. Freycinetia sumatrana Hemsley from the Gunung Tukung Gede Nature Reserve in
Banten, western Java. A. Habit. B. Young terminal ternate infructescence with the characteristic
creamy white-pale yellow bracts with reddish purple-tinged basal part. C. Mature terminal
ternate infructescence showing the distinctive lobed auricle. D. Young terminal quaternate
infructescence, not uncommon in F. sumatrana. Photos: T. Djarwaningsih & A. Supriatna.
Pandanaceae in Bukit Baka Bukit Raya National Park 4]
relatively narrow Sunda Strait.
Freycinetia walkeri shares many morphological similarities with F- swmatrana,
particularly the lobed auricle and the number of stigmas. Indeed, other than the colour
of the bracts, there is no substantial difference between this species and F. sumatrana.
Stone (1975) mentioned that F’ walkeri differs from F) suwmatrana only in the smaller
size of the cephalium (Stone wrote “fruits”) and red floral bracts. Solms (1878) did
not mention bract colour and it was Stone (1969: 1975) who mentioned the red bract
colour based on observations he made on several non-type specimens, particularly N.
Wirawan 818.
The BO Herbarium has two specimens identified by Stone (1975) as F. walkeri,
N. Wirawan 818 (BO!) and Kostermans 24071 (BO!). Wirawan 818 is a staminate
collection from the vicinity of Ratnapura, Ceylon in 1969, in which the bract colour
is noted as bright red, while Kostermans 24071 (collected also in Ratnapura vicinity,
Ceylon in 1973) is a pistillate collection with a fieldnote mentioning the fruit as dark
red. Stone did visit Ceylon, but mostly worked in the Peradeniya Botanic Garden,
thus he had never seen or collected F’ walkeri in the field and was apparently using
only specimens and citing from fieldnotes. The result of this current study is not in
accordance with Stone in recognising the two specimens as F. wa/keri. Neither has the
distinctive lobed auricle, and the specimens are unlikely to be F) walkeri. Therefore,
Stone’s description of the bract colour in F’ walkeri as red is considered doubtful.
Until better specimens become available F- walkeri is regarded as a distinct species but
closely allied to F sumatrana. Further study is still required. When the matter of bract
colour is resolved, F- swmatrana is likely to become a synonym of F. walkeri based on
the law of priority.
Specimens examined: INDONESIA. West Kalimantan, Amai-Ambit, 1893-1894, H.
Hallier 3188 (holo BO! iso L); Sungai Betas Dalam, 29 Apr 2006, A.P. Keim 757 (BO!);
Bukit Siman, Camp 35, 30 Apr 2006, A.P Keim 764 (BO!); Central Kalimantan,
Area Bukit Raya, Km 44, Sungai Wah, Kasongan, Sinamang Mentikay, 1 May 2006,
A.P. Keim 766 (BO!): East Kalimantan, Mount Kemoel, Km 47, 10 Oct 1925, F-H.
Endert 3875 (BO!): Nunukan Island, N of Tarakan, Nov 1953, W. Meijer 2163 (BO!);
South Kalimantan, Bandjarmasin-Martapoera, Km 14, 12 Oct 1939, B. Polak 494
(BO!); Moluccas, Ambon, 7eysman s.n. (holo BO! iso L); West Java, Banten, Gunung
Tukung Gede Nature Reserve, Cikolelet, 02 Oct 2009, T. Djarwaningsih 1499 (BO!).
3. Freycinetia cf. tenuis Solms (Fig. 6)
Slender climbing pandan, up to 40 m. Leaves small, short, lanceolate-elongate, c. 6
cm long, 0.5 cm wide, apex acuminate and with minute spines, spines only on terminal
and basal parts of leaf; adaxial surface green, glabrous; abaxial surface pale green,
glabrous; leaf sheath deep purplish red, apical part with minute spines; auricle tapered,
small, deep purplish green to red.
42 Gard. Bull. Singapore 63(1 & 2) 2011
Distribution. Sumatra and Borneo. In the Bukit Baka Bukit Raya National Park the
species was only found at one site and appears to be rare.
Habitat. Lowland tropical rain forest. In Bukit Baka Bukit Raya, it is found in a rather
open area close to a creek.
Vernacular name and uses. Not recorded.
Notes. Unfortunately the only specimen collected in this current study (AK 769) is a
sterile one. Identification of this material as possibly belonging to F- tenuis is based on
the habit, leaf dimensions, and colouration of auricles, which are in accordance with
the protologue (Solms 1879; see also Warburg 1900), particularly the distinctive deep
purplish red leaf sheath and tapered auricles. Freycinetia tenuis is the smallest member
of the genus in Borneo. The species was recorded for Borneo by Stone (1970a) in
Sabah, North Borneo; thus, as in the case of F’ sumatrana, the presence of F. tenuis
in Borneo strengthens the biogeographical link between Sumatra and Borneo (1.e.,
Sahulland).
Specimens examined: INDONESIA. Central Kalimantan, Kantor PT Sari Bumi
Kusuma Timber, Katingan Hulu, Kota Waringin Timur, | May 2006, 4.P. Keim 769
(BO!); South Kalimantan, Bandjermasin to Martapoera, Km 14, 12 Oct 1939, B.
Polak 513 (A, BO! L).
Fig. 6. Freycinetia cf. tenuis Solms (A.P. Keim 769) showing the very slender habit and small
lanceolate-elongate leaves; the overlapping leaf sheaths along the stem are a characteristic
deep purplish red in live specimens, with tapered auricles. Photo: A.P. Keim, Rugayah & H.
Rustiami.
Pandanaceae in Bukit Baka Bukit Raya National Park 43
Pandanus
1. Pandanus aristatus Martelli, Webbia 4 (1): 6 (1913). TYPE: H. Hallier 2250 (holo
BO! iso FI), Indonesia, Kalimantan (then Dutch Borneo), Sungai Djemala, Gunung
Kelam, 1893-1894. (Fig. 7)
Pandanus andersonii H.St.John, Pacific Sci. 15(4): 576 (1961). TYPE: J/.A.R.
Anderson s.n. (holo SAR), Malaysia, Sarawak, Lawas, Kayangeran Forest Reserve,
Nov 1960. PARATYPE: J/.A.R. Anderson $2815 (SAR), Malaysia, Sarawak, Loba
Kabang Protected Forest, 16 May 1954.
Robust shrubby pandan, 2-3 m high. Prop roots absent. Stem very short (1.5—2 cm) or
stemless. Leaves in a rosette of 20 or more, spirally arranged in 3 ranks (tristichous);
each leaf lanceolate-elongate, 250-350 cm long, 4-10 cm wide, stiff, apex acute to
acuminate, margin with obvious spines throughout the length; adaxial surface green,
glabrous, main vein apparent, adaxial ventral pleats not observed or present; abaxial
surface pale green, glabrous, main vein obvious, recurved spines obvious, very sharp;
leaf sheath white to yellowish white. Staminate inflorescence terminal, creamy white,
107-115 cm long; rachis 44-66 cm long, with 8—10 rachillae; rachilla elongate, 28—
30 cm long; peduncle 68-70 cm long, glabrous; bracts 13, each 10-140 cm long,
persistent, boat shaped, pale brown. Staminate flower pale creamy yellow, odorous;
stamens numerous. Pistillate inflorescence terminal, ascending; peduncle 10 cm long,
diameter 2.5 cm, obtusely trigonous. /nfructescence terminal, pendulous, 15—20 cm
long, with a solitary cephalium or a spike consisting of 2 to 3 unequally sized cephalia;
peduncle c. 8.5 cm. Cephalium elongate-ellipsoid or broadly ellipsoid to subglobose,
obtusely trigonous, light to dark brown, 8.5—-14 cm, c. 3 cm diameter, consists of
numerous drupes. Drupe oblong-ellipsoid, 35-80 mm long, 6-7 mm wide, 5—6 mm
thick, fusiform, 5S—6 angled; pileus conical, 13—17 mm long, gradually narrowed into
the style; style 5-7 mm long, subulate, arcuate, beaked, glabrous; stigma 4-5 mm
long, appointed, needle-like and sharp.
Distribution. Borneo.
Habitat. Peat swamp forest, riversides and riverine forest. In villages close to the Bukit
Baka Bukit Raya National Park, this species is cultivated. Pandanus aristatus is the
most abundant species in the National Park and surrounding areas, particularly in the
peat swamp forests.
Vernacular name. Kajak (Dayak, Belaban dialect).
Uses. Leaves used for baskets, mats, bags, hats, and handicrafts.
Notes. Previously P. aristatus was known only from the type (H. Hallier 2250), which
is a staminate collection. As the classification of the Pandanaceae is basically based
44 Gard. Bull. Singapore 63(1 & 2) 2011
Fig. 7. Pandanus aristatus Martelli. A. Robust-shrubby and stemless habit with large male
inflorescence, beside the Nenga Pinoh-Sintang Highway, West Kalimantan. B. Staminate
inflorescence at anthesis showing numerous minute, crowded, pale creamy yellow male
flowers and layers of pale brown bracts; same venue as A. C. Within the Bukit Baka Bukit
Raya National Park, P. aristatus is common along riversides. Photos: A.P. Keim & H. Rustiami.
Pandanaceae in Bukit Baka Bukit Raya National Park 45
on pistillate characters, the identity of P aristatus was thus unresolved. Pistillate
material is apparently rare, as even the present study has only succeeded in obtaining
five staminate collections from various area and habitats within the National Park to
the Nenga Pinoh-Sintang highway (Fig. 7), including the Gunung Kelam Ecopark (the
type locality); these were from riversides to peat swamp forests or open areas. These
findings not only increase the known distribution of the species, but also add new
information on preferred habitats. A visit to the type location was also without success
as individuals observed were not in flower or fruit. Nonetheless, the comparison
between these five specimens and the holotype available at BO indicate that they
undoubtedly belong to P. aristatus.
Apart from the lack of pistillate material in P. aristatus and staminate material
in P. andersonii, no other significant morphological character could be used to
differentiate P. andersonii from P. aristatus. Also, besides sharing the same habit as
robust shrubby pandan with extremely short stems (less than 50 cm high); the two
taxa also inhabit the same habitat, peat swamp forests. St. John (1961) described P.
andersonii as the dominant species in peat swamp areas. Stone (1993) even stated that
P. andersonii was a distinctive species of freshwater swamp forest, as is the case with
P. aristatus. Now that the pistillate data for P. aristatus has become available; it is clear
that P. aristatus belongs to subgenus Acrostigma and section Acrostigma. A specimen
from South Kalimantan (J. Dransfield & D. Saerudin 2102) has been identified by the
senior author as belonging to P. aristatus. Pandanus aristatus is regarded as one of the
three main species in the peat swamps of Borneo, the other two being P. atrocarpus
and P. yvanii.
Specimens examined: INDONESIA. West Kalimantan, Kampung Belaban, | May
2006, A.P. Keim 768 (BO!); Km 35, Kampung Belaban, 3 May 2006, 4.P. Keim 776
(BO!); Nenga Pinoh-Sintang, Kampung Pandan, 5 May 2006, 4.P. Keim 778 (BO!);
Central Kalimantan, Katingan Hulu, Sungai Sahaur, Km 54, 2 May 2006, 4.P. Keim
771 (BO!); A.P. Keim 772 (BO!); South Kalimantan, Djaro Dam, Muara Uja, 11 Nov
1971, J. Dransfield & D. Saerudin 2102 (BO! KLU, L).
2. Pandanus discostigma Martelli, Webbia 4 (1): 12 (1913). TYPE: Jaheri 662 (holo
BO!), Indonesia, Central-West Kalimantan, Maguc River, Nieuwenhuis Expeditie,
1896-1897. (Fig. 8 & 9)
Pandanus matthewsii Merrill, J. Str. Br. Roy. Asiat. Soc. 85: 153 (1922). TYPE: Ramos
1321 (holo PNH7# iso BO!), Malaysia, Sabah, Sandakan.
Slender clustered shrubby pandan, 0.5—1 m high. Prop roots very short, not obvious.
Stem very short, not obvious, brown to reddish brown. Leaves in a rosette, spirally
arranged in 3 ranks (tristichous); each lanceolate-elongate, c. 56 cm long, c. 1.5 cm
wide, apex acuminate, margin with minute spines throughout the length; adaxial
surface green, glabrous, adaxial ventral pleats not obvious; abaxial surface pale green,
Gard. Bull. Singapore 63(1 & 2) 2011
Fig. 8. Pandanus discostigma Martelli. A. Slender clustered habit of plants beside the Betas
Dalam River in the Bukit Baka Bukit Raya National Park, with other populations found
submerged. B. Young solitary and terminal pale yellow cephalium. Photos: A.P. Keim, Rugayah
& R. Asmarayani.
Pandanaceae in Bukit Baka Bukit Rava National Park AT
glabrous, recurved spines present: leaf sheath purplish red. Infructescence terminal,
solitary. Cephalium globose, c. 9 cm long, 11.5—16.5 cm in circumference, pale creamy
yellow when young, turning brownish yellow when mature, cephalium consists of
numerous creamy yellow drupes: style short: stigma depressed to form a disc-like
structure, brown.
Distribution. Bomeo.
Habitat. Riversides in lowland tropical rain forest. In the National Park, the species
occurs sparsely along riversides (Fig. 8) and plants are sometimes submerged.
Vernacular name. Ries (Dayak, Belaban dialect).
Uses. Cephalium and leaf are used to cure (stop) hyperurination in children.
Notes. Prior to this, P discostigma was only known from the type, (Martelli 1913;
see also Stone 1993). Although the exact location of Maguc River remains unknown,
the Nieuwenhuis expedition covered most of the central-western part of Kalimantan,
including the numerous tributaries that run through Menikung and Melawi. Thus, it is
possible that the locations where the collections were made during the present study
were in the vicinity of the Maguc River.
Our study agrees with Stone that P matthewsii is a synonym of P. discostigma,
based on the unique disc-like stigma (Fig. 9) and also comparison with the two
collections of P. matthewsii available at BO (Endert 4904 and B.C. Stone 6690). P.
discostigma is now considered a widespread species found along the rivers of lowland
forests in Borneo.
Specimens examined: INDONESIA. West Kalimantan. Sungai Betas Dalam, Km 37,
Menikung, Melawi. 28 Apr 2006, A.P. Keim 75] (BO!); Bukit Siman, Sungai Ela, 30
Apr 2006, A.P Keim 765 (BO!): Central Kalimantan. Maguc River, Nieuwenhuis
Expedition, 1896—1897, Jaheri 662 (holo BO! iso L); Kuala Kuangan. Sei Sampit, 27
Feb 1982. J_J. Afriastini 428 (BO!): East Kalimantan. West Koetai. Km 19, Poekoes
Hill, 14 Nov 1925, FH. Endert 4904 (BO!) — MALAYSIA. Sabah, Sandakan, Ramos
132] (PNH? iso BO!); Leila FR, 17 Mar 1967, B.C. Stone 6683 (BO!).
3. Pandanus epiphyticus Martelli, Nuovi. Bull. Soc. Bot. Ital. 11: 304 (1904). TYPE:
Beccari s.n. (holo Fl), Malaysia, Sarawak, Mt. Mattang, near Kuching, Jun 1866. (Fig.
10)
Pandanus trigonus H.St-John, Pacific Sci. 19(1): 98, f. 207 (1965). TYPE: HN. Ridley
s.n. (holo K: iso SING), Malaysia. Sarawak, Bau. PARATYPE: 4.D.E. Elmer 21022
(BO! C, M, NY, PNH7 SING), Malaysia, Sabah (then British North Borneo), Tawao,
Oct 1922—Mar 1923.
Gard. Bull. Singapore 63(1 & 2) 2011
Fig. 9. Pandanus discostigma Martelli has a fairly globose cephalium consisting of numerous
drupes with distinctive disc-like stigmas that characterise this species, a character which
supports the placement of P. matthewsii into its synonymy. Photo: A.P. Keim, Rugayah & R.
Asmarayani.
Pandanaceae in Bukit Baka Bukit Raya National Park 49
Epiphytic pandan, c. 2 m high. Leaves in rosette, spirally arranged in 3 ranks
(tristichous); each leaf lanceolate-elongate, 262-300 cm long, 9-10 cm wide, apex
acuminate, margin with spines throughout the length: adaxial surface deep green,
glabrous, adaxial ventral pleats not observed; abaxial surface pale green, glabrous,
basal part with spines, recurved spines obvious: leaf sheath white and yellow.
Infructescence terminal, 60-65 cm long, a spike consisting of 10 cephalia, cephalia
not uniform in size, the most terminal one being the smallest; rachis c. 38 cm long,
glabrous; peduncle 22—27 cm long, glabrous; bracts persistent. Cephalium elongate
ellipsoidal, sausage-like, noticeably trigonous, yellowish white to dull greyish white,
consisting of numerous compactly arranged drupes: style very short or sessile, not
pointed; stigma short, not pointed, deep brown.
Distribution. Borneo, Malay Peninsula, and presumably also in Sumatra. Stone (1993)
mentioned that the species used to be fairly frequently seen in Johor, Malay Peninsula
before severe deforestration took place. One of the authors of this current study (APK)
reported seeing P. epiphyticus in the peat swamp forest of Pelalawan in Riau, Sumatra
in 2007, but no collection was made.
Fig. 10. Pandanus epiphyticus Martelli. An infructescence spike consisting of 10 elongate
ellipsoid and trigonous, dull greyish cephalia. Photo: A.P. Keim & R. Asmarayani.
50 Gard. Bull. Singapore 63(1 & 2) 2011
Habitat. This species is an epiphytic plant in lowland tropical rain forest, commonly
found along gorges or riversides. In Bukit Baka Bukit Raya National Park, it is
abundantly found in lowland forest, especially in the foot hills or close to rivers.
Vernacular name. Kajak (Dayak, Belaban dialect).
Uses. Leaves are used for thatching. Local people reported that the cephalium is eaten
by orang utan and gibbons.
Notes. The presence of P. epiphyticus in Indonesian Borneo (Kalimantan) was reported
by Stone (1993) based on a single collection, Kostermans 9115 (BO! duplicate at
L; see Stone 1993) from Nunukan, East Kalimantan. Nunukan is an island off the
mainland East Kalimantan, so the present study confirms the presence of this species
on the mainland.
Specimens examined: INDONESIA. West Kalimantan, Km 37, Sungai Betas Dalam,
Menukung, Melawi, 28 Apr 2006, 4.P. Keim 747 (BO!); East Kalimantan, Nunukan,
Northern part, 19 Dec 1953, A. Kostermans 9115 (BO!) — MALAYSIA. Sabah,
Tawao, Elphinstone Province, Oct 1922—Mar 1923, A.D.E. Elmer 20490 (BO! PNH7*);
A.D.E. Elmer 21022 (para BO! C, M, NY, PNH+ SING).
4. Pandanus pachyphyllus Merrill, J. Str. Br. Roy. Asiat. Soc. 85: 154 (1922). TYPE:
Ramos 1541 (holo PNH# 1so BO! A), Malaysia, Sabah, Sandakan. (Fig. 11 & 12)
Pandanus apicalis H.St.John, Pac. Sci. 22: 523, f. 276 (1968). TYPE: Motley 1247
(holo K; iso BO! SING), Indonesia, Kalimantan (then Dutch Borneo), Banjarmasin,
1857-1858.
Robust shrubby pandan, 1—1.5 m high. Prop roots present, very short. Stem very short,
not obvious. Leaves in a rosette, spirally arranged in 3 ranks (tristichous), 20—more
leaves in a rosette; each leaf lanceolate-elongate, 205-206 cm long, c. 3.5 cm wide,
apex acuminate, margin with obvious spines throughout the length; adaxial surface
deep green, glabrous, adaxial ventral pleats present; abaxial surface pale green,
glabrous, recurved spines present, obvious, very sharp. /nfructescence terminal,
solitary, c. 10 cm long; rachis c. 5 cm long, glabrous; peduncle c. 5 cm long, glabrous.
Cephalium globose, yellowish green; style pointed, ascending; stigma pointed, sharp;
in general appearance the cephalium superficially resembles the fruit of durian (Durio
zibethinus, Malvaceae).
Distribution. Borneo.
Habitat. Foothills and gorges in lowland tropical rain forest. In Bukit Baka Bukit Raya
National Park, the species is commonly found in gorges. Although abundant, most
were not fruiting.
Pandanaceae in Bukit Baka Bukit Raya National Park
Fig. 11. Pandanus pachyphyllus Merrill: a robust shrubby pandan with a very short, almost
invisible, stem, and a terminal infructescence with a globose cephalium. Photo: A.P. Keim,
Rugayah & H. Rustiami.
2 Gard. Bull. Singapore 63(1 & 2) 2011
Vernacular name. Selinsik (Dayak, Belaban dialect).
Uses. Leaves are used for mats.
Notes. The record of P. pachyphyllus in Kalimantan was only based on the type of
P. apicalis, which was placed as a synonym of P. pachyphyllus by Stone (1978).
Observations made on the types of P. apicalis and P. pachyphyllus at BO, and the
collection made in this current study, indicate that there is no substantial difference.
The distribution of this species could be more widespread in Borneo than previously
thought.
Specimen examined: INDONESIA. Central Kalimantan, Km 44, Sungai Wah, Kota
Waringin Timur, Kasongan, Sinamang Mentikay, 1 May 2006, 4.P. Keim 767 (BO!).
Fig. 12. Pandanus pachyphyllus: a terminal infructescence with extremely short peduncle and
solitary yellowish green durian-like cephalium. This photo also shows the conspicuous spines
all along the leaf margin. Photo: A.P. Keim, Rugayah & H. Rustiami.
5. Pandanus yvanii Solms, Linnaea 42: 20 (1878). TYPE: Yan s.n. (holo Herb.
Delessert), Malaysia, Malay Peninsula, Malacca. (Fig. 13—16)
Pandanus motleyanus Solms, Linnaea 42: 21 (1878). SYNTYPES: Korthals s.n.
(L), Indonesia, Kalimantan (then Dutch Borneo); Motley 1057 (K), Malaysia, North
Borneo.
Pandanaceae in Bukit Baka Bukit Raya National Park 53
Pandanus ridleyi Martelli, Bull. Soc. Bot. Ital. (1904) 303. SYNTYPES: Cantley s.n.
(K, SING), Malaysia, Sungai Ujong, Gunong Berumban; Kunstler s.n. (K, SING),
Malaysia, Perak.
Pandanus brevifolius Martelli, Bull. Soc. Bot. Ital. (1914) 302. TYPE: Beccari PB 273
(holo FI), Malaysia, Sarawak, Siul near Kuching.
Pandanus sigmoideus H.St.John ex B.C. Stone, Fed. Mus. J. 17: 124, f. 15 (1972).
TYPE: Brunig S12384 (holo L; iso K, SAR), Malaysia, Sarawak, Marudi, Baram,
Lobok Pasir, Apr 1961.
Slender clustered tree pandan, 2—3 m high, commonly forming dense thickets. Prop
roots short, 20 cm or less. Stem unbranched (4.P. Keim 777) or branched in the terminal
part (4.P. Keim 779), slender, older bark dark purplish brown, spiny, diameter c. 1.5
cm. Leaves in a rosette, spirally arranged in 3 ranks (tristichous); each lanceolate-
elongate, 40-45 cm long, I—1.5 cm wide, apex acute to acuminate, margin with spines
throughout the length; adaxial surface green, glabrous, adaxial ventral pleats absent;
abaxial surface pale green, glabrous, recurved spines present, small, brown; leaf sheath
yellowish green to yellow. /nfructescence solitary, terminal, 17—20.5 cm long; bracts
persistent, each 10—28 cm long, c. 2.5 cm wide, brown to deep brown. Cephalium
ellipsoidal elongated, creamy yellow to dull creamy yellow or dull yellowish orange
(A.P. Keim 779), 7.5—11 cm long, 13—14 cm in circumference; style ascending, needle-
like, 2-2.5 mm long; stigma pointed, sharp.
Distribution. Malay Peninsula, Sumatra (including Bangka Island), and Borneo.
Previously, P. yvanii was known as an endemic species of Malacca in the Malay
Peninsula; thus we newly record the species for Sumatra, Bangka Island, and Borneo.
In Bukit Baka Bukit Raya National Park, the species is hardly noticeable at higher
altitudes but in the lowland open areas in full sun, and peat swamps in the vicinity
of the National Park, the species is abundant (Fig. 13) and this is in accordance with
Stone (1966).
Habitat. Peat swamps, where it usually forms dense thickets. Pandanus yvanii
sometimes grows along riversides or in riverine forest.
Vernacular name. Rasau or Rassau (Dayak, Nanga Pinoh).
Uses. Not recorded.
Notes. Pandanus yvanii is a common species of peat swamp forests in the Malay
Peninsula, Sumatra and Borneo (Stone 1993, as PR. motleyanus). In a number of places
in these areas, the species is widely known by the vernacular name rassau. This
vernacular name was first recorded by Kurz (see Kurz ex Miquel 1866) for P. helicopus
Kurz ex Miq.; however, a string of morphologically similar taxa have been published
54 Gard. Bull. Singapore 63(1 & 2) 2011
Fig. 13. Pandanus yvanii Solms. A. Collector holding a severed stem in Nanga Pinoh. B. A
dense thicket of stems. C. Stems with sharp nodules on the outer surface (and a characteristic
deep reddish brown when fresh). D. A solitary, terminal elongate-ellipsoidal young cephalium
(creamy yellow when fresh) with persistent deep brown bracts; notice the obvious dark spines
along the leaf margin. Photos: A.P. Keim, Rugayah & H. Rustiami.
Nn
N
Pandanaceae in Bukit Baka Bukit Raya National Park
with almost all bearing “rassau” as one of their vernacular names. The classic case
invoives P. helicopus, P. motleyanus, and P. yvanii, which share a common habitat,
peat swamps.
Unlike P. helicopus, the protologues of both P. motleyanus and P. yvanii
(Solms 1878) describe fewer morphological details. The situation is worsened by the
fact that instead of comparing P. motleyanus or P. yvanii with P. helicopus, Solms
(1878) compared P. motleyanus only with P. yvanii, which was also published as a
new species in the same publication, only a page earlier. The identity of P. yvanii as
a species in its own right was first questioned by Hooker (1894), who described the
morphological characters given in the protologue of P. yvanii as being worthless for
the identification of a species of Pandanus. Apparently due to the limitted specimens
available, Warburg (1900) distinguished P. motlevanus from P. yvanii based only on
a minor morphological character, the colouration of the leaf spines (P. motlevanus
has dark brown leaf spines; P. yvanii, light brown) and distribution (P. motlevanus
in Borneo, P. yvanii in Malacca in the Malay Peninsula). Warburg numbered the two
species one after the other (P. yvanii was 107, P. motleyanus was 108), thus suggesting
their close affinity. Ridley (1925) followed Hooker in arguing that P yvanii was too
imperfectly described to identify the species and placed it as a synonym of Martelli’s
P. ridleyi.
Prior to the present study, P. yvanii was known only from a very few number
of localities. BO has a specimen collected by Teijsmann in Muntok, Bangka Island
(Teijsmann s.n., BO! Fig. 14 & 15), which was named P. yvanii by Stone, but never
published. This specimen possesses many morphological characters that match the
protologue of P. yvanii, particularly the leaf and cephalium dimensions, and also the
appearance of the style. We now have a firm record for the first time of its presence
outside the type locality.
In contrast to Ridley (1925), this study places P. rid/eyi in the synonymy of P.
yvanii. Aithough there is a noticeable difference in the length of the cephalia, it is still
within the range of P. yvanii and the synonymous P. motleyanus.
We place P. sigmoideus as a synonym of P. yvanii. St. John & Stone (1972) and
Stone (1993) regarded P. sigmoideus as a distinct species from the then P. motleyanus
based on the nature of leaf spines (reduced or obsolute, at most | mm long in P.
sigmoideus, compared to developed, |—2 mm long in P. motleyanus) and pollen surface
(smooth in P. sigmoideus, minutely spinulose in P. motleyanus). We regard these two
characters as less important and not discrete.
Martelli (1914) mentioned four distinctive characters of P. brevifolius: low
shrub habit, brownish red stem, small leaves (hence the epithet “brevifolius”), and
glabrous leaf margin. Observations made on specimens from Bukit Baka Bukit Raya
and 7Jeijsmann s.n. (BO! Fig. 14 & 15) indicate that those characters are also found in
P. yvanii. Indeed, in the field P. yvanii can be seen as low shrubs with short brownish
red stems, small narrowed glabrous leaves and subglobose cephalia; thus there is not
sufficient reason to place P. brevifolius as a species on its own and it is regarded here
as a synonym of P. yvanii.
56 Gard. Bull. Singapore 63(1 & 2) 2011
fees eae
“Dewar me BO~ 1576042
rian dt
Hitt Hah a i ia
Fig. 14. Pandanus yvanii Solms (Teijsmann s.n. “Muntok, Bangka”, BO). This specimen shows
two characters that match the protologue of P. brevifolius: small leaves without spines along the
margin (see Martelli 1904). Photo: A.P. Keim.
Pandanaceae in Bukit Baka Bukit Raya National Park 57
mrs
| - Remittendum | BO- 1576043
REVISIO PANDANACEARUM
Pandanus_ yvanii_Serms__1878 —_——
co Shincwa sae,
{eet a ol Stone + f6?.
HERB. HORT. BOT. BOG.
ya
vss
Archip. Ind.
Det.
Leg.
_ /% j Herb. Hort. Bot. Bog.
Rogatar hoc as) cum nomine remittere
pet art ted aneand eS a=
determ.
190
Golrcs@ Cable
Le wnt nt Zee
Fig. 15. Pandanus yvanii Solms (Teijsmann s.n. “Muntok, Bangka”, BO). This specimen shows
two characters that match the protologue of P. brevifolius (see Martelli 1904): the deep reddish
brown slender stem and small cephalium (apparently immature). Photo: A.P. Keim.
58 Gard. Bull. Singapore 63(1 & 2) 2011
gap
This present study does not agree with Stone (1993) in assigning P. fruticosus
as a synonym of P. motleyanus. Pandanus fruticosus shares more similarities with P.
helicopus in being a more robust pandan with longer (up to 5-10 mm long) drupes;
whereas P. motleyi (or P. yvanii) is a more slender pandan with small (3 mm long)
drupes. We regard P. fruticosus as a synonym of P. helicopus. Stone (1993), however,
did recognise P. yvanii (then P. motleyvanus) and P. helicopus as two distinct species.
Despite both being found in peat swamps, P. yvanii and P. helicopus occupy
different ecological niches (Partomihardjo 2010 pers. comm.). Pandanus yvanii 1s
commonly found forming dense thickets further inland in the peat swamps and has
never been found on open riverbanks. On the contrary, P helicopus 1s always found
forming dense thickets along open riverbanks and streamsides in peat swamp areas
and has never been found further in the forest. In other words, the two species are not
fully cohabitant.
Also, the two species can almost instantly be identified in the field. Pandanus
yvanii 1s a low shrub, possessing deep purplish brown stems, wide leaf scars, and
bright yellow to yellowish green leafsheaths (Fig. 16). On the contrary, P. helicopus
is larger, has pale brown stems, narrow leaf scars, and eye-catching bright orange-red
leaf sheaths (Fig. 16).
Pandanus pumilus H.St.John shares many morphological characters and
a habitat preference with P yvanii; however, P. pumilus possesses a spike-like
infructescence consisting of 4—5 cephalia (St. John 1961). Pandanus yvanii, so far
as known, always has an infructescence with a solitary cephalium. The possibility
of a variable pistillate inflorescence or infructescence structure needs to be further
investigated. Until better data becomes available, we merely suggest that P pumilus
has a close affinity with P. yvanii.
Specimens examined: INDONESIA. West Kalimantan, Soengei Kelassar, 1893—
1894, H. Hallier 1549 (BO! L); West Koetei, 30 Nov 1925, F'H. Endert 5413 (BO!
L); Pontianak, Sei Raja, 12 Mar 1931, Mondi 9 (BO! L); Pontianak, Kampoeng
Mandor, 23 Dec 1931, J.P. Schuitemaker 139 (BO! L); Mampawah, Mengkatja, 29
Sep 1948, M. Enoh 399 (BO! K, L); Nanga Pinoh to Sintang, Kampung Pandan, 5
May 2006, 4.P. Keim 777 (BO!); A.P. Keim 779 (BO!); A.P. Keim 780 (BO!); Central
Kalimantan, Sampit, 22 Aug 1940, P. Buwalda 7647 (BO! L); Kapuas, Tewah, Desa
Kasintu, 13 Oct 1999, S. Riswan et al. TWH 025 (BO!); East Kalimantan, Nunukan
Island, Tarakan, Samenggaris, Dec 1912, Amdjah 1077 (BO!); Samarinda, Sungai
Mukun, near Sango-Sango, 05 Aug 1952, W. Meijer 1100 (BO!); Nunukan Island, N
of Tarakan, SE of Kampong, 22 Nov 1953, W. Meijer 2303 (A, BO! K, L, P, PNH,
SING); Nov—Dec 1953, W. Meijer 2308 (BO!); Nov—Dec 1953, W. Meijer 2351 (BO!,
L); 13 Dec 1953, W. Meijer 2479 (BO!); West Kutei, Mount Palimasan near Tabang,
on Belajan River, 09 Sep 1956, 4. Kostermans 12823 (BO!, L); South Kalimantan,
Bandjarmasin-Martapoera, Km 19, 04 Oct 1939, Polak 439 (BO!); Banjarmasin to
Martapura, Km 22, 26 Jun 1974, J. Dransfield & G. Hambali 4312 (BO!, L); 26 Jun
1974, J. Dransfield & G. Hambali 4313 (BO!, L); Sumatra, Bangka Island, Muntok,
Teijsmann s.n. (BO!).
tn
\O
Pandanaceae in Bukit Baka Bukit Raya National Park
=
«'
>
%
ti
binds a
Se
~
DD pacaaah 4
Fig. 16. Pandanus yvanii (A) and P. helicopus (B) compared. Pandanus yvanii individuals are
more slender compared to P. helicopus (photo taken in the Sebangau National Park, Central
Kalimantan). Stems of P. yvanii are characteristically deep reddish brown with wider leaf scars;
those of P. helicopus are bright reddish brown with distinctively narrow, dense and crowded
leaf scars. The leaf sheath of P. yvanii is yellow or yellowish green, those in P. helicopus are
strikingly reddish brown. Photos: A.P. Keim, Rugayah & H. Rustiami (A); E.A. Widjaja & M.
Amir (B).
60 Gard. Bull. Singapore 63(1 & 2) 2011
ACKNOWLEDGEMENTS. The authors would like to express their deepest appreciation to
the field officers of the Bukit Baka Bukit Raya National Park for their constant support and
help during exploration. We are very grateful to our colleagues Arief Supriatna, Elizabeth A.
Widjaja, Rani Asmarayani, Susi Susiarti, Tutie Djarwaningsih, and Yessi Santika for letting
us use their amazing photographs. We also thank Tukirin Partomihardjo for good discussions
and suggestions regarding the ecology of rassau in Kalimantan and Sumatra. This paper is
dedicated to the people of Belaban village, with whom it was our privilege to experience their
most precious heritage: the lush tropical rain forest of Bukit Baka.
References
Anonymous (2002) Data dan Informasi Kehutanan Propinsi Kalimantan Barat
(electronic version). Jakarta: Pusat Informasi dan Statistik Kehutanan - Badan
Planologi Kehutanan Departemen Kehutanan Republik Indonesia.
Hemsley, R. (1894) A new genus of Pandanaceae. J. Linn. Soc. Bot. 30: 216.
Hooker, J.D. (1894) The Flora of British India, vol. 6, Orchidaceae to Cyperaceae.
London: L. Reeve & Co.
Kanehira, R. (1938) A contribution to our knowledge of Pandanus of Malaya and
Borneo. J. Jap. Bot. 14(7): 429-438.
Keim, A.P. (2009) Three new species of Freycinetia (Pandanaceae) from Kalimantan,
Indonesia. Reinwardtia 13(1): 15—20.
Keim, A.P, Purwanto, Y. & Muirmanto, E. (2006) Keanekaragaman Suku Palem
(Palmae) dan Pandan (Pandanaceae) di Lahan Gambut PT. Putra Riau Perkasa
& Balai Kayang Mandiri (Sinar Mas Group), Riau Daratan. Bogor: Herbarium
Bogoriense (mimeograph).
Keim, A.P, Susiarti, S. & Amir, M. (2008) Taksonomi Pandanaceae Pulau Séram, pp.
1281—1326. Cibinong: Laporan Teknik Pusat Penelitian Biolog: LIPI.
Kurz, S. (1866) Pandanaceae. In: Miquel, F. (ed) Flora van Nederlandsch Indié. Ann.
Mus. Bot. Lugduno-Batavi 2: 52.
Martelli, U. (1904) Pandani Asiatici Nuovi. Bull. Soc. Bot. Ital. 11: 298-304.
Martelli, U. (1910) Enumerazione delle Pandanaceae. Part 1: Freyvcinetia. Webbia 3:
307-327.
Martelli, U. (1913) Enumerazione delle Pandanaceae. Part 2: Pandanus. Webbia 4 (1):
140.
Martelli, U. (1914) Le specie e varieta nuove di ‘Pandanus’ menzionate nella
‘Enumerazione delle Pandanaceae’. Webbia 4: 399-435, t. 1-43.
Martelli, U. (1929) Pandanaceae. In: Merrill, E.D. (ed) Plantae Elmerianae Borneenses.
California: University of California Publ. in Botany.
Merrill, E.D. (1908) Philippine Freycinetia. Philipp. J. Sci., C. 3: 307-315.
Merrill, E.D. (1922) New or Noteworthy Bornean Plants, Pandanaceae, Pandanus. J.
Straits Branch Roy. Asiat. Soc. 85: 152-155.
Nieuwenhuis, A.W. (1898) La récente expedition scientifique dans l’ile de Borneo.
Tijdschr. N. I. Taal-, Land- en Volkenk. 40: 508-541.
Pandanaceae in Bukit Baka Bukit Raya National Park 61
Reuler, H. van (1987) Introduction. In: Nooteboom, H.P. (ed) Report of the 1982—
1983 Bukit Raya expedition. Leiden: Rijksherbarium.
Ridley, H.N. (1907) Materials for a Flora of the Malay Peninsula, vol. 2. Singapore:
Methodist Publishing House.
Ridley, H.N. (1925) The Flora of the Malay Peninsula, vol. 5. London: L. Reeve &
Co.
Solms-Laubach, H. (1878-1879) Monographia Pandanacearum. Linnaea 42: 85-105.
St. John, H. (1961) Revision of the genus Pandanus Stickman, part 7: New species
from Borneo, Papua, and the Solomon Islands. Pacific Sci. 15(4): 576-590.
St. John, H. (1963) Revision of the genus Pandanus Stickman, part 15: Malayan
species described by H.N. Ridley. Pacific Sci. 17(3): 329-360.
St. John, H. (1965) Revision of the genus Pandanus Stickman, part 17: Species mostly
new in Borneo, Cambodia, and Vietnam. Pacific Sci. 19(1): 96-112.
Steenis, C.G.G.J. van (1950) Flora Malesiana. Vol. |. Ser. 1: Spermatophyta. Jakarta:
Noordhoff-Kol ff.
Stone, B.C. (1967a) A new species of Pandanus from Sarawak with notes on section
Multidens St. John. Fed. Mus. J. 11: 113-119.
Stone, B.C. (1967b) A new species of Pandanus from East Malaysia. Malayan Sci. 3:
23-25.
Stone, B.C. (1967c) Materials for a monograph of Freycinetia (Pandanaceae) I. Gard.
Bull. Singapore 22(2): 129-152.
Stone, B.C. (1968) Materials for a monograph of Freycinetia Gaud. IV. Subdivision of
the genus with fifteen new sections. Blumea 16(2): 361-372.
Stone, B.C. (1969) Materials for a monograph of Freycinetia (Pandanaceae) VIII. A
revised list of Philippine species, with critical notes and some new taxa. Webbia
23: 597-607.
Stone, B.C. (1970a) Materials for a monograph of Freyvcinetia Gaud. (Pandanaceae).
VI. Species of Borneo. Gard. Bull. Singapore 25(2): 209-233.
Stone, B.C. (1970b) Materials for a monograph of Freycinetia Gaud. (Pandanaceae).
V. Singapore, Malaya, and Thailand. Gard. Bull. Singapore 25(2): 189-207.
Stone, B.C. (1970c) Malayan climbing pandans—the genus Freycinetia in Malaya.
Malayan Nat. J. 23: 84-91.
Stone, B.C. (1972) Studies in Malesian Pandanaceae. VII. A review of Javanese
Pandanaceae, with notes on plants cultivated in the Hortus Bogoriensis.
Reinwardtia 8(2): 309-318.
Stone, B.C. (1975) Notes on the Pandanaceae of Ceylon, with a review of the species
cultivated in the Royal Botanic Gardens, Peradeniya, and those found in its
herbarium. Ceylon J. Sci. (Biol. Sci.) 11(2): 109-122.
Stone, B.C. (1978) Revisio Pandanacearum. Part 1: Pandanus subgenera Coronata
and Acrostigma. Fed. Mus. J. 23: 1-74.
Stone, B.C. (1980) The vegetation and plant communities of Pulau Balambangan,
Sabah, East Malaysia. J. Malayan Branch Roy. Asiat. Soc. 53(1): 68-89.
62 Gard. Bull. Singapore 63(1 & 2) 2011
Stone, B.C. (1982) Two new species of Pandanus from Gunung Mulu National Park.
In: Jermy, A.C. & Kavanagh, K.P. (eds) Notulae et Novitates Muluenses. Bot. J.
Linn. Soc. 85: 31-35.
Stone, B.C. (1983a) A guide to collecting Pandanaceae (Pandanus, Freycinetia and
Sararanga). Ann. Missouri Bot. Gard. 70: 137-145.
Stone, B.C. (1983b) Studies in Malesian Pandanaceae 19: New species of Freycinetia
and Pandanus from Malesia and Southeast Asia. J. Arnold Arbor. 64(2): 309-324.
Stone, B.C. (1993) Studies in Malesian Pandanaceae 21: The genus Pandanus in
Borneo. Sandakania 2: 35-84.
Warburg, O. (1900) Pandanaceae. In: Engler, A. (ed) (1898-1923) Das Pflanzenreich.
Vol. 4. Part 9(3): 1-100.
Gardens’ Bulletin Singapore 63(1 & 2): 63-70. 2011 63
A taxonomic study of the Pandanus furcatus
and P. tectorius complexes (Pandanaceae) in Java
Sri Endarti Rahayu', Alex Hartana’, Tatik Chikmawati?
and Kuswata Kartawinata*
‘Graduate School of Bogor Agricultural University,
Biology Department, National University, Indonesia
endarti2004(@yahoo.com (corresponding author)
*Bogor Agricultural University, Indonesia
* Herbarium Bogoriense, Botany Division, Research Center for Biology,
Indonesian Institute of Sciences, Cibinong Science Center (CSC),
Jl. Raya Jakarta-Bogor Km 46, Cibinong, Bogor 16911, Indonesia;
and Botany Department, Field Museum, Chicago, Illionis, U.S.A.
ABSTRACT. Current taxonomic problems in Pandanus in Java include the interpretation
of the Pandanus furcatus complex as well as the P. tectorius complex. A study of general
morphological, stomatal and molecular characteristics (viz., the noncoding chloroplast
intergenic spacer region atpB-rbcL) showed that P. bantamensis Koord., P. pseudolais Warb.,
and P. scabrifolius Martelli, previously considered synonyms of P. furcatus, and P. tectorius
var. littoralis Martelli and P. odoratissimus L.f. are all distinct species.
Keywords. Anatomy, atpB-rbcL, Java, morphology, Pandanus furcatus complex, Pandanus
tectorius complex
Introduction
The term species complex was used to describe a species aggregation sharing specific
morphological and molecular features (Judd et al. 1999). Within such a complex, a
complicated morphological overlap, without any discontinuities, has led to taxonomic
difficulty (Pak and Kawano 1992). Although their taxonomic affinity may be difficult
to determine, some form of taxonomic resolution is desirable.
According to Stone (1972), Pandanus Parkinson in Java contains many
rather problematic species. He suggested that detailed studies were required to obtain
a more refined taxonomic scheme. The main problem, as far as Javan plants are
concerned, appears to be the status of taxa which are given as synonyms of P. furcatus
Roxb. by Backer and Bakhuizen van den Brink f. (1968). These synonyms include
P. bantamensis Koord., P. oviger Martelli, P, pseudolais Warb. and P. scabrifolius
Martelli. Also, P. odoratissimus L.f was thought to be synonymous with P. tectorius
var. littoralis Martelli. Stone (1972) made a short revision of Pandanaceae in Java
based on herbarium specimens in the BO herbarium and living plants cultivated in the
Hortus Bogoriensis, attempting to develop a more stable species concept for Pandanus
64 Gard. Bull. Singapore 63(1 & 2) 2011
furcatus Roxb. In Stone’s opinion, Backer’s species concept for Pandanus furcatus was
far too comprehensive and required readjustment because distinct species were lumped
with Pandanus furcatus. Stone regarded P. bantamensis, P. oviger, P. pseudolais and
P. scabrifolius as four different species, whereas the status of P. tectorius var. littoralis
and P. odoratissimus was still more or less in question. The circumscription of these
taxa are reviewed in light of new studies based on material from our recent fieldwork
in Java. A practical difficulty here was that almost all morphological characters and/or
character states used for evaluating a species complex are only slightly differentiated
from one another and usually show considerable overlap. Therefore the identification
of taxa within a species complex usually required a combination of several characters.
The aim of the study was to provide a taxonomic resolution of the P. furcatus complex
and P. tectorius complex, based on general and stomatal morphology, and a molecular
approach utilising the noncoding chloroplast intergenic spacer region atpB-rbcL.
Materials and methods
Studies of herbarium specimens were conducted in the Herbarium Bogoriense
(BO), Herbarium of the Royal Botanical Gardens, Kew (K) and National Herbarium
of the Netherlands, Leiden (L). Observations of living plants and stomatal studies
were undertaken at BO, while the molecular data was analysed in the Van der Klauw
Laboratory, Leiden.
Data on morphology were collected from herbarium specimens and fresh
field collections. The procedure for treating morphological variation followed that
described by Rifai (1976) and Vogel (1987). Measurements were taken from spirit-
preserved material, dried herbarium specimens and living material. Floral parts were
measured from spirit-preserved material, and material rehydrated (by boiling) from
dried specimens.
Leaf stomatal characteristics were investigated by first fixing the leaves (a
small part of the middle to basal area) in FAA. Paradermal sections were taken from
the upper and lower surfaces of leaves, then stained with safranin 1% in water and
mounted in glycerine (Johansen 1940).
Genomic DNA was extracted from leaf material dried in silica gel according
to the protocol described by Doyle & Doyle (1987). Double-stranded DNA was
directly amplified by PCR. Reaction volumes were 25 ul and contained 2.5 ul PCR
buffer, 2.5 ul dNTPs, 1 ul each of the 5 mM primers, 0.3 ul Zag Pol and 12.7 pl
ddH,O. Approximately 5 wl genomic DNA was added to the PCR mixture. PCR
was performed 3 min at 94°C for the activation of the polymerase, followed by
35 cycles of 49 sec at 94°C, 45 sec at 55°C, 2 min at 72°C, with a final extension
period of 10 min at 72°C. The primers used in this study for atpB-rbcL intergenic
spacers are (forward 5’-GAAGTAGTAGGATTGATTCTC- 3’) and (reverse 5’-
TACAGTTGTCCATGTACC AG-3’). The PCR product was checked on 1% agarose
gel, and purified using a purification kit of Wizard SV Gel and PCR clean up system
(PROMEGA) following the manufacturer’s protocol prior to sequencing. The DNA
Pandanus complexes in Java 65
concentration was measured with the nanodrop. Cycle sequencing was performed by
Macrogen Korea. The sequences were edited using sequencher 4.6 and MEGA 3.0
(Kumar et al. 2004).
Results
Morphological characters of the Pandanus furcatus complex
For this study, we found only three of the four species recognised within the P. furcatus
complex, 1.e., P- bantamensis, P. pseudolais and P. scabrifolius. Several characters
studied are compared in Table 1. Prickles on prop-roots did not appear to display
consistent differentiation among taxa, although P. scabrifolius did not have prickles on
the prop-roots. Leaf dimensions were highly variable. Pandanus pseudolais tended to
have longer leaves, whereas the leaves of P. scabrifolius had a shorter range of length
measurements. Pandanus bantamensis had intermediate leaf length. The number of
drupes per cephalium of P. pseudolais was higher than in the other two taxa. Pandanus
scabrifolius had longer drupes compared with the other two taxa.
Four other characters studied were leaf base colour, peduncle shapes, fruit
shapes and style shapes, that appeared to be useful for discriminating among species
(Table 1). Pandanus bantamensis and P. pseudolais had a reddish brown leaf base, and
P. scabrifolius a yellowish white leaf base. The peduncle was stout and quite straight
in P. scabrifolius, and curved in the other two taxa. Pandanus scabrifolius also had
a broadly ellipsoid cephalium, compared to the narrower cephalia in the other two
species; it also had comparatively shorter style bifurcations (divisions a third the style
length or less) than in the other two species (divisions about half the style length).
Table 1. The Pandanus furcatus complex: some morphological features of the vegetative parts,
peduncle, fruiting cephalium, drupes and style in the three taxa studied.
Character
P. bantamensis
P. pseudolais
P. scabrifolius
Leaf length (cm)
Prickles on prop-
root
Number of drupes
Drupe length (cm)
Leaf base colour
Peduncle shape
Cephalium shape
Style bifurcation
216-441
present, prickles in
rows
475-795
DANS)
Reddish brown
Slightly curved at
the base
Narrow-ellipsoid
Divisions about half
the style length
299.4-574.5
present, prickles in
rows
724-1053
2.6—4.5
Reddish brown
Strongly curved at
the base
Narrow-ellipsoid
Divisions about half
the style length
204-372
smooth
473-483
4.85.2
Yellowish white
Straight
Broad-ellipsoid
Divisions a third of
the style length or less
66 Gard. Bull. Singapore 63(1 & 2) 2011
Morphological characters of the Pandanus tectorius complex
Leaf dimension, leaf shape, cephalium size and shape, and the number of phalanges per
cephalium did not have any consistent differences between the taxa analysed (Table
2). However, P. tectarius var. littoralis had leaf apex prickles only on one surface (P.
odoratissimus has prickles on both surfaces of the leaf apex); and carpel apices that
were essentially fused, withour any deep grooves between carpels (P. odoratissimus
had carpel apices that were free, leaving deep grooves between carpels) (Table 2).
Table 2. The Pandanus tectorius complex: morphological characteristics of P. tectorius var.
littoralis and P. odoratissimus.
Crmaeias P. AEDT var Ries . P. aaiomasn
“Leaflength(em) = MI2-199 98-126 af
Leaf shape ligulate or sword-shaped ligulate or sword-shaped
Cephalium size (cm) 26 x 25 22% 25
Cephalium shape broad-ellipsoid broad-ellipsoid
Number of phalanges 79-83 73-83
Prickles on leaf apex only on one side of leaf, the — on both sides of leaf
other side smooth
Carpel apices fused, without deep grooves not fused, with deep
between carpels grooves between carpels
Stomatal characters
Five classes of stomatal features (Tomlinson 1965, Kam 1971) are known in Pandanus,
depending on the number of papillae that develops on the subsidiary and neighbouring
cells. In the Pandanus furcatus complex, P. bantamensis, P. pseudolais and P.
scabrifolius had stomatal Class 2 (papillae only occurring on lateral subsidiary cells),
Class | (papillae absent: unspecialized stomata) and Class 3 (papillae in both terminal
and lateral subsidiary cells) of Tomlinson (1965), respectively. Within the Pandanus
tectorius complex, P. tectorius var. littoralis had stomatal Class 4 (papillae occurring
in both lateral subsidiary and neighbouring epidermai cells) and P. odoratissimus had
stomatal Class 2 (papillae only occurring on lateral subsidiary cells).
Molecular characteristics
Although the small number of taxa investigated would not be expected to yield any
meaningful phylogenetic analysis, the atpB-rbcL intergenic spacer region provided
potential markers for distinguishing the different species in both of the complexes.
Pandanus complexes in Java 67
There were 29 polymorphic sites in this region for the three species of the Pandanus
furcatus complex investigated (Table 3), of which 19 sites were different between P.
bantamensis and P. pseudolais, 24 sites were different between P. pseudolais and P.
scabrifolius, and 21 sites were different between P. bantamensis and P. scabrifolius.
Likewise, there were 71 polymorphic sites in this region for the two species of the
Pandanus tectorius complex investigated (Table 4).
Table 3. The Pandanus furcatus complex: polymorphic sites in the atpB-rbcL intergenic spacer
region.
Species Nucleotide . eer site at indicated position
Ce SS CC CI ea ep
P. bantamensis wane GG GAA GT AAT 1 T A
P. pseudolais Werte Om OUA © eA AA A TA 1G, (Gea Aer A
P. scabrifolius amG? Coes ihe Ee
>
(>)
Q
--
Q
4
Q
>
Q
4
>
4
Nucleotide at polymorphic site at
indicated position
Species
iS (we) TE beh eo!
OS Se SP UN- NN wD |}
QPS oN shi wy on @) va) a)
Ravanamensis, ) Gb AG Ae € ff €
P. pseudolais T
Ge | 552
P. scabrifolius Ag Dyke Ad Ta An 'G
Discussion
Pandanus furcatus complex
The species recognised within the Pandanus furcatus complex shared some
morphological similarities, but could be distinguished by leaf base colour, peduncle
shape, fruit shape, style shape, and stomatal characteristics. A number of sites in the
atpB-rbcL intergenic spacer region were also polymorphic for these species. We
conclude that Pandanus bantamensis, Pandanus pseudolais and Pandanus scabrifolius
are three distinct species. Our study thus corroborates the conclusions of Stone (1972).
68 Gard. Bull. Singapore 63(1 & 2) 2011
Table 4. The Pandanus tectorius complex: polymorphic sites in the atpB-rbcL intergenic
spacer region.
Nucleotide at polymorphic site at indicated position
Species
Mm= OA eG =
a A Oo = Ao oo SS He a Ao oS
(om) (02) <eay Ne) No) Oo Cc CO 600 _ — = — SING
P. tectorius var. littoralis AG TY 7 © TW at AW Ay Gaal Saas Com Came ee ©
P. odoratissimus G TA 3G “A 7 1GY TT AA A Ge Ge
; Nucleotide at polymorphic site at indicated position
Species
DAosaranmne Aoakt 4 © SOM © =
SNM Matt COM KH HA DOM ST
ON N NN N AN NN AN N NN loo) loa) (oe) (oe) wt bs +
P. tectorius var. littoralis T A GG AAT Poe GeAr it SAS Ge Cae
P. odoratissimus A G TY TCT As AD AG LMG Tera Awe
; Nucleotide at polymorphic site at indicated position
Species =e
wr Wm oO > | ta => —! — mM Co OD Oo HA © — CO
Yor No IS iS oS S| aS Se Ee tS ce es CF G&G &
+ + + bs 2 wy LV ay LV ay LV ay yD yD Dy ~y Dy ~y ~y No)
P. tectorius var. littoralis A; Ge As AnkG- WT: CoA CA AC © Ga Age Agama
P. odoratissimus G AV Te) Te C2 °AG AS ee Ge Ga Ca CCG
. oe ; Nucleotide at polymorphic site at indicated position
Species DE os
AnmtaMaaAOCwWOn TMA R © OE
onto Aon SS SS ES SES Oo SS j] wh ww
\O Oo © No) ti | ~~ i ~~ ls FG 200s OC OC OO
P. tectorius var. littoralis G A Cow «E- AVE EFAS A CC WP Gr eeeAeG
P. odoratissimus A T Tae sAU“G sGy Ac GGG sO eA Ages lame
Nucleotide at polymorphic
site at indicated position
Species
"395
896
397
90
902
903
P. tectorius var. littoralis A A
CG
P. odoratissimus
|4 >
A
©
l4 a
(er fe
| >
Pandanus complexes in Java 69
Key to three species of the Pandanus furcatus complex
la. Leaf base reddish brown; cephalium narrowly ellipsoid (the length almost 3 times
the width); style bifurcations about half the style length 0.0.0.0... eeeeeseeeeeee 2
lb. Leaf base yellowish white; cephalium broadly ellipsoid (the length at most twice
the width); style bifurcations a third of the style length or less ...... P. scabrifolius
2a. Peduncle slightly curved at the base; cephalium with 475—795 drupes; stomata type
“Class 2”, with papillae on lateral subsidiary cells ..................... P. bantamensis
2b. Peduncle strongly curved at the base; cephalium with 724-1053 drupes; stomata
em ASS lee VILMOU LD APN AGC. <2. acesoncscaeieacsaseeceescaceadeadavecasncaseceiees P. pseudolais
Pandanus tectorius complex
Although Pandanus tectorius var. littoralis and P. odoratissimus are very closely
related (Stone 1994). Stone (1967) had proposed delimiting P. odoratissimus by just
two characters: fleshy shoulders on phalanges and large white spines on the leaves.
In our study, P. odoratissimus was found to have large white spines on the leaves as
described, but the same could also be found in P. tectorius var. littoralis. Pandanus
odoratissimus also does not have fleshy shoulders on the phalanges. Our observations
tally with those of Stone (1979) who stated that P. odoratissimus has large white spines,
but does not have fleshy shoulders. In this study, we have found several contrasting
morphological characters that could be useful for distinguishing these two species.
Pandanus tectorius var. littoralis has a leaf apex that is prickly on only one side of
the leaf and fused carpel apices without any significant grooves in between; whereas
P. odoratissimus has a leaf apex that 1s prickly on both sides of the leaf and free
carpel tips separated by deep grooves in between. As many as 7] sites in the atpB-rbcL
intergenic spacer region were also found to be polymorphic between these two taxa.
We therefore accept these two taxa as specifically distinct.
Key distinguishing Pandanus tectorius and P. odoratissimus
la. Leaf apex prickly only one one side; stomata with papillae on neighbouring
epidermal and subsidiary cells; ala a fused and without any significant
grooves in between . sess Sen oe . P. tectorius var. littoralis
lb. Leaf apex prickly on hon silos stomata oid panilice on lateral noes cells
only; carpel apices free and separated by deep grooves in between . ¥
56 PhcR ORS PaO cee EO ee oer ee Pp we atissimus
ACKNOWLEDGEMENTS. We thank Dr. Barbara Gravendeel, Marcel Eurlings (Van der
Klauw Laboratory, Leiden University — Netherlands) for laboratory assistance and for sequence
determination. We are grateful to Herbarium Bogoriense (BO) for supplying the materials;
70 Gard. Bull. Singapore 63(1 & 2) 2011
National Herbarium Netherland — Leiden (L) and the Herbarium, Royal Botanic Gardens, Kew
(K) for hospitality provided during work; and Prof. Dr. Mien A Rifai for valuable suggestions
offered. This work was supported by the National University and the Directorate General of
Higher Education of Indonesia through research grants 109/SP2H/PP/DP2M/III/2008 and 028/
SP2H/PP/DP2M/IV/2009, and a Sandwich Programme scholarship provided by the Directorate
General of Higher Education of Indonesia through research grant B18140/Setneg/Setmen/
KTLN/8/2008.
References
Backer, C.A. & Bakhuizen van den Brink Jr., R.C. (1968) Flora of Java Vol. 3.
Groningen: N.V.P. Noordhoff.
Doyle, J.J. & Doyle, J.L. (1987) A rapid DNA isolation procedure for small quantities
of fresh leaf tissue. Phytochem. Bull. 19 (1): 11-15.
Johansen, D.A. (1940) Plant Microtechnique. New York: McGraw-Hill, Book
Company Inc.
Judd, W.S., Campbell, C.S., Kellogg, E.A. & Stevens, P.F. (1999) Plant Systematics, A
Phylogenetic Approach. Massachusetts: Sinauer Associate.
Kam, Y.K. (1971) Comparative systematic foliar anatomy of Malayan Pandanus. Bot.
J. Linn. Soc. 64: 315-353.
Kumar, S., Tamura, K. & Nei, M. (2004) MEGA3: Integrated software for molecular
evolutionary genetics analysis and sequence alignment. Brief. Bioinform. 5: 150-
163:
Pak, J.H. & Kawano, S. (1990) Biosystematic studies on the genus /xoris (Compositae-
Lactucaceae). II. Karyological analysis. Cytologia 55: 553-570.
Rifai, M.A. (1976) Sendi-Sendi Botani Sistematika. Herbarium Bogoriense-LIPI,
Bogor (unpublished).
Stone, B.C. (1967) Studies of Malesian Pandanaceae, I. Polymorphism in Pandanus
odoratissimus L.f. of Asia. Gard. Bull. Singapore 22(2): 231-257.
Stone, B.C. (1972) A review of Javanese Pandanaceae with notes on plants cultivated
in Hortus Bogoriensis. Reinwardtia 8: 309-318.
Stone, B.C. (1979) Revision of the Genus Pandanus Stickman. Part 42. Pandanus
tectorius Parkins. and Pandanus odoratissimus L.f. Pacific Sci. 33(4): 395401.
Stone, B.C. (1994) A note on Pandanus (Pandanaceae) in Taiwan. Bot. Bull. Acad. Sin.
35: 129-132.
Tomlinson, P.B. (1965) A study of stomatal structure in Pandanaceae. Pacific Sci. 19
(1): 38-54.
Vogel, E.F. de (1987) Manual of Herbarium Taxonomy. Theory and Practice, pp. 14—
64. UNESCO. Regional Office for Science and Technology for Southeast Asia,
Indonesia.
Gardens’ Bulletin Singapore 63(1 & 2): 71-76. 2011 71
Bothriochloa (Poaceae: Andropogoneae) in Malesia
A. Sumadijaya' and J.F. Veldkamp
‘Herbarium Bogoriense, Botany Division, Research Center for Biology,
Indonesian Institute of Sciences, Cibinong Science Center (CSC),
Jl. Raya Jakarta-Bogor Km 46, Cibinong, Bogor 16911, Indonesia
alexsumadijaya@ gmail.com (corresponding author)
"Netherlands Centre for Biodiversity — Naturalis,
Section National Herbarium of The Netherlands, Leiden University,
PO Box 9514, Leiden 2300 RA, The Netherlands
veldkamp@nhn leidenuniv.n!
ABSTRACT. In Malesia there are four species of Bothriochloa (Poaceae: Andropogoneae).
Andropogon modesta is lectotypified.
Keywords. Andropogoneae, Bothriochloa, Malesia, Poaceae
Introduction
Bothriochloa Kuntze is a small genus of grasses with about 35 subtropical and tropical
species (Clayton et al., 2008). It belongs to the Andropogoneae, a subtribe that is
especially developed in the tropics. The species are usually found in areas with a
pronounced dry season, where they may become vegetation forming. There is no recent
overall revision, only some local accounts, e.g., Deshpande (1984) for India, Vega
(2000) for South America, Neamsuvan et al. (2009) for Thailand, and Sumadijaya &
Veldkamp (2009) for Malesia.
The generic delimitation is problematic, as B. bladhii (Retz.) S.T. Blake, better
known as B. intermedia (R.Br.) A-Camus or B. glabra (Roxb.) A.Camus, is a most
curious species (see e.g., Wet & Harlan 1970). In general, the plants are hexaploid and
cleistogamous, but occasionally some are outbreeding. The off-spring then is diploid,
tetraploid and hexaploid and can hybridise with species with the same ploidy level,
not only of Bothriochloa, but also with some of Capillipedium Stapf and Dichanthium
Willemet. Thus it is able to transfer genetic information from one genus to another. The
F1 of these is cleistogamous again, and so within its area of distribution from Africa to
Australia and the Pacific there are a great number of local clonal forms, differing more
or less, and causing the description of numerous taxa in all three of these genera. The
synonymy is therefore bewildering.
Some have advocated to join the genera into one, the oldest name being
Dichanthium for it, but then immediately recognise infrageneric taxa in it, one
step down, with no knowledge gained (Wet & Harlan, 1968). It may be noted that
intergeneric hybrids in grasses are just as common as in orchids, so hybridisation is no
argument to join genera.
2 Gard. Bull. Singapore 63(1 & 2) 2011
The diagnostic characters for the taxa are as follows: Bothriochloa with spikes
in a panicle or subdigitate, partial peduncles not capillary, racemes with more than 10,
very slender joints, the pedicels and joints with a longitudinal, translucent, resinous
channel. Capillipedium has panicles with capillary branches, racemes with up to 9
joints, the pedicels and joints have resinous channels. Dichanthium has inflorescences
similar to those of Bothriochloa, but the pedicels and joints have no resinous channel.
Distribution
Note that B. bladhii is the most widespread species which locally may become
dominant, yet as far as herbarium records are concerned it is very rare in Borneo and
apparently introduced in New Guinea. Only three Bornean specimens were seen in L
and SING. Moreover, there is only single specimen collected from Buru island, in
the Moluccas. These facts give these areas a higher priority for collection in the future.
Interesting is the disjunct distribution of B. pertusa. It occurs from Africa [as
B. insculpta (Hochst. ex A.Rich.) A.Camus] to Burma and then in Java, Madura, and
the Lesser Sunda Islands (Flores, Sawu, Sumba, Sumbawa, Timor). Said to have been
introduced, but already in 1858 it was collected by Zollinger in Madura. It may well
be an Ice age relictual distribution. This is a pattern that is often seen in species that
need a more or less seasonal climate with pronounced dry periods. They follow the
former drought tracks from Burma over the Sunda platform, and from Taiwan through
the Philippines to Australia during the last Ice Age, when the sea was perhaps 120 m
lower than today.
A surprising result of our study was that B. ewartianus (Domin) C.E.Hubb.
reported in the herbarium and literature for the Lesser Sunda Isl. (Sumbawa, Timor),
and Papua New Guinea (Central, Madang) could not be distinguished from B.
ischaemum (L.) Keng. For its wide distribution in Australia see the map in Mallett &
Orchard (2002). For that in Eurasia see Conert (1979).
s ‘ = N
i Lie 10, >
Fos e 3 x
1,000 km
f
5
x 1s
¢ | aa 4 oF = Q
Rel, etic
Si a5 ~~
- ra) F AE J}
x a fh a
sod foe {
oe ~ leds 5 gay yee e = acy
RIS See) ea a 5 Le
Fig. 1. Distribution of Bothriochloa Kuntze in Malesia. Range marked with dots is used for
B. bladhii, marked with crosses for B. ischaemum, and marked with dashes for B. modesta;
blackened areas indicate B. pertusa.
Bothriochloa in Malesia 73
The species are easy to distinguish. Important characters are the relative
lengths of the main axis of the inflorescences and of the racemes: main axis longer or
shorter than the lowermost racemes: also whether the lowermost racemes are whorled,
or fascicled, or solitary; whether there are “pits” on the lower glume. It is not clear
what the function of the latter is: it has been suggested that they would be extra-floral
nectaries.
Key to the Malesian species
la. Axis of the panicle much longer than the lowermost branches ...............::0cce0 2
ipeeAxis Of the panicle much shorter than the branches ..........2...:::.2ccc.csseeseccccleeeeecees 3
2a. Axis of the panicle 10—20 cm long, possibly shorter in under-developed specimens.
Racemes whorled, the lowermost often branched, with 8—many joints. Widespread
eee nee RIN iets NESS nO roche vsseiakocde su cusecnesaticebsnseve acest B. bladhii
2b. Axis of the panicle 4-8 cm long. Racemes solitary, simple, with 3—6 joints. (E Java,
eae Aileely AGU ESAT irs ea Scdescn ccs lave cnccisthsvacedededdcvanadead toacdasteeesucnes B. modesta
3a. Nodes usually glabrous. Blades usually glabrous. Upper glume setulose. (Lesser
BarmathalSteSnE Apa INC WAC TUIIEA) \<2sc.c.ccecocsetecnceconcenaccenncnbeceeetservoseets B. ischaemum
3b. Nodes bearded. Blade usually hairy. Upper glume slightly rough. (Lower glume of
sessile spikelet 1—3-pitted.) (Java, Madura, Lesser Sunda Isles) ............ B. pertusa
LECTOTYPIFICATION: Andropogon modesta Backer, the basionym of Bothriochloa
modesta (Backer) Backer & Henrard is lectotypified here with Bewmée 2672 (lecto
BO: isolecto L, PNH7).
An index to Malesian specimens identified as the various species of Bothriochloa is
given at the end of this account.
Excluded species
A single collection of Bothriochloa saccharoides (Sw.) Rydb. from an experimental
garden in Manila was seen in L, with two other collections from that garden (without
exact date and year) in BO. Therefore its presence is excluded from Malesia.
Bothriochloa kwashotensis (Hayata) Ohwi and B. parviflora (R.Br.) Ohw1 var.
mutispicula Ohwi belong to Capillipedium.
ACKNOWLEDGEMENTS. The authors would like to express their gratitude to Wita Wardani
(BO) and Bryan Simon (BRI) for help with literature and pictures.
74 Gard. Bull. Singapore 63(1 & 2) 2011
References
Clayton, W.D., Harman, K.T. & Williamson, H. (2008) GrassBase, the online world
grass flora. http://www.kew.org/data/grasses-db/sppindex.htm (accesed 28 Jan
2008)
Conert, H.J. (1979) Familie Poaceae, in G. Hegi, ///ustrierte Flora von Mitteleureuropa,
ed. 3, I, 3: 18, t. 16. Parey Buchverlag, Berlin.
Deshpande U.R. (1984) Poaceae: tribe Andropogoneae, Dichanthium. Fasc. Fl. Ind.
15: 1-30.
Mallett, K. & Orchard, A.E. (eds) (2002) Flora of Australia 43: 364, map 1287. ABRS/
CSIRO Australia, Melbourne.
Neamsuvan, O., Seelanan, T. & Veldkamp, J.F. (2009) A revision of the genus
- Bothriochloa (Poaceae) in Thailand. Gard. Bull. Singapore 61: 129-143.
Sumadijaya, A. & Veldkamp, J.F. (2009) Notes on Bothriochloa Kuntze (Gramineae:
Andropogoneae) in Malesia. Reinwardtia 12: 415-417.
Vega, A.S. (2000) Revision taxonomica de las especies americanas del género
Bothriochloa (Poaceae: Panicoideae: Andropogoneae). Darwiniana 38: 127-186.
Wet, J.M.J. de & Harlan, J.R. (1968) Taxonomy of Dichanthium section Dichanthium
(Gramineae). Bol. Soc. Argentina Bot. 12: 207-227.
Wet, J.M.J. de & Harlan, J.R. (1970) Bothriochloa intermedia — a taxonomic dilemma.
Taxon 19: 339-340.
Appendix A. Index to the specimens.
bla = Bothriochloa bladhii (Retz.) S.T. Blake
isc = Bothriochloa ischaemum (L.) Keng
mod = Bothriochloa modesta (Backer) Backer & Henrard
per = Bothriochloa pertusa (L.) A. Camus
Afriastini 1664 (BO): bla; 1808 (BO): bla; Alston 15312 (BO): bla; Alvarez & Fernando 4195
(BO): bla; Arendsen Hein 22 (BO): isc; Arsin s.n. (BO): bla; Ass. Resident Timor s.n.
(BO): bla;
Backer Jul 1918 (L): bla; 134 (BO): bla; 6424 (BO): bla; 6630 (BO): bla; 6883 (BO): bla; 6986
(BO): bla; 9542 (BO): bla; 12950 (BO, L): bla; 13375 (BO, L): bla; 13405 (BO): per;
Backer 13496 (BO): bla; 13887 (BO): bla; 1/8089 (BO, L): bla; 203176 (BO): per; 20342 (BO):
bla; 20342 (L): per; 20929 (BO): per; 2/145 (BO): mod; 22807 (BO): bla; 235/2 (BO,
L): bla; 24784 (BO): isc; 24186-bis (BO, L): mod; 2429] (BO): mod; 2432/7 (BO): per;
24402 (BO, L): bla; 24485 (BO): bla; 2453/1 (BO): bla; 246/3 (BO): per; 24882 (BO,
L): mod; 26407 (BO, L): bla; 26558 (BO): bla; 27055 (BO, L): bla; 27735 (BO): bla;
27987 (BO): bla; 28342 (L): per; 28530 (BO): bla; 28770 (BO): bla; 28996 (BO): bla;
29513 (BO): bla; 29670 (BO): bla; 30035 (BO): per; 30054 (BO): mod; 30409 (BO): isc;
30746 (BO): mod; 33579 (BO, L): bla; 33685 (BO): per; 36018 (BO): per; 36019 (L):
per; 36096 (L): bla; 36217 (BO, L): mod; 36925 (BO): per; 37558 (L): per; Bakhuizen
Bothriochloa in Malesia 75
van den Brink 4877 (BO, L): bla; Brink 6755 (BO, L): per: Balansa 1 Dec 1886 (L):
per: Beumée A 674 (BO): per: A 788 (BO): bla; 1073 (BO): mod; 1042 (BO, L): mod:
1498 (BO): bla; 2672 (BO, L): mod; 5396 (BO): bla; 5506 (BO): per; Bloembergen
3207 (BO, L): bla; 3296 (BO, L): bla; 3320 (BO, L): bla; Bouman 4 (L): bla; Brown 8
(SING): bla; BS 4858 (Ramos): (BO): bla; 6734 (Merrill) (SING): bla; 30072 (Fénix):
(BO): bla; 84957 (Ramos & Edano) (SING): bla; Biinnemeijer 1284 (BO, L): bla; 6001
(B), L, SING: bla; 3338 (BO, L): mod: 38/2 (BO): bla; 80// (BO, L): bla:
Carr 11329 (L, SING): bla; CHB (BO): bla: Cinatti 38 (L): bla; 44 (L): bla; 46 (L): isc; 98 (L):
per; Classon K 43 (BO): mod; Clemens 18198 (SING): bla; Cuming 1400 (CGE, G,
GOET, K, L, P,W): bla;
De Voogd 2443 (BO): bla; 2503 (BO): mod; 2530 (BO): bla: De Wilde 22 Jun 1946 (L): bla:
De Wit 4114 (L): bla; Docters van Leeuwen 19 Mar 1911 (L): per; Dorgelo 3012 (L): per; 3021
(L): per;
Edeling 141 (BO): per; Elbert 2885 (L): bla; Elmer 12024 (BO): bla;
Fukuoka & Sukasdi J-1583 (BO): bla:
Gezagh. Sawoe 2 (BO): bla; 4 (BO): per; Gouv. Veearts 4 (BO): per: 18 (L): per;
Gouvern. Veearts Soembawa B. (BO): bla; Gouvern. Veearts te Watampone (BO): bla; Gutterink
3165 (BO): bla;
Hallier M 39 (BO): per: M 41 (BO): per; Henty 22] (L): mod; Hoekstra 19 (BO, L): per; Hoft
2793 (L): bla; 306/ (L): bla; 3072 (L): bla; Holttum 26 Oct 1946 (SING): bla:
Jaag 1200 (L): bla; 1395 (L): bla; 852 (L): bla; Javasuikerindustrie 92 (BO): bla;
Kjellberg 3008 (BO): bla; 3704 (BO): bla; Knaap 15 (BO): bla; Kooy 660 (L): per; 688 (L):
isc: Kuswata 180 (BO): bla;
Lambinon 87/134 (L): bla; Leefmans 94 (BO): per; Lérzing 3732 (BO): bla; 58/7 (BO): bla;
8069 (BO): bla: 88/] (BO): bla: 9033 (BO): bla; 9/02 (BO, L): bla; 1/010 (BO): bla:
11101 (BO): bla; 12885 (BO): bla:
Malvins 31 Jan 1886 (SING): bla: Mehra 4 (BO): bla: 6 (BO): bla; // (BO): bla; 5/7 (BO):
bla; 84 (BO): bla; 97 (BO): bla; /// (BO): bla; Mehra & Dadi Supriadi 1173 (BO):
bla; Metzner 90 (L): per; Monod de Froideville 995 (BO, L): bla: 1028 (BO): bla; 12/8
(BO): bla; 1243 (BO): per: 1376 (BO): per; 1444 (BO, L): bla; 1498 (BO, L): bla; 1515
(BO): bla; 15/5a (BO): bla; 1598 (BO, L): bla; 1627 (L): per; 1721 c (BO): per; 1760
(BO): bla: 7878 (BO, L): bla; 7882 (BO): bla: 1/884 (BO): bla; 1886 (BO): bla; 1893
(BO): bla; 1982 (BO, L): per; 1996 (BO, L): per; 1999 (BO): per; 2029 (BO): per; 2041
(BO): ise;
NGF 20987 (Henty) (L): bla; 22058 (Gillison) (L): bla; 39411 (Streimann & Kairo) (BO): bla
49815 (Henty & Katik) (L): bla:
76 Gard. Bull. Singapore 63(1 & 2) 2011
PNH 11401 (Farinas & Abordo) (L): isc; 17024 (Sulit) (L): bla; 19065 (Conklin) (L): bla;
20464 (Mendoza) (L): bla; 82029 (Mendoza) (L): bla; Posthumus 2652 (BO): bla;
Proppe 22 (BO): bla; 23 (BO): bla; 33 (BO): per; Pullen 3142 (L): isc; 6748 (L): bla;
Ramos 1842 (BO, SING): bla; Reid May 1956 (SING): bla; Ridley 14 (SING): bla; 11689
(SING): bla; 14844 (SING): bla:
Saakov 45 (BO): bla; SAN 151251 (Laegaard et al.) (L): bla; 151306 (Laegaard et al.) (L):
bla; Santos J.V. 4607 (L): bla; 5095 (L): bla; 5104 (L): ise; 5240 (L): bla; 5826 (L):
bla; 6089 (L): bla; 6280 (L): bla; 6365 (L): bla; 6673 (L): bla; 6763 (L): bla; 6883 (L):
bla; 6906 (L): bla; 6950 (L): bla; 7373 (L): bla; 7323 (L): bla; 7364 (L): bla; 7387 (L):
bla; 7422 (L): bla; 7426 (L): bla; 7578 (L): bla; 7587 (L):-bla; 7648 (L): bla; 7718 (L):
bla; 78/5 (L): bla; 87/6 (L): bla; 8253 (L): bla; Schmutz 4969 (L): bla; 5047 (L): bla;
5933 (L): bla; SF 7284 (Nur) (SING): bla; 7771 (Ridley) (SING): bla; 13301 (Burkill
& Haniff) (SING): bla; 22899 (Henderson) (SING): bla; Simon 4228 (L): per; Sinclair
8879 (SING): bla; Siwon 1026 (L): bla; Sohns 4 (BO): bla; 1/0 (BO): per; 7/8 (BO): bla;
Sumadijaya & Fanani 5 (BO): bla; Sunarti & Hamzah PTU 34 (BO): bla;
Van Balgooy 5094 (BO, L): bla; Van Borssum Waalkes 3157 (BO): bla; Van der Meer & De
Hoed 2099 (L): per; Van Harreveld s.n. (BO): per; Van Leeuwen JEF 6 (L): bla; TSIOF
2 (L): bla; Van Ooststroom 12645 (L): per; Van Slooten 2065 (BO): per; Van Steenis
6669 (BO): bla; 7514a (BO, L): bla; 7763 (BO): mod; 7/229 (BO): bla; 17(8)73 (BO):
per; 17450 (BO): per; 17450 (mixed with Zoysia matrella) (L): per; 17473 (L): per;
17978 (L): per; 18062 (BO, L): bla; 18084 (BO, L): bla; 18476 (L): bla; Veearts Sibolga
20 (BO): bla; Veeartsenijk. Dienst 40 (BO): bla; Veldkamp 6955 (L): mod; Veldkamp
7165 (BO, L): bla; 8905 (BO, L): bla; 8967 (BO, L): bla; Verheijen 3148 (L): per; 5277
(L): bla; Volkens 196 (SING): bla;
Walsh 13 (BO): bla; 42 (BO): bla; Widjaja 4690 (BO): bla; Widjaja & Hamzah 2950 (BO): bla;
Wisse 429 (BO): isc; 683 (BO): isc;
Yakob 6 (SING): bla:
Zollinger 3960 (Madura, 6 Jun 1858!) (L): per.
Gardens’ Bulletin Singapore 63(1 & 2): 77-82. 2011 4
Saurauia (Actinidiaceae) of New Guinea:
current status, future plans
Marie Briggs
Herbarium, Library, Art and Archives,
Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3AE, U.K.
m.briggs@kew.org
ABSTRACT. Saurauia, with approximately 300 species, is the largest of three genera within
the family Actinidiaceae and is found in the tropical and sub-tropical regions of Asia, Central
and South America. The family placement of the genus has changed several times, at times
being placed in Ternstroemiaceae, Dilleniaceae and its own family, Saurauiaceae. The island
of New Guinea may be a centre of diversity for Saurauia in South East Asia with more than
50 species. No comprehensive treatment of New Guinean Saurauia has been attempted since
the work of Diels in 1922, despite complaints by later researchers that this publication is out of
date and the subdivisions of the genus proposed therein are unsatisfactory. A full account of the
family, including Saurauia, has yet to be covered in Flora Malesiana. This paper presents an
introduction to the genus Saurauia in New Guinea and communicates plans for future research.
Keywords. Actinidiaceae, New Guinea, Saurauia
The family Actinidiaceae
The family Actinidiaceae Gilg & Werdermann contains c. 355 species within three
genera—A ctinidia Lindl. (which includes the kiwi-fruit, c. 30 species), Saurauia Willd.
(c. 300 species) and Clematoclethra (Franch.) Maxim. (c. 25 species). The family
occurs in tropical and subtropical Central America, South America and South East
Asia and also in temperate Asia and northern Australia (Heywood 2007). According to
the Angiosperm Phylogeny Group (APG) 3 (Stevens 2001 onwards), Actinidiaceae sits
in the order Ericales as a sister group to the families Roridulaceae and Sarraceniaceae.
Actinidiaceae is a family of trees, shrubs and woody lianas with alternate or
spiral, simple leaves with entire or serrate margins and no stipules. Inflorescences are
axillary, with few to many unisexual or hermaphroditic flowers with free or fused petals,
10 to many stamens which may be fused to the base of the petals and a superior ovary,
usually with three to five carpels (Heywood 2007). Flowers are usually pentamerous
but exceptions occur. In pentamerous flowers the aestivation is quincunical (Dressler
& Bayer 2004). Flowers are often white but can also be pink, red or yellowish brown.
The fruit is a berry or (often in Saurauia) a capsule, usually containing many small
seeds. Raphides occur in many plant parts (Dressler & Bayer 2004). Several species
(of mainly Actinidia) are cultivated world-wide for their edible fruit and ornamental
value.
78 Gard. Bull. Singapore 63(1 & 2) 2011
Sap
The genus Saurauia
Saurauia, by far the largest genus in Actinidiaceae, occurs from Mexico southwards to
Chile in the New World and then in the Old World from China to New Guinea (with
one species in Queensland, northern Australia). They are conspicuously absent from
Brazil and Africa. There are c. 60 species across Central and South America (Hunter
1966, Soejarto 1980), the remainder being found in the Old World.
Saurauia 1s a genus of small to medium trees and shrubs (sometimes
scrambling). A prominent feature of many species of Saurauia is the distinctive
indumentum of (often stiff) hairs and scales that cover many plant parts. In dried
herbarium specimens, the indumentum is often chaffy and tan-coloured, making it
easy to spot with a 10 hand-lens, and sometimes with the naked eye. The flowers of
Saurauia may be subtended by a bract and two bracteoles (Dickison 1972). Sepals are
connate at the very base, filaments are adnate to petals at the base; filament bases may
be connate, forming a ring. Anther dehiscence is via poricidal slits.
Saurauia was first described by Carl Ludwig von Willdenow in 1801 and
placed by him in the family Tiliaceae (Willdenow 1801). Two different spellings were
used in this original publication—Saurauia on the description of the illustration plate
and Saurauja in the main body of text—most likely due to a printing error (Hoogland
1977). Both were used for over 170 years until Hoogland’s 1977 proposal, that
Saurauia be conserved over Saurauja, was accepted.
Between Willdenow’s 1801 publication and the present day Saurauia has
been assigned by different authors to a number of different families—most commonly
Ternstroemiaceae (now Pentaphylacaceae), Dilleniaceae and Saurautaceae (Fig. 1). In
1972 Dickison provided evidence based on detailed studies of floral morphology and
anatomy that Saurauia belonged in the family Actinidiaceae.
Saurauia in South East Asia and current status in New Guinea
Saurauia has yet to be treated for the Flora Malesiana series (Actinidia 1s the only
genus in the family Actinidiaceae to have been covered so far; Steenis 1950). A
revision of the Peninsular Malaysia taxa is currently in progress (Rafidah Rahman,
pers. comm.) which will be the first for the Flora Malesiana region since Diels’s (now
outdated) treatment of the New Guinea taxa in 1922. Diels’s classification, looking
only at New Guinea species, divided the genus into 10 sections: Uniflorae, Ramiflorae,
Calyptratae, Squamulosae, Setosae, Armatae, Obtectae, Rufae (also inconsistently
referred to in the manuscript as Zomentosae by Diels), Bibracteatae and Obvallatae.
The sections are divided mainly on inflorescence architecture, bract characters, and
leaf, bract and sepal indumentum.
Gilg & Werdermann (1925) later adapted Diels’s system to encompass other
Old World species (no extra New Guinean taxa were added) and modified it by sinking
seven of Diels’s 10 sections into one, thus dividing the genus into four sections:
Uniflorae, Ramiflorae, Calyptratae and Pleianthae. The seven sections of Diels sunk
Saurauia in New Guinea 79
into Pleianthae were retained as series.
These systems are now desperately outdated. as many more New Guinean
species have been published (e.g., Diels 1929, Smith 1941, Kanehira & Hatusima 1943,
Royen 1982, Takeuchi 2008) and authors have experienced difficulties in fitting new
taxa into the generic framework laid down in the Diels (1922) and Gilg & Werdermann
(1925) papers. Burtt (1936) noted that under Gilg & Werdermann’s system, several
closely related Old World species fell into different sections, leading him to believe
that the classification was largely artificial. The same has been noted specifically in
New Guinea taxa, in Saurauia trugul P.Royen, for example. Royen (1982) noted that
the pubescence of the bracts and sepals placed the species in series Bibracteatae, but
the species showed a closer affinity to Obvallatae. More recent publications (e.g.,
Takeuchi 2008) continue to complain about this lack of clarity. Diels, Gilg and
Werdermann were however aware of the limitations of their work, with Diels (1922)
noting that “the series suggested are intended only as a preliminary grouping” and Gilg
& Werdermann (1925) stating: “we believe that future monographic work will change
and improve the groupings”.
A search of Index Kewensis records, via the International Plant Names Index
(2010) reveals that c. 100 taxon names have been published for New Guinea. The first
two New Guinea taxa were published by Miquel in 1869. During the 20 years from
1922 to 1941, over 50 species were described (in Diels’s 1922 paper he almost doubled
the number of species known from New Guinea by publishing 24 species that were
new to science). In contrast. less than 10 New Guinea species have been published
in the more than 60 years since. Very little mention is made regarding synonymy in
any of the papers on New Guinea taxa and it is possible that some of the 100 taxon
names will be sunk as future research finds them to be conspecific. Royen (1982), for
example, estimates there to be c. 50 species in New Guinea. Saurauia appears to be
Fig. 1. A small selection of different family names under which Saurauia has been published.
80 Gard. Bull. Singapore 63(1 & 2) 2011
particularly species-rich in New Guinea when compared with other areas: there are
6 species in Vietnam, 13 in China, 10 in Peninsular Malaysia (and the whole of the
New World only has c. 60 species in total) (Cuong et al. 2007, Li et al. 2006, Rafidah
Rahman, pers. comm. 2010, Hunter 1966 and Soejarto 1980, respectively). Many New
Guinean Saurauia species appear to favour disturbed or partially disturbed habitats—
the edges of clearings and trails for example. New Guinea ts a very geologically active
area, where frequent earthquake disturbances and landslips may provide a plethora of
suitable habitats and may go some way to explaining the level of species diversity seen
there.
Fig. 2 gives an overview of the number of taxon names published from New
Guinea by year and, when considered alongside the fact that there is a substantial
amount of indeterminate Saurauia material from New Guinea in herbaria around the
world, suggests that the Sauwrauia of this region have been somewhat neglected in
recent years. The reasons for this are perhaps manifold—the revisions are outdated—a
significant number of species have been described since they were published and these
new taxa are not included in the keys. There may also be a significant number of new
species yet to be recognised as such. Smith (1941) noted that only a few Saurauia
species (usually those occurring in lowland rain forest) were abundant, with most
taxa from altitudes higher than 1200 m occurring in very narrow ranges. Many of the
indeterminate specimens have been collected from high altitude areas previously not
researched botanically.
Saurauia in New Guinea have long since been considered a ‘difficult’ group
to work with due to the lack of a solid and taxonomically sound generic framework
around which new taxa can be added. There is no definite understanding of which
characters are taxonomically significant. Recent studies on New World taxa (by
Soejarto in 1980), however, are promising and may hold the key to bringing a sense
of order to the New Guinea taxa. The distribution of indumentum on sepals and also
the type of indumentum on different plant parts has been found to be taxonomically
significant. This may also be true of the New Guinea taxa where a wide variety of
No. Spp. Described
| coat all | ;
© ©
9 6 8 © © & 5D 6S SPD OS 6 DOS SHH DS 2
ECE ELE EF & FEF SEP HP LPF SPF oF FPF LFS SF SH Ss FSF Fs
Year of Description
Fig. 2. The number of Saurauia species described for New Guinea each year, the first by
Miquel in 1869, the last by Takeuchi in 2008.
Saurauia in New Guinea 81
indumentum is evident and should be a consideration of any future research on the
group in New Guinea.
The long reaching aim of future work on Saurauia taxa from New Guinea
will be to look for taxonomically significant characters that can be used to identify
species, to write a comprehensive key to the species and to get an overview of how
these species relate to others in the Flora Malesiana region. No molecular studies have
been published to date and such work may yield useful information when looked at
in parallel with morphological studies. There 1s evidently much to learn about the
genus Saurauia in New Guinea and it is hoped that, with careful research and access
to herbarium specimens, DNA samples, field notes and photographs, definite progress
towards these aims can be made.
ACKNOWLEDGEMENTS. Thanks to Tim Utteridge, Rogier de Kok, Melanie Thomas, Eve
Lucas and Frances Crawford from the Herbarium at RBG, Kew for useful discussions and
valuable feedback regarding this paper; to Craig Brough from the Library at RBG, Kew and the
library staff at the Natural History Museum, London for sourcing Willdenow’s 1801 paper; to
Monika Shaffer-Fehre at RBG, Kew for the translation of Diels’s 1922 paper; and finally to the
Flora Malesiana organising committee and TOBU Foundation for giving me the opportunity to
present this paper at the Flora Malesiana conference.
References
Cuong, N.M., Soejarto, D.D. & Li, J. (2007) A taxonomic revision of Actinidiaceae of
Vietnam. Blumea 52(2): 209-243.
Dickison, W.C. (1972) Observations on the floral morphology of some species of
Saurauia, Actinidia, and Clematoclethra. J. Elisha Mitchell Sci. Soc. 88: 43-54.
Diels, L. (1922) Die Dilleniaceen von Papuasien. Bot. Jahrb. Syst. 57: 436-459.
Diels, L. (1929) Descriptions of new species collected in British Papua by L.J. Brass.
J. Arnold Arbor. 10: 75-84.
Dressler, S. & Bayer, C. (2004) Actinidiaceae. In: Kubitzki, K. (ed) The Families and
Genera of Vascular Plants, vol. 6, pp. 14-19. Berlin: Springer.
Gilg, E. & Werdermann, E. (1925) Actinidiaceae. In: Engler, H.G.A. & Prantl, K. (eds)
Die Natiirlichen Pflanzenfamilien (ed. 2) 21: 36-47.
Heywood, V.H., Brummitt, R.K., Culham, A. & Seberg, O. (2007) Flowering Plant
Families of the World. Kew: Royal Botanic Gardens, Kew.
Hoogland, R.D., Schultes, R.E., Soejarto, D.D. (1977) Proposal for the conservation
of the generic name Saurauia Willdenow (1801) against Saurauja Willdenow
(1801). Taxon 26(1): 147-148.
Hunter, G.E. (1966) Revision of Mexican and Central American Saurauia
(Dilleniaceae). Ann. Missouri Bot. Gard. 53(1): 47-89.
International Plant Names Index (2010) http://www.ipni.org (accessed 18 Aug 2010).
82 Gard. Bull. Singapore 63(1 & 2) 2011
Kanehira, R. & Hatusima, S. (1943) The Kanehira-Hatusima 1940 collection of New
Guinea plants. XVII. Bot. Mag. (Tokyo) 57: 63-85.
Li, X.-W., Li, J.-Q. & Soejarto, D.D. (2006) New combinations in Actinidiaceae from
China. Novon 16: 362-363.
Miquel, F.A.G. (1869) Adnotationes des Ternstroemiaceis. Ann. Mus. Bot. Lugduno-
Batavi 4: 103-114.
Royen, P. van (1982) Saurauiaceae. Taxonomic part Winteraceae to Polygonaceae.
The Alpine Flora of New Guinea 3: 1367-1396.
Smith, A.C. (1941) Studies of Papuasian Plants, IV. J. Arnold Arbor. 22: 497-528.
Soejarto, D.D. (1980) Revision of South American Saurauia (Actinidiaceae).
Fieldiana, Bot. N.s. 2: 1-141.
Steenis, C.G.G.J. van (1948) Actinidiaceae s. str. In: Steenis, C.G.G.J. van (ed) Flora
’ Malesiana Series 1, 4, pp. 37-39.
Stevens, P.F. (2001 onwards) Angiosperm Phylogeny Website. Version 9. June 2008
[and more or less continuously updated since]. http://www.mobot.org/MOBOT/
research/APweb/ (accessed 18 Aug 2010).
Takeuchi, W. (2008) Saurauia taylorii (Actinidiaceae), a distinctive new species from
the Kaijende Highlands of Papua New Guinea. Blumea 53: 335-340.
Willdenow, C.L. von (1801) Drei neue Pfanzen-Gattungen. Saurauia. Ges. Naturf.
Freunde Berlin Neue Schriften 3: 407-408.
Gardens’ Bulletin Singapore 63(1 & 2): 83-96. 2011 83
Searching for Sumatran Begonia described by
William Jack:
following in the footsteps of a 19th century
Scottish botanist
Mark Hughes! and Deden Girmansyah?
'Royal Botanic Garden Edinburgh, 20a Inverleith Row, Edinburgh, EH3 5LR, U.K.
m.hughes@rbge.ac.uk (corresponding author)
"Herbarium Bogoriense, Botany Division, Research Center for Biology,
Indonesian Institute of Sciences, Cibinong Science Center (CSC),
Jl. Raya Jakarta-Bogor Km 46, Cibinong, Bogor 16911, Indonesia
ABSTRACT. Eight species of Begonia were described from Sumatra in 1822 by the Scottish
botanist William Jack. All of the type material associated with these names was destroyed in
a fire in 1824, and an expedition was mounted in August 2010 to re-visit Jack’s collecting
localities in an effort to find material suitable for neotypification. Of the eight species, two
(Begonia bracteata Jack and B. racemosa Jack) could be neotypified with certainty, whilst
others require further work. It is possible that some of the species described from Bengkulu
province may have become extinct due to loss of forest habitat.
Keywords. Begonia, Sumatra, William Jack
Introduction
William Jack (1795-1822) was the son of the Principal of Aberdeen University.
He is celebrated as “one of the most able botanists ever to become associated with
the incredibly rich and the then very little known flora of the Malay Peninsula and
Archipelago” (Merrill 1952). According to Don (1834), his “well known indefatigable
labours in natural history have long ago entitled him to the highest respect” and tribute
has been paid to “the astonishing accuracy of [his] work and his descriptive powers”
(Noltie 2009). Two tragedies throw these accolades into relief; Jack’s untimely death
at the age of 27, and the destruction of the bulk of his biological collections by fire.
After finishing his medical training in London, Jack left England for India
in 1813 to work for the British East India Company as a surgeon. Whilst in India he
commenced correspondence with Nathanial Wallich in Calcutta, who recommended
Jack to Sir Stamford Raffles as a suitable appointment to his staff as both a medical
man and botanist. This appointment led Jack to Sumatra, where the British East India
Company was hoping to strengthen its influence, and where Jack would be able to
make his name as one of the most prolific and brilliant botanists of his time. He made
collections from the island during 1819-1822, visiting North Sumatra (Tapanoeli),
84 Gard. Bull. Singapore 63(1 & 2) 2011
West Sumatra (Pulau Pegang and neighbours, and the Nias Islands) and Bengkulu
(Gunung Bungkuk and the interior of the province). Jack published a fascinating
account of his 1821 ascent of Gunung Bungkuk (“Gunong Benko” or “Sugar Loaf
Mountain”; Jack 1822a), where locals pleaded with Jack and his party not to climb
the mountain, as they feared the vengeance of evil spirits if they achieved the sacred
summit. However, he was undaunted:
“The next acclivity terminated at the head of another ravine, where their progress was again
checked by a jutting rock rendered moist by the trickling of a small spring of water from
among its crevices. Here the guides declared that further ascent was impracticable, and that
from thence the party might return as soon as they pleased. The fact is, they were extremely
averse to their proceeding, fearing the vengeance of the evil spirits if they conducted
strangers to the summit; they were, therefore, advising to return at every difficulty, and the
ascent was ultimately accomplished without their aid, or rather in spite of them.”
And reached the summit in spite of the very difficult terrain:
“The last of these precipices was perhaps the most dizzy and dangerous, as it was necessary
to make a step or two on a narrow ledge on the face of a cliff of such height that the
eye could not discern the bottom, and thence catch at a dry stump barely within reach,
by swinging from which it was possible with a considerable effort to clear the rock. The
denseness of the moss and the stunted appearance of the trees now indicated their approach
towards the top, and at length about two o’clock they found themselves on the summit. This
was a bare spot of not more than four or five yards in breadth with a precipice on each side
partly concealed by brushwood. Of those who set out together from the foot of the hill a
few only reached this point, by far the majority giving up in despair at different parts of the
ascent, but the labour of those who persevered was amply recompensed by the view which
opened from the summit.”
Within a year, two of the three “mountain defilers” (Noltie 2009), Captain
Harry Auber and Jack himself, were dead. Jack contracted malaria during an excursion
to Moco-Moco in March 1822, which in combination with the consumption caught
during his time in Nepal proved more than his health could bear. His condition became
so grave that he was placed on a passage to England aboard the Layton; however the
departure was delayed to bad weather and due his rapidly deteriorating condition,
Jack was moved to Government House in Bengkulu, where he died shortly after. As
if to further avenge the trespass, in 1824 all of Jack’s wonderful collections were lost.
Raffles had them loaded on the Fame along with other irreplaceable manuscripts and
drawings and set sail from Bengkulu on the 2nd of February 1824. Allegedly the ship
caught fire after a sailor tried to illicitly tap a brandy cask by candle light; all onboard
were saved but the entirety of the collections were lost, an enormous tragedy for
Malesian botany second only to the death of Jack himself.
In his short career, Jack managed to describe about 200 species and 31 genera
of plants. His account of the Begonia he collected on Sumatra was published in 1822
in Malayan Miscellanies (Jack 1822b):
“The island of Sumatra abounds with Begoniae, a tribe of plants which are chiefly found
in moist shady situations at the foot of hills and in the recesses of forests. Being succulent
herbs they are with difficulty preserved in herbaria, and the specimens are frequently
deficient in one or other of the parts of fructification. Descriptions from the living plants in
their native soil are therefore particularly desirable, and in this view the following account
Sumatran Begonia described by William Jack 85
of the species which have fallen under my observation will not be uninteresting.”
The loss of Jack’s collections, of which only a few scraps reached Europe
prior to the Fame disaster, means that vast majority of his plant names have no type
material, and this is certainly true for his Begonia names. Although Jack had a deserved
reputation for writing excellent descriptions, the variation between Begonia species on
Sumatra can be subtle, and the absence of any comparative diagnoses in Jack’s account
means neotypes are a necessity to fully understand the application of his names. To this
end, the authors undertook an expedition during August 2010 to re-visit the localities
where Jack collected Begonia specimens, in an attempt to find material suitable for
neotypification. The expedition was based, firstly, in Padang, then Bengkulu (Fig. 1).
This permitted collecting on Pulau Pasumpahan and Gunung Bungkuk, and in the
rapidly diminishing remaining forest scattered throughout Bengkulu province. The
trip to Gunung Bungkuk was navigated using Jack’s 1822 manuscript and we used
the same village as a base, Rejak Bessi. The dense forests which obscured Jack’s
view of the mountain have now been cleared (Fig. 2), allowing us to reach the base
of the mountain in less than one hour with the aid of motorcycles with tyre chains.
It was possible to reach the mountain and return to Bengkulu in the same day, in
contrast to Jack’s expedition which lasted eight days in all. Another contrast to Jack’s
expedition is that a full ascent was not attempted, due to failing light and a desire by
the participants not to suffer the same fate as their 19th century predecessors. The fact
that the expedition took place during the Muslim fasting month was another, minor,
consideration, although by the time of the descent the elusive Begonia bracteata plus
another new species had been found.
Taxonomic treatment
Sectional placement
Following examination of descriptions, herbarium specimens and living plants,
insights have been gained to allow a more informed sectional classification of
Sumatran Begonia, leading on from the excellent groundwork laid down by Doorenbos
et al. (1998). Begonia caespitosa, B. orbiculata and B. sublobata are transferred from
Begonia sect. Diploclinium to Begonia sect. Reichenheimea; the former section is no
longer represented on the island and the latter probably represents a local radiation,
possibly including some species from Peninsular Malaysia such as B. forbesii and
allies. Begonia sect. Reichenheimea is currently united by the presence of entire
placentae and, on Sumatra at least, a functionally scapigerous habit. The Sumatran
species and their allies may eventually require a section of their own to accommodate
them, depending on further study of the type of the section, B. tenera Dryand. from
Sri Lanka. Jack’s suggestion of a resemblance of Begonia sublobata to B. grandis is
most definitely not supported. Begonia fasciculata and B. pilosa are transferred from
Begonia sect. Petermannia to Begonia sect. Bracteibegonia, due to the presence of red
hairs on the leaves which is so characteristic of the section on Sumatra. Begonia sect.
Bracteibegonia seems to be much more species rich on Sumatra than Begonia sect.
86 Gard. Bull. Singapore 63(1 & 2) 2011
95°E 100°E 105°E
0 50100 200 Km
¢
Tapanoeli,.
_Pulau Pasumpahan
Pulau Pegang
"Batang Aia Manjuto
lami maida
0255 10 Ken
gounung Bungkuk
gbengkulu
ee
o 1 20 40 Kem
95°E 100°E 105°E
Fig. 1. Map of Sumatra showing localities mentioned in the text.
Fig. 2. The isolated peak of Gunung Bungkuk, approximately 30 km northeast of Bengkulu
town.
Sumatran Begonia described by William Jack 87
Petermannia, whichis characterised by being glabrous and having larger inflorescences.
Begonia racemosa and a B. geniculata remain in Begonia sect. Petermannia.
Species descriptions
Jack’s original descriptions and diagnoses (Jack 1822b) are reproduced here, arranged
first by current sectional placement and then alphabetically. Following each is a
discussion of the success or failure to find a neotype, and where possible, an account
of comparative diagnostic characters to related taxa.
Begonia bracteata Jack [§ Bracteibegonia|, Malayan Misc. 2(7): 13 (1822); Miquel,
Pl. Jungh. 417 (*1855°, 1857); Candolle, Prodr. 15(1): 316 (1864); Golding, Phytologia
54(7): 494 (1984). Diploclinium bracteatum (Jack) Migq., Fl. Ned. Ind. 1(1): 688 (1856).
TYPE: Sumatra, Bengkulu Province, Gunung Bungkuk, 3°35°3”S 102°25°24”E, 610
m, 15 Aug 2010, D. Girmansyah & M. Hughes DEDEN 1495 (neotype here designated
BO; isoneotypes ANDA, E, K, SING). (Fig. 3)
Foliis duplicato-serratis acuminatis pilosis, pedunculo I—3 floro bracteis numerosis
appressis vestito, capsulis basi bibracteatis; alis equalibus rotundatis.
Near the foot of Gunong Bunko in the interior of Bencoolen.
Suberect, strong and branching, very villous, shaggy. Leaves alternate, short petioled,
ovate, semicordate at the base, acuminate, duplicato-serrate, pilose, 34 inches long.
Stipules large, pilose. Peduncles oppositifolious, generally supported by a smaller leaf,
invested particularly towards the base with many pair of opposite ovate acute pilose
ciliate bracts, which are pressed flat against each other; the uppermost pair is distant
from the rest and supports from one to three pedicels. Flowers white. Male. Corolla
four petalled; the outer two large subrotund. Stamina numerous. Female. Corolla
five petalled; petals nearly equal. Styles three. Stigmata lunate, villous with yellow
short glandular hairs. Capsule embraced by two bracts at the base, three celled, three
winged; wings equal, rounded.
Notes. Neotypification was very straightforward for this species, due to the precise
locality and this being the only Begonia sect. Bracteibegonia in the vicinity, in addition
of course to matching the clear description perfectly. This species has been considered
a synonym of Begonia lepida Blume from Java (Koorders 1912), but is a much hairier
plant; Jack’s description as ‘shaggy’ is very apt; it differs in having much longer, erect,
translucent hairs on the stem (not reddish appressed hairs), hairier stipules, noticeably
bullate leaves and also lacks any red colouration on the stems, leaves and young tepals.
It is similar in habit and leaf shape to an undescribed species from Aceh (Wilkie et al.
PW621]a, Gunung Leuser National Park, Ketambe Research Station, BO, E, SING),
which differs in having appressed, matted, red hairs rather than erect, translucent hairs
on the stem; differences between this taxon and Begonia lepida (considered endemic
to Java) need to be clarified.
88 Gard. Bull. Singapore 63(1 & 2) 2011
Za
Fig. 3. Begonia bracteata Jack. Main image, habit; top right, female flower; middle right,
female flower after pollination; bottom right, female flower with developing fruit; bottom left,
opening male flower. All from DEDEN/ 495, Gunung Bungkuk, Bengkulu.
IUCN category. In the absence of collections from elsewhere, it 1s assumed that this
species is endemic to Gunung Bungkuk and nearby forests. As the foothills have been
cleared of forest for coffee plantations up to the base of the mountain, Begonia bracteata
will have undergone a significant contraction in range since Jack’s expedition, and is
now effectively endemic to the slopes of the mountain itself. The steepness of the
terrain will afford some protection, but as the forests on Gunung Bungkuk do not
belong to a formally designated protected area, B. bracteata should be considered as
Vulnerable (VUD2).
Begonia fasciculata Jack § [Bracteibegonia], Malayan Misc. 2(7): 12 (1822); Candolle,
Prodr. 15(1): 322 (1864). Petermannia fasciculata (Jack) Klotzsch, Monatsber. Kon.
Preuss. Akad. Wiss. Berlin 1854: 124 (1854); Klotzsch, Abh. Kon. Akad. Wiss. Berlin
1854: 195 (1855); Klotzsch, Begoniac. 75 (1855). Diploclinium fasciculatum (Jack)
Miq., Fl. Ned. Ind. 1(1): 690 (1856). TYPE: Sumatra, Tappanuly, Jack (destroyed).
Foliis inferioribus alternis, stiperioribus oppositis, oblongo-ovatis basi semicordatis
duplicato-serratis pilosis, perianthiis masculis diphyllis, capsulae alis equalibus
obtusangulis.
Sumatran Begonia described by William Jack 89
Found at Tappanuly on the west coast of Sumatra.
Caulescent. Stem weak, jointed, thickened at the joints, round, covered with red hairs.
Leaves petiolate, the lower ones alternate, the upper ones opposite, oblong-ovate,
inequilateral, semicordate at the base, acuminate, irregularly serrate, covered above
with red erect subspinescent hairs, beneath with softer and weaker hairs. Petioles
densely pilose. Stipules linear, acuminate, pilose. The flowers come in fascicles
from the middle of the petioles, and these flower bearing leaves are always opposed
to another without flowers; hence it is that the upper leaves are opposite while the
lower are alternate. Fascicles composed of male and female flowers; pedicels slender,
smooth, white. Bracts several at the base of the blades, acute, pilose, red. Male perianth
diphyllous, white. Stamina numerous. Anthers yellow. Female perianth superior,
white, cup-shaped five leaved; petals ovate, acute, with a few short red hairs on the
outside. Style deeply trifid lobes convolute, infundibuliform. Capsule three-winged,
three-celled, wings equal, obtuse-angled.
Notes. The specimen Argent & Igbar 9968 (E) from Tapak Tuan 250 km northwest of
Tapanoeli 1s the closest match to this species, but does not have leaves “covered above
with red erect subspinescent hairs”. There are hardly any collections from the vicinity
of Tapanoeli and the location was not on the itinerary for the 2010 expedition; the area
needs to be further explored to search for suitable material.
IUCN category. Data Deficient.
Begonia pilosa Jack {§ Bracteibegonia| Malayan Misc. 2(7): 13 (1822). Diploclinium
pilosum (Jack) Mig. Fl. Ned. Ind. 1(1): 688 (1856). TYPE: Sumatra, Bengkulu
Province, Jack (destroyed).
Foliis subsessilibus irreguliter serratis acumiatis pilosis subtus rubris, bracteis
ad basin pedicellorum subrotundis ciliatis, capsulae alis subaequalibus parallelo
rotundatis.
Interior of Bencoolen.
Caulescent, pilose. Leaves alternate, scarcely petiolate, ovate, inequilateral, acuminate,
slightly and irregularly serrate, pilose with long red hairs, under surface of a bright red
colour; about three inches long. Stipules large, lanceolate, pilose externally. Peduncles
oppositifolious, subidichotomous. Bracts at the base of the pedicels, roundish, ciliate.
Flowers white. Male: Corolla four petalled, the inner pair smaller. Stamina numerous.
Female: Corolla five petalled; the two outer petals larger. Capsule three winged; wings
nearly equal, parallel and rounded.
Notes. In addition to Begonia bracteata, two other members of Begonia sect.
Bracteibegonia were encountered on the expedition, represented by D. Girmansyah
90) Gard. Bull. Singapore 63(1 & 2) 2011
& M. Hughes DEDEN1493 & DEDEN1507, both collected in Bengkulu. Neither
have large stipules, one has leaves which are glabrous above and the other has leaves
which are pilose but considerable less than three inches long, and hence neither can
be ascribed to Begonia pilosa Jack. Examination of other specimens of this section in
ANDA and BO also fails to provide a convincing match.
IUCN category. Data Deficient.
Begonia geniculata Jack [§ Petermannia] Malayan Misc. 1(7): 15 (1822); Candolle,
Prodr. 15(1): 321 (1864). Petermannia geniculata (Jack) Klotzsch, Monatsber. Kon.
Preuss. Akad. Wiss. Berlin 1854: 124 (1854); Klotzsch, Abh. Kon. Akad. Wiss. Berlin
1854: 196 (1855); Klotzsch, Begoniac. 76 (1855). TYPE: Sumatra, Jack (destroyed).
Caule geniculato, foliis ovato-oblongis denticulatis acuminatis glabris, pedunculis
divaricato dichotomis, floribus superioribus masculis dipetalis, inferioribus femineis,
capsulae alis equalibus obtus angulis.
Sumatra.
Caulescent; stems smooth, compressed, channelled, jointed, thickened at the
articulations. Leaves alternate, petiolate, semicordate at the base, ovate oblong,
acuminate, denticulate smooth. Peduncles oppositifolious, dichotomous, divaricate,
many flowered, lower flowers female, upper male. There is often a female flower from
the axil. Male perianth two petalled, white. Stamina numerous; anthers oblong, broader
above. Female. Capsules long, three winged, wings obtuse-angled, equal, smooth.
Observations by Jack. The leaves of this plant are used by the natives for cleaning
and taking out rust from the blades of creeses. It has considerable resemblance to the
preceding species [Begonia racemosa].
Notes. The description of Begonia geniculata is uncharacteristically short for Jack,
and it reads like a brief description of Begonia sect. Petermannia generally. Combined
with the lack of a specific locality, it means this species will be near impossible to
neotypify. It is feasible that its resemblance to Begonia racemosa, typified below, may
be of some use in sorting out its true identity once Begonia sect. Petermannia becomes
better known on Sumatra.
IUCN category. Data Deficient.
Begonia racemosa Jack [§ Petermannia] Malayan Misc. 2(7): 14 (1822); Candolle,
Prodr. 15(1): 322 (1864). Petermannia racemosa (Jack) Klotzsch, Monatsber. Kon.
Preuss. Akad. Wiss. Berlin 1854: 124 (1854); Klotzsch, Abh. Kon. Akad. Wiss. Berlin
1854: 196 (1855); Klotzsch, Begoniac. 76 (1855). Diploclinium racemosum (Jack)
Miq., Fl. Ned. Ind. 1(1): 691 (1856). TYPE: Sumatra, Bengkulu Province, Bukit
Menyan, 3°36°26”S 102°39°39”E, 1110 m, 19 Aug 2010, D. Girmansyah & M. Hughes
Sumatran Begonia described by William Jack 91
DEDEN1509 (neotype here designated BO: isoneotypes ANDA, E, K, SING). (Fig. 4)
Foliis obovato oblongis irregulariter dentatis acuminatis glabris, racemis erectis
masculis, flore femineo axillari, perianthiis masculis diphyllis, capsulae alis equalibus
parallelo-rotundatis.
Interior of Bencoolen.
Suberect; stem smooth, jointed. Leaves alternate. short petioled, obovate oblong,
attenuated towards the base which is unequally cordate, acuminate, irregularly
and unequally dentate, smooth; 6—7 inches long. Stipules large, oblong. Racemes
oppositifolious, long, erect, bearing numerous fasciculate male flowers, and having a
single female one in the axil. Male. Corolla two petalled. petals very thick. Stamina
numerous. Female. Capsule with three equal parallel rounded wings, three celled.
Specimens: Sumatra. Bengkulu, Kaba, 1 Mar 1931, C_N_A. de Voogd 1053 (BO, L):
Bengkulu, Kaba, 10 Mar 1932, C_N.A. de Voogd 1325 (BO, L): Bengkulu, Sungei
Gembung, 100 m, 12 Oct 1993, JJ. Afriastini 2620 (BO): Bengkulu, Sungei Gembung,
100 m, 12 Oct 1993, JJ. Afriastini 2625 (BO); Bengkulu, Road from Kapahiang,
3°39°47°S 102°33°46"E, 660 m, 17 Aug 2010, D. Girmansyah & M. Hughes
DEDEN1498 (ANDA, BO, E).
Notes. Matching this name to collections was initially confounded by Jack’s description
of the inflorescences as oppositifolius, as in Begonia sect. Petermannia they are
usually terminal. as was the case in all the specimens of this section collected during
the expedition. However, one can easily interpret this terminal inflorescence syndrome
as oppositifolius when considering a highly branched specimen (e.g., Fig. 4). This
species was observed at three localities during the 2010 expedition (Bukit Menyan,
Kapahiang and Bukit Kaba), and it seems likely therefore that Jack would also have
happened upon it. They key characters are the leaf shape, the long inflorescences, the
thick tepals of the male flower and the capsule with rounded wings, all of which match
between the description and specimens D. Girmansyah & M. Hughes DEDEN1498
& DEDEN!509. Strictly speaking the inflorescence is not racemose, but an elongated
cyme. A further distinctive but previously unknown feature of B. racemosa is that the
female flowers have three tepals, an unusual feature in Begonia generally.
IUCN category. Of the known populations of this species, only one is confirmed as
being in a protected area (Bukit Kaba). One of the localities, Bukit Menyan, is not
under protection and the already small forest fragment is under active encroachment
from coffee plantations. However, it is extremely likely that Begonia racemosa is
present in the extensive nearby Bukit Hitam protection forest, and as long as Bukit
Kaba and Bukit Hitam remain in good condition the IUCN category Least Concern is
considered appropriate.
92 Gard. Bull. Singapore 63(1 & 2) 2011
b3i ibe
Fig. 4. Begonia racemosa Jack. Main image, habit; top left, male flower; middle left, female
flower, all from DEDENI509, Bukit Menyan, Bengkulu. Bottom left, male portion of
inflorescence, Gunung Kaba, Bengkulu.
Begonia caespitosa Jack |§ Reichenheimea| Malayan Misc. 2(7): 8 (1822); Candolle,
Prodr. 15(1): 397 (1864). Diploclinium caespitosa Miq., Fl. Ned. Ind. 1(1): 685 (1856).
TYPE: Sumatra, Bengkulu, Jack (destroyed).
Subacaulis, foliis inequaliter cordatis angulatis acuininatis glabris, pedunculis
dichotome cymosis, capsulae alis equalibus obtusangulis v. rotundatis.
At Bencoolen.
Nearly stemless. Leaves petiolate, oblique, cordate at the base with rounded slightly
unequal lobes overlapping each other a little, somewhat falcate, rounded and sublobate
on one side, straighter on the other, attenuated into a long acumen or point, spinulose
but scarcely serrated on the magin, smooth, shining above, pale and punctato papillose
beneath; nerves 5—9, branched towards the margin. The leaves are of unequal size
and vary somewhat in shape, the old ones being much rounder and more decidedly
lobed than the younger ones, which have the point so much incurved as to be nearly
falcate on one side. Petioles red, pilose. Peduncles often as long as the leaves, smooth,
bearing a dichotomous cyme of white flowers. Bracts ovate, concave. Male perianth
four leaved, the inner pair smaller. Stamina numerous, collected into a head. Female
Sumatran Begonia described by William Jack 93
perianth superior, three leaved, two exterior large, subrotund, applied to each other as
in the male flowers, and enclosing the third which is much smaller and oblong. Style
trifid. Stimata lunato bifid, yellow and glanduloso-pilose. Capsule three winged, wings
nearly equal, obtuse angled or rounded.
Notes. Four species in Begonia sect. Reichenheimea are represented by specimens
from Bengkulu. Species A, with long petioles and large flowers (e.g., D. Girmansyah
& M. Hughes DEDEN1508); species B, smaller with slightly lobed leaves (e.g., D.
Girmansyah & M. Hughes DEDEN1506); species C, with peltate leaves (e.g., de Voogd
1055); and species D, with leaves mottled green and brown (e.g., D. Girmansyah &
M. Hughes DEDEN1496). None of these match Begonia caespitosa perfectly; species
B does have leaves which vary in the degree of lobing according to their age, but no
evidence of falcate leaves was observed. The lack of size measurements in Jack’s
descriptions means none of the species can ruled out on what would otherwise have
been a very simple character. It is possible, given its location (“at Bencoolen’) that
this is a lowland species which is no longer extant given the almost complete lack of
indigenous forest cover on the vicinity of the coastal lowlands.
IUCN category. Data Deficient.
Begonia orbiculata Jack [§ Reichenheimea| Malayan Misc. 2(7): 9 (1822); Candolle,
Prodr. 15(1): 398 (1864). Diploclinium orbiculatum (Jack) Miq., Fl. Ned. Ind. 1(1):
688 (1856). TYPE: Sumatra, Bengkulu, Jack (destroyed).
Subacaulis, foliis orbiculatis cordatis crenatis glabris, pedunculis subdichotomis,
capsulae alis subequalibus obtusangulis.
Interior of Bencoolen.
Nearly stemless. Leaves petiolate, subrotund, from three to four inches in diameter,
slightly oblique, cordate at the base where the lobes overlap each other, remotely
crenate, rounded at the point, smooth except for the nerves of the under surface,
beautifully and finely punctate above. Stipules scariose, acute. Peduncles erect,
subdichotomous, nearly as long as the leaves, i.e. about six or eight inches in height.
Flowers white. Male. Corolla four petalled, the outer pair large, oblong; the inner
small. Stamina numerous. Female. Capsule three celled, many seeded, three winged;
wings obtuse-angled, nearly equal.
Notes. As the inflorescences and flowers of many species of Begonia sect. Reichenheimea
on Sumatra are extremely similar, we are left with only Jack’s description of the leaf as
diagnostic. Species D, collected from the base of Gunung Bunkuk, matches in having
orbiculate crenate leaves, but differs in being distinctly bullate between the veins
and being variegated green and purplish brown, rather than “beautifully and finely
punctate”. The bullate leaves on species D are very obvious, and it seems likely that
94 Gard. Bull. Singapore 63(1 & 2) 2011
gay
Jack would have noticed this: hence we assume there is no material available for the
typification of Begonia orbiculata.
IUCN category. Data Deficient.
Begonia sublobata Jack [§ Reichenheimea| Malayan Misc. 2(7): 10 (1822); Candolle,
Prodr. 15(1): 355 (1864). Diploclinium sublobatum (Jack) Mig., Fl. Ned. Ind. 1(1):
690 (1856). TYPE: Sumatra, West Sumatra, Pulau Pegang, Jack (destroyed).
Repens, foliis cordatis subquinquelobis vel angulatis dentato serratis margine reflexis
glabris, capsulae alis equalibus obtusangulis.
Found under moist rocks on Pulo Pegang, West coast of Sumatra.
Repent with a thick knotty root. Leaves alternate, petiolate, cordate, sometimes
unequally, large and broad, often six or seven inches long, angulate, sometimes with
five acute lobes, sometimes nearly ovate, acuminate, dentato-serrate, edges recurved,
very smooth, 5—7 nerved, finely punctate, the dots appearing elevated on the upper
surface and depressed on the lower. Petioles 4-6 inches long, nearly smooth, furnished
immediately below their junction with the leaf with a semiverticil of linear acute
appendices or scales. Stipules large, ovate, rather laciniate towards the apex, one on
each side the petiole. Peduncles axillary, erect, 6—8 inches long, red, very smooth,
terminated by a dichotomous divaricate panicle of white flowers tinged with red.
Bracts roundish. Male. Perianth four leaved, leaflets rather thick and fleshy, the two
outer ones much larger and subrotund, before expansion completely enclosing the
inner two, and having their edges mutually applied to each other in such a manner
that they form an acute carina round the unexpanded flower. Stamina, numerous in
a roundish head; filaments short, inserted on a central column which rises from the
base of the flower. Anthers oblong, cells adnate to the sides of the filaments, bursting
longitudinally. Female. Capsules with three equal obtusely angled wings, three celled,
three valved, valves septiferous in the middle, sutures corresponding to the wings.
Seeds numerous, attached to placentae which project from the inner angle of the cells.
Observations by Jack. The serratures are hard and cartilaginous and recurved in such
a manner along with the margin of the leaf, that when only observed on the upper
surface their place is perceived by an indentation. It seems to resemble the B. grandis
Dryand which differs however in having oblique, doubly serrated leaves, and purple
flowers.
Notes. A photograph of a plant with red scales at the petiole apex, a unique character in
Sumatran Begonia and diagnostic for Begonia sublobata, led us to Pulau Pasumpahan
(Nurainas, pers. comm.), 5 km from the type locality of Pulau Pegang (Fig. 1). About 20
m back from the beach the (limestone?) rocky core of the island proved to be covered
on one side with a Begonia species which appeared to match Begonia sublobata Jack
ah
, s
Fig. 5. Begonia sp. aff. sublobata Jack. Main image, habit; inset, top left, apex of petiole and
leaf underside; top right, male flower: bottom ght, female flowers and ovaries; bottom left,
juvenile plant showing leaves with reduced lobing. All from DEDEN 1/486, Pulau Pasumpahan,
West Sumatra.
in every detail (Fig. 5). This is one of Jack’s more lengthy and detailed descriptions,
and every character seemed to tally perfectly; the lobed leaf shape, the cartilaginous
serratures on the leaf margin, the semiverticil of scales at the petiole apex, the fleshy
outer tepals on the male flower, each of which is diagnostic enough when considered
singly. However, in the excitement of the moment of discovery, the authors failed
to notice that the collection from Pulau Pasumphan did not have leaves which were
“finely punctate, the dots appearing elevated on the upper surface and depressed
on the lower’; they were completely smooth. However by the time this was noted
several days had passed and there was no time to visit Pulau Pegang to investigate
further. Another species, as yet unnamed and manifestly allied to Begonia sublobata
given the red scales at the petiole apex, was discovered nearby on the mainland at
Batang Aia Manjuto. Hence it seems possible that the ancestor of Begonia sublobata
has fragmented locally into a number of taxa, and that the islands in the bays south
of Padang may harbour a number of endemic species. Consequently further work is
required to resolve the true identity of Begonia sublobata.
IUCN category. Data Deficient.
96 Gard. Bull. Singapore 63(1 & 2) 2011
ACKNOWLEDGEMENTS. This research was facilitated by the Indonesian Ministry of
Research and Technology (RISTEK), the Indonesian Institute of Sciences (LIPI), Direktorat
Jenderal Perlindungan Hutan dan Konservasi Alam (DITJEN PHKA) and the Scottish
Government’s Rural and Environment Research and Analysis Directorate. Nurainas and Roki
from Universitas Andalas and Mr. Wahyudi from Universitas Bengkulu are thanked for their
assistance in the field.
References
Don, G. (1834) A General System of Gardening and Botany III. Rivington, London.
Doorenbos, J., Sosef, M.S.M. & Wilde, J.J.F.E. de (1998) The sections of Begonia
‘including descriptions, keys and species lists (Studies in Begoniaceae VI).
Wageningen Agric. Univ. Pap. 98(2): 1-266.
Jack, W. (1822a) Memorandum of a journey to the summit of Gunong Benko, or the
Sugar Loaf Mountain in the interior of Bencoolen. Malayan Miscellanies 2: 1-11.
Jack, W. (1822b) Descriptions of Malayan plants. Malayan Miscellanies 2(7): 1-96.
Koorders, S.H. (1912) Exkursionsflora von Java 2: 640-653.
Merrill, E.D. (1952) William Jack’s genera and species of Malaysian plants. J. Arnold
Arbor. 33(3): 199-251.
Noltie, H.J. (2009) Raffles’ Ark Redrawn: Natural History Drawings from the
Collection of Sir Thomas Stamford Raffles. London & Edinburgh: British Library
& RBGE, in association with Bernard Quaritch Ltd.
Gardens’ Bulletin Singapore 63(1 & 2): 97-103. 2011 o7
Materials for a revision of Erycibe (Convolvulaceae)
in Peninsular Malaysia
S. Syahida-Emiza'*, G. Staples? and N.W. Haron?’
‘Forest Research Institute Malaysia (FRIM), 52109 Kepong, Selangor, Malaysia
syahida@frim.gov.my
*Singapore Botanic Gardens, | Cluny Road, Singapore 259569
‘Institute of Biological Sciences, Faculty of Science,
University of Malaya, 50603 Kuala Lumpur, Malaysia
ABSTRACT. Information from the literature, new observations based on field study, and new
distribution data gathered from herbarium specimens and new collections are assembled in
preparation for a revision of the genus Erycibe in Peninsular Malaysia. Significant new data are
discussed and a conservation status is assigned to each of the 19 taxa recognised in Peninsular
Malaysia. Problems still to be resolved are highlighted.
Keywords. Convolvulaceae, Erycibe, Peninsular Malaysia
Introduction
Erycibe includes about 75 species distributed mainly in tropical Asia and Malesia with
outlying species in Australia, Japan and Taiwan (Staples 2010). The genus Erycibe
was first described by Roxburgh (1798), based on E. paniculata Roxb. from India.
The taxonomic framework for understanding the genus was established by Hoogland
(1953a) who provided a nomenclatural review of all taxa described at that time (70
accepted species). Detailed descriptions, keys for identification, selected illustrations,
ecological information and biogeographical distribution summaries were provided for
the 53 Malesian species (Hoogland 1953b). Later, an index of all Erycibe specimens
examined was prepared (Hoogland 1961) that remains useful for naming older
herbarium specimens and is invaluable for understanding the taxonomic concepts
Hoogland employed. Subsequent to Hoogland’s work another five Asian species have
been described although not all are accepted.
Hoogland made a thorough revision based solely on herbarium material and his
keys and descriptions rely heavily on reproductive (floral) characters. He was the first
to use trichome characters, particularly the hairs from the calyx and the midpetaline
bands, to distinguish the species in certain groups. Yet, while the keys and descriptions
work well enough in the herbarium, they are not practical for field identification. Using
the hair characters requires a compound microscope with an optical micrometer for
measuring accurately; many botanists and collectors have lamented the lack of good
field characters for recognising Erycibe. This is problematic especially when trying to
key out sterile or fruiting materials.
98 Gard. Bull. Singapore 63(1 & 2) 2011
In the account of Ervcibe in Peninsular Malaysia, Ridley (1923) recognised
15 species, documented in the Flora of the Malay Peninsula; some of these were later
reduced to synonymy by Hoogland (1953a, 1953b). Hoogland accepted 16 species
and 2 varieties in Peninsular Malaysia. Of these, four taxa were considered endemic
(E. magnifica, E. praecipua ssp. praecipua, E. sapotacea and E. strigosa). Later, Ng
(1989) recognised two more species that he named ‘species A’ and ‘species B’ based
on leaf characters. Until now, the two latter species have not been described due to
incomplete material. In total, Peninsular Malaysia has 19 recognised taxa.
Almost six decades after Hoogland’s revision for Flora Malesiana, there is no
updated taxonomic revision for the genus Erycibe in Peninsular Malaysia. The Flora
of Peninsular Malaysia Project now provides an impetus to re-examine the genus and
synthesise new information. Today, there are more collections of Erycibe available,
which makes it possible to review taxonomic concepts for the recognised species
and the two new taxa recognised by Ng. The full revision of the genus in Peninsular
Malaysia, with a new key, detailed descriptions, distribution maps for each species
and colour photographs will be published in the Flora of Peninsular Malaysia account.
The purpose of the present paper is to bring together new information gathered from
the literature, from field observations of living plants, and from study of the herbarium
specimens that have accumulated since the 1950s. Furthermore, the conservation status
for each species has been assessed based on the Malaysia Plant Red List Guideline
(Chua & Saw 2006).
Materials and methods
Field study and specimen collection
Nine field trips were carried out from January 2009 till April 2010 at known localities
as well as in new collection areas, while specialised trips were carried out to relocate
rare species to obtain fresh materials. Materials for flowers and fruits were preserved
in the spirit collection in addition to voucher specimens. Further information of the
habitat and habit characters based on personal observations made in the field was
added. In addition, close-up colour photographs were taken, especially of flower and
fruit parts, as an aid in distinguishing the species.
Comparative morphology based on herbarium specimens
This study was conducted on herbarium specimens from the following herbaria: BKF,
K, KEP, KLU, L, SING and UKMB. A total of 586 specimens of Erycibe collected
from Borneo, Singapore, Sumatra and Thailand were borrowed and compared with
specimens collected from Peninsular Malaysia. Of these, 241 collections of Ervcibe
collected from Peninsular Malaysia were examined.
Scanning Electron Microscope (SEM) studies
The structures of the floral parts, especially trichomes on the mid-petaline bands
and calyx, have been observed by Hoogland (1953b) to be of taxonomic value for
Erycibe in Peninsular Malaysia 99
distinguishing the species. However, during that time, Hoogland observed this character
through light microscopy and no figures or plates were provided in his account to
illustrate this character. Today, Scanning Electron Microscope (SEM) offers a powerful
technique for observation of trichome characters and making precise measurements.
In this study, 15 species were studied using Scanning Electron Microscope (FEI
Quantum 200) using herbarium specimens or fresh materials.
Conservation status of Erycibe
The distribution of the Peninsular Malaysian Erycibe species has not been mapped,
so their conservation status is unknown, particularly for the endemic species. The
conservation status assessment of each taxon is being carried out based on the
guidelines and criteria of the Malaysia Plant Red List (Chua & Saw 2006). The final
result for all 19 taxa is currently in preparation.
Results and discussion
Distribution of Erycibe
Based on the data from recently collected material together with that on herbarium
specimen labels, all Ervcibe species are found and distributed in lowland to hill forest,
ranging from 20 m to 1200 m a.s.]. From the field work conducted, only four of the
19 taxa, namely, E. albida, E. sapotacea, E. stapfiana and E. rheedii were found and
studied in the forest at the base of Gua Wang Buluh and Temurun Waterfall (Kedah
state), Penang Hill (Penang), base of Gunung Korbu and Bubu Forest Reserve
(Perak), Forest Research Institute Malaysia (Selangor), Pasoh Forest Reserve (Negeri
Sembilan), Gunung Belumut (Johor) and Tembat Forest Reserve (Terengganu)
(Fig. 1).
Erycibe albida was found flowering at Temurun Waterfall (Kedah), Pasoh
Forest Reserve (Negeri Sembilan) and Tembat Forest Reserve (Terengganu). All
collections were made in lowland areas. However, no fruits were obtained.
Erycibe sapotacea was again found on Penang Hill (type locality). It has been
recorded as endemic to Peninsular Malaysia in the past. Unfortunately, no flowers were
obtained (December 2009). However, specimen W.J.J.O. de Wilde & B.E.E. de Wilde-
Duyfjes 21199, 29 July 1981, from Sumatra (deposited in the Leiden herbarium) looks
similar to E. sapotacea in fruit and leaf characters. For the time being, FE. sapotacea
is considered as an endemic to Peninsular Malaysia. However, further study is needed
and perhaps this species has a wider distribution extending to Sumatra.
Erycibe stapfiana was observed fiowering in April at the lower trail to Gunung
Korbu (Perak). The flowering season is about 2—3 weeks only. It is a climber, reaching
30-35 m tall in the forest canopy. We found this climber on a hillside near a Saraca
stream, which is relatively undisturbed with quite an open forest canopy.
Erycibe rheedii was found to be quite common in Pulau Tuba (Kedah) near
Gua Wang Buluh (a limestone cave). It occurs along the trail to the cave’s base.
Similar to other species, E. rheedii also favours gaps where sunlight is available. From
100 Gard. Bull. Singapore 63(1 & 2) 2011
0 100 a 200 Kilometers x
| vO ~ oe
Fig. 1. Location of field collecting trips for Ervcibe spp.
our observations, it only flowers once a year (early March) and the fruiting season is
towards the end of the month.
Based on the data gathered, the distribution of the two undescribed species is
now known. Erycibe sp. A is believed to be endemic to Gunung Belumut, Johor, and
Erycibe sp. B is endemic to lowland forest of Pahang, Selangor and Negeri Sembilan.
Morphological observations
Generally, Eryvcibe species are small shrubs, woody climbers or lianas, climbing by
twining high in the forest canopy. Plants are always found on forest margins, in forest
gaps and sometimes near roadsides. In Peninsular Malaysia, only E. a/bida has been
recorded consistently as a shrub. From the observations made in the field, the outer
bark is normally light or pale grey, sometimes with lenticels or low longitudinal ridges
and sometimes very smooth when the climbers become huge. However, characters
such as plant height, bark texture and bark colours are not good taxonomic characters
for identification because they are related to age of the climbers. Nevertheless, these
characters are able to provide supplementary evidence for field identification.
Erycibe in Peninsular Malaysia 101
As for the leaves, there are a few characters that are quite useful for the
identification of Peninsular Malaysia species. The size and shape of the lamina in E.
leucoxyloides is very distinct, oval-elliptic to lanceolate, 1.1—3.9 cm long and 0.5—
1.3 cm wide. In the examination of herbarium specimens, some of the species such
as E. magnifica have very clear venation underneath with thick pubescence. These
characters are very consistent and can be useful to distinguish the species.
Erycibe has two types of inflorescence: racemose/paniculate or glomerulate
at either a terminal or axillary position. Many Erycibe species have very light sweet-
scented flowers like jasmine, although there is a species recorded with a strong odour:
E. rheedii. The flower of Erycibe is either white or creamy in colour. Erycibe has a
deeply 5-lobed corolla, with each lobe having a bilobed apex and very dense hairy
outside on the mid-petaline bands. The filament is either triangular or laterally concave.
A few species have truncate anther and many have acute anther apices.
The fruit is a berry, with a little flesh surrounding the single seed, seated on the
persistent calyx. Generally, the shape of the fruits is ovoid or ellipsoid or sometimes
obpyriform. In Hoogland’s revision for Flora Malesiana, important characters such as
colour of the fruits was unknown for some species. For example, a recent collection
made from Penang Hill added another important character for E. sapotacea: this
species has pale grey fruits, which was not mentioned in Hoogland’s account.
Micromorphological observations
For first time SEM technique is used for the micromorphological study on trichome
structure of the midpeline bands and calyx. In this study, two main hair types were
found on the midpetaline bands: two-branched hairs and three- to many-branched
hairs (stellate hairs). Two-branched hairs are found in E. festiva and E. maingavi.
Three- to many-branched hairs are found in other species and also in FE. maingayi.
For the calyx, a glabrous calyx surface is found only in E. a/bida, while two-branched
hairs are found in E. festiva, E. griffithii and E. maingayi and three- to many-branched
hairs (stellate hairs) are found in other species. Thus, the results from the SEM images
support Hoogland’s findings in 1953.
An examination of the mid-petaline and calyx hairs of the specimen Sidek bin
Kiah SK513, 19 February 1976, from Kuala Dipang Forest Reserve, Perak (with very
typical Erycibe leaf shape) shows it belongs to E. festiva, which has two- branched
hairs. Initially this specimen had been wrongly identified as E. griffithii by the collector
and later identified by an unknown person (in 2008) as E. maingayi.
However, from the analysis made, not all species can be distinguished by the
trichome type. Trichome type is an additional character useful to distinguish a few
species only.
Ecology and life history
In the study of plant dispersal by Ridley (1930), birds are reported as seed dispersers of
Erycibe tomentosa var. tomentosa (synonym E. princei) and E. malaccensis. Erycibe
tomentosa var. tomentosa produces large panicles of drupes; the fruits do not all ripen
at once. The contrast of bright-orange unripe fruits (very conspicuous) with dark
102 Gard. Bull. Singapore 63(1 & 2) 2011
gay
red or almost black ripe fruits is attractive to frugivorous birds. Besides that, there
are anecdotal records by some biologists observing hornbills eating Erycibe fruits.
However, recent books on hornbill biology such as Kinnaird & O’Brien (2007) and
Poonswad (1998) do not list Erycibe (or any Convolvulaceae) among the food plants
eaten by hornbills. Besides birds, the fruits of Erycibe are also eaten by mammals and
McConkey & Galetti (1999) reported the sun bear (Helarctos malayanus) eating E.
maingayi fruits and dispersing the seeds in Central Kalimantan, Indonesia. . The bear’s
droppings contained establishing E. maingayi seedlings found around 150 m from the
adult liana.
Presently unresolved problems
After almost 60 years, the number of new herbarium collections for Erycibe in
Peninsular Malaysia has not greatly increased. There were less than 100 collections
of Erycibe collected after Hoogland’s time. Several species, for example E. strigosa,
has very little information known; the only collection is still only the type specimen
collected in 1886 from Taiping, Perak, with very limited locality information.
Therefore, further collecting would be important. Until now, incomplete material in
herbarium specimens (fruits and flowers) prevents the description and formal naming
of the two taxa recognised by Ng (1989). Besides, the locality data as stated on the
specimen label are insufficient, thus it is hard to relocate the plant.
Even with new collections available, there are still major gaps to be filled,
especially in life history data. The study of ecology, phenology, pollination, seed
predation, herbivory, or seedling establishment is still lacking. Pollen grains have also
still to be studied.
ACKNOWLEDGEMENTS. We are very grateful to the Ministry of Natural Resources and
Environment Malaysia (NRE) for the scholarship given; the Flora of Peninsular Malaysia
Project funded by the Ministry of Science, Technology and Innovation (MOSTI) through the
National Council for Scientific Research and Development (MPKSN), under Project No. 01-
04-01-0000 Khas 2 entitled ‘Safeguarding the Forest Plant Diversity of Peninsular Malaysia’;
and the University of Malaya, Kuala Lumpur for Grant No. PS172 2008B. We wish to express
our sincere thanks to Dr. Ruth Kiew for her help and comments on the manuscript, Dr. Richard
Chung for his help and Dr. Lilian Chua for help with species conservation assessment. Thanks
are due to the Directors or Curators of the herbaria BKF, K, KEP, KLU, L, SAN, SING and
UKMB for allowing us to borrow specimens.
References
Chua, L.S.L. & Saw, L.G. (2006) Malaysia Plant Red List, Guide for Contributors.
Kepong: Forest Research Institute Malaysia.
Hoogland, R.D. (1953a) A review of the genus Erycibe Roxb. Blumea 7: 342-359.
Ervcibe in Peninsular Malaysia 103
Hoogland, R.D. (1953b) Ervcibe. In: Ooststroom, S.J. van, Convolvulaceae; Steenis,
C.G.G.J. van (ed) Flora Malesiana ser. 1, 4: 404-431.
Hoogland, R.D. (1961) /dentification Lists of Malaysian specimens. 13. Erycibe
(Convolvulaceae), pp. 165-171. Foundation Flora Malesiana.
Kinnaird, M. & O’Brien, T.G. (2007) The Ecology & Conservation of Asian Hornbills.
Farmers of the Forest. Chicago: University of Chicago Press.
McConkey, K. & Galetti, M. (1999) Seed dispersal by the sun bear Helarctos malayanus
in Central Borneo. J. Trop. Ecol. 15: 237-241.
Ng, F.S.P. (ed) (1989) Tree Flora of Malaya, vol. 4. Malayan Forest Records No. 26.
Petaling Jaya: Longman Malaysia Sdn. Bhd.
Poonswad, P. (ed) (1998) The Asian Hornbills: Ecology and Conservation. Thai
Studies in Biodiversity no. 2.
Ridley, H.N. (1923) Convolvulaceae. Flora of the Malay Peninsula 2: 443-463.
Ridley, H.N. (1930) The Dispersal of Plants Throughout the World. Ashford: L. Reeve.
Roxburgh, W. (1802) [t.p. 1798]. Plants of the Coast of Coromandel 2: 31-32, plate
159.
Staples, G. (2010) Convolvulaceae. Flora of Thailand 10(3): 330-468.
— ee
vee h es an
i aoaoteten @
bas
Gardens’ Bulletin Singapore 63(1 & 2): 105-118. 2011 105
Updating Malesian Icacinaceae
T.M.A. Utteridge' and M. Schori?
'SE Asia and Pacific Regional Team,
Herbarium, Library, Art and Archives,
Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3AE, U.K.
t.utteridge@kew.org (corresponding author)
Jodrell Laboratory,
Royal Botanic Gardens, Kew, Richmond, Surrey, TW9 3AE, U.K.
schori@ohio.edu
ABSTRACT. The Icacinaceae were traditionally considered difficult to recognise because of
extremely diverse vegetative anatomy and an enormous range in structure. Using a traditional
circumscription of the family, the Icacinaceae of Asia were revised by Sleumer in 1969 and
published in the Flora Malesiana in 1971, and included 100 species in 21 genera. Since
the publication of the FM account, a new understanding of relationships within the group,
stimulated by molecular phylogenetic data, has resulted in these genera being assigned to
several different, more morphologically homogeneous families. In addition, an increase in
collections has allowed species-level taxonomy to be revised in several groups, resulting in
new species from the region, as well as a new genus from Borneo. In this paper these changes
are reviewed, with a discussion of useful characters for identification, and an updated list of
families, genera and species presented.
Keywords. Cardiopteridaceae, checklist, classification, Icacinaceae, Malesia, Stemonuraceae
Introduction
The Icacinaceae as historically circumscribed are a pan-tropical family of trees, shrubs
and lianas with c. 50-55 genera and 300-400 species worldwide (Mabberley 1997,
Karehed 2001). However, recent studies have shown that the family as traditionally
circumscribed was polyphyletic (Savolainen et al. 2000, Soltis et al. 2000, Karehed
2002); this had resulted in a morphologically heterogeneous group that was difficult to
recognise in the field and herbarium (e.g., van Balgooy 1998). Genera were traditionally
placed in the family if they possessed a superior, unilocular ovary with two pendulous
ovules, of which only one matures. In addition simple, exstipulate, alternate leaves,
free petals with valvate aestivation, free stamens, and drupes with a single seed were
used to identify members of the Icacinaceae. The family has now been split into five
families residing in three different orders (see Karehed 2001 and Stevens 2001 for an
updated family classification).
The family was revised for the Flora Malesiana by Hermann Otto Sleumer
(1971a), and our knowledge of the group in Malesia is based on this solid foundation.
106 Gard. Bull. Singapore 63(1 & 2) 2011
Sleumer was best known for his work on Ericaceae and Flacourtiaceae, especially
the former which he revised for the Flora Malesiana (published in 1966). His work
on the Icacinaceae was something that he commenced in 1942 but, when compared
to his treatments of Ericaceae, was considered a sideline, e.g., it has been noted that
Sleumer’s “taxonomic studies have included Ericales, Proteaceae, Flacourtiaceae, and
various minor groups” (Doleshy 1966).
In addition to the new family limits, several new taxa have been described
recently, including a new genus and a complete revision of the taxonomically complex
genus Gomphandra, and these changes are summarised here with an updated key and
checklist to the families in the Malesian region. We have kept the key to genera in
Sleumer’s original format as we note that many herbaria have yet to re-arrange their
collections to anew APG III system and the collections will still be kept together in the
traditional Icacinaceous circumscription. (We also feel that some botanists may still
have a mental concept of Icacinaceae being those plants that are neither Euphorbiaceae,
Flacourtiaceae nor anything else!) Distributions are given for each species following
the TDWG distribution scheme (Brummitt 2001).
Key to the genera of
Cardiopteridaceae, Icacinaceae s.str., and Stemonuraceae in Malesia
The following key is adapted from Sleumer (1971la) with the new families in
brackets after each genus, and with the addition of Sleumeria, Cardiopteris, and a
brief distribution statement for each genus (distribution statements are given for each
species below).
la, Trees‘or (mot climbing) /shiubs 22.ecscresee cece eres eee ete ee eee p
1b. Climbing shrubs or lianas, sometimes with tendrils. Flowers unisexual or
fumCtionally: $0) ...2.5.00csevosdissesbsaun voadee eee hee eee she oe Rae eee ee 16
2a. Sepals essentially free at least in their upper 3/4 and imbricate ................. eee 3
2b. Sepals connate into a cup-like calyx, its upper free part (or lobes), if any, short and
FeC8 1010010) 9 Cr: | eR ERs nce arth. Se casera ci Ser ens aotecrgoanstiotnogeus: 6
3a. Flowers bisexual oi.c6c:.cc.2.0t.cssttihe ete eet eee meee cere Re eee 4
3b. Flowers amise xual cicc.5 cece Ra eae ae a ee ee 5
Malaysia) :..s02s..sde.cssoonssssaethetsvesstnedadeussooarece aed Glranena||Cardioptendaceae]
4b. Petals connate below into a tube, their upper part free and valvate in bud. New
GUINEA. cessscdivssasetianch Sara ee ee 4. Pseudobotrys |Cardiopteridaceae |
Sa. Filaments free, fixed to the very base of the petals. Leaves with a layer of rounded
to star-shaped appressed scales underneath at least in the young state. Malesia ....
sion bg Bde bussiness isda ermal to are ae eae ee 11. Platea {Icacinaceae s.str. ]
Malesian Icacinaceae updated 107
5b.
6a.
6b.
Ta.
7b.
8a.
8b.
9a.
Ob.
10a.
10b.
lla.
11b.
12b.
13a.
13b.
14a.
Filaments adnate for almost their entire length to the lower tubular part of the petals.
SEES SET FS 3. Gonocaryum [Cardiopteridaceae]
ances uamisexual (or functionally so) (2. 2...2.2...20 Kens ed este a
DERE DURES 722 anc oad ae nee ce 11
Drupe ovoid-ellipsoid or oblongoid, without a fleshy lateral appendage. Flowers in
rather short cymes. Malesia ..........................- 19. Gomphandra {|Stemonuraceae]
Banc iaictally conipressed (almond-like) -.:......:.....-<.:-.-..<----c0---0c2-0nsdecseseeeenees 8
Drupe without a fleshy lateral appendage. Flowers in spikes (very rarely in panicles
composed of spikes, or almost fascicled). New Guinea and the Moluccas ............
nas 22 uodligc tos oes ae alae eee 14. Rhyticaryum {\cacinaceae s.str. |
Drupe with a thick fleshy, laterally borne, practically entirely adnate appendage.
SDDS SEE SPSS SE Be ee a ey Cr 9
Disk unilateral, thick-squamular. New Guinea ..... 20. Hartleya [Stemonuraceae]
oe LEP EELE oS SLs aR ee ore aap a eee nt Stee ee 10
Filaments glabrous. Fleshy appendage of drupe covering two pronounced ribs of
memesocarp. Philippines 22.0. 2.222.23....02.0ee2ceeess 18. Codiocarpus [Stemonuraceae |
Filaments with apical, longish, club-shaped hairs at least in the fertile stamens.
Prominent ribs under the appendage of the drupe less pronounced or absent.
Malesia (except Borneo) ...................-2.:--..- 21. Medusanthera [Stemonuraceae]
Ovary with a lateral swelling which in the fruit develops into a thick succulent
appendage, appearing perpendicular to the drupe. Disk absent. Peninsular
RRRALGSE¥ 05 AM IHIES aso ooo ceded wets 5. Apodytes [Icacinaceae s.str. ]
Ovary and fruit without such an adnate appendage. Disk present or not............ 12
. Connective surpassing the anther cells as a marked glabrous apiculus. Outer part
of the stone finally spongious-corky and deeply irregularly lacunose. Disk absent.
Philippines and New Guinea ................... 7. Merrilliodendron |\cacinaceae s.str. |
Connective, if any, hardly or not surpassing the anther cells. Outer part of the stone
fibrous, slightly ribbed or grooved lengthwise, or smooth outside .................... 13
Peduncle of inflorescence with numerous knob-like bracts which form alveoles.
Stigma peltate. Disk absent. Sumatra, Peninsular Malaysia and Borneo ...............
See ER tier 7 Bees Piers nce ee en 17. Cantleya |Stemonuraceae]
Peduncle quite smooth. Stigma small, subcapitate or point-like. Disk + cup-
SN De ee ete re ac ek ee eas ec ce cde sasseb Sone cdessccdchdavxsotesctstics 14
Inflorescence usually terminal. Anthers glabrous. Philippines, Sumatra and Lesser
SSPSANEL USE ANNES ete 98 codec Sard a Saeed ntwtc con iowes 9. Nothapodytes |Icacinaceae s.str. |
108
14b
15a.
15b.
16.
16.
17a.
17b.
18a.
Gard. Bull. Singapore 63(1 & 2) 2011
. Inflorescence axillary. Anthers with an apical tuft of club-shaped hairs ........... 15
Flowers sessile. Petals up to 8 mm, free to almost the base. Stigma point-like at the
top of the + attenuate (sometimes shortly style-like) part of the ovary. Malesia ......
sd shad vaewswicecandast Sen hesse eee ee 22. Stemonurus [Stemonuraceae]
Flowers 1—2 mm pedicelled. Petals (12—)13—15 mm, free in the distal part only.
Stigma small on one side of the inverted, 1.e. cup-like distal part of the ovary.
SOLOMON 2.8 Rak se oe eek re eee ees 23. Whitmorea |Stemonuraceae |
LEAVES. OPPOSUES <..-scasceccte shes coeeee ese eee eee 17
Leaves: spirally anraneed 2: ccsc2..c..g--eee eee eee ee 18
Anthers broadly club-shaped to subglobular, many-celled, with numerous pollen-
bearing alveoles. Moluccas, New Guinea and the Solomons .................:cc:ccsseeeee
ssa canatvzhaceisausan vi ovese seen cote eee 12. Polyporandra [Icacinaceae s.str. |
Anthers as usual, with 2 cells. Malesia except New Guinea ...............::cccssscceeeseees
Geta cdalte cucu cathe eek CAA one ee dS cast os OE a ee 6. Iodes [Icacinaceae s.str. ]
Twining liana with white milky juice. Flower bisexual (or with plants
andromonoecious). Malesia ..............::cc0008 1. Cardiopteris |Cardiopteridaceae |
. Scandent shrubs, lianas. Plants dioecious (except Slewmerid) ..........cccccceeeeeeeees 19
. Flowers in elongate spikes or spike-like racemes, these solitary or sometimes
composed of panicles: ..0i2.s.4t nace eee eee ee 20
. Flowers in peduncled heads or umbels, these solitary or composed of racemes or
PAMICLSS oie soc cvcwsccsads ba deeah endear kcese se ae eee OO Pd
. Leaves with rather lax but slightly raised reticulation. Sepals absent. Philippines
Vvahastusiavuakertbasseans chceSesdore ce Satter 13. Pyrenacantha [Icacinaceae s.str. |
. Leaves markedly prominently tessellate on both faces. Sepals persistent ......... 21
. Leaves glabrous or very sparsely hairy. Flowers usually 5-merous. Anther
connectives not auriculate. Peninsular Malaysia to Philippines (including Java) ..
weve te ou Ge abcde ERE RUOL Sec Aes re Re See CAE A 15. Sarcostigma [Icacinaceae s.str. |
. All parts hairy with yellow hairs. Flowers usually 4-merous. Anther connectives
auniculate SB One Ot.c.c:c1n eee eee 16. Sleumeria [Icacinaceae s.str. ]
. Style absent, i.e. stigma sessile, thick-peltate. Peninsular Malaysia to the
Philippines. 212 eicrcn see 8. Miquelia [Icacinaceae s.str. ]
. Style (very) shortly thick-columnar, with 2—4 stigmatic lobes. Malesia ................
ss sasledeaanasuade eda aceon cccame toon ee aeeeee tee eee eee 10. Phytocrene [Icacinaceae s.str.]
Malesian Icacinaceae updated 109
Checklist of Cardiopteridaceae, Icacinaceae s.str., and Stemonuraceae in Malesia
CARDIOPTERIDACEAE
A family difficult to characterise, but Karehed (2001) lists free, imbricate sepals,
sympetalous corollas, and epipetalous stamens as possible synapomorphies, which,
together with no stipules, entire leaf margins and one seeded drupes, provide a
combination of characters to help identify the family. The following genera and species
are found in the Flora Malesiana region with the species composition unchanged since
Sleumer (197 1a, b).
1. Cardiopteris Wall. ex Royle
1.1. Cardiopteris moluccana Blume.
42 MOL, PHI, SUL; 43 BIS, NWG-IJ, NWG-PN, SOL-NO.
1.2. Cardiopteris quinqueloba (Hassk.) Hassk.
Distribution: 41 THA; 42 BOR-KA, JAW, LSI-BA, LSI-ET, LSI-LS, MLY-PM,
SUL, SUM.
2. Citronella D.Don
2.1. Citronella latifolia (Merr.) R.A.Howard
Distribution: 42 PHI.
2.2. Citronella philippinensis (Merr.) R.A.Howard
Distribution: 42 PHI.
2.3. Citronella suaveolens (Blume) R.A.Howard
Distribution: 42 BOR-SB, BOR-KA, SUL, SUM, MOL; 43 NWG-IJ, NWG-PN.
3. Gonocaryum Miq.
3.1. Gonocaryum calleryanum (Baill.) Becc.
Distribution: 42 BOR-KA, MOL, PHI, SUL.
3.2. Gonocaryum cognatum Elmer
Distribution: 42 BOR-SB, PHI.
3.3. Gonocaryum crassifolium Ridl.
Distribution: 42 MLY-PM.
3.4. Gonocaryum gracile Miq.
Distribution: 42 MLY-PM, SUM.
3.5. Gonocaryum impressinervium Sleumer
Distribution: 42 BOR-KA, BOR-SR.
3.6. Gonocaryum litorale (Blume) Sleumer
Distribution: 42 LSI-ET, LSI-LS, MOL, SUL; 43 BIS, NWG-IJ, NWG-PN.
3.7. Gonocaryum lobbianum Kurz
Distribution: 41 THA; 42 MLY-PM.
3.8. Gonocaryum macrophyllum (Blume) Sleumer
Distribution: 42 BOR-BR, BOR-KA, BOR-SB, BOR-SR, MLY-PM, SUM.
3.9. Gonocaryum minus Sleumer
Distribution: 42 BOR-BR, BOR-SB, BOR-SR.
110 Gard. Bull. Singapore 63(1 & 2) 2011
4. Pseudobotrys Moeser
4.1. Pseudobotrys cauliflora (Pulle) Sleumer
Distribution: 43 NWG-IJ, NWG-PN.
4.2. Pseudobotrys dorae Moeser
Distribution: 43 NWG-PN.
ICACINACEAE s:str.
The revised concept of Icacinaceae s.str. may be recognised by its racemose
inflorescence of small flowers that are usually pentamerous (although they can be
4—6-merous); glabrous, alternipetalous stamens; unilocular ovary with 2 pendent
ovules, and drupaceous fruits. In addition, the family 1s always woody (climbers or
trees), the leaf margins are entire (although they can be palmately lobed) and exstipulate,
but the leaves can be opposite or alternate. Apart from the description of the new
genus Sleumeria (Utteridge et al. 2005) and a new species of Platea (Utteridge 2010),
the members of this family are unchanged since Sleumer’s (1971a) Flora Malesiana
treatment. The genera Phytocrene and Rhyticaryvum are the most likely candidates to
yield new taxa or to have taxonomic changes made, especially the latter genus which
has several species known only from a few collections from New Guinea.
5. Apodytes E.Mey. ex Arn.
5.1. Apodytes dimidiata E.Mey. ex Arn.
Distribution: 42 BOR-SB, JAW, LSI-LS, MLY-PM, MOL, PHI, SUM.
6. Jodes Blume
6.1. odes cirrhosa Turez.
Distribution: 42 BOR-BR, BOR-SB, BOR-SR, JAW, 7MOL, MLY-PM, MLY-SI,
PHISSUES SUM:
6.2. lodes ovalis Blume
Distribution: 42 JAW, MLY-PM, MLY-SI, SUM.
6.3. lodes philippinensis Mert.
Distribution: 42 BOR-BR, BOR-SB, MOL, PHI, SUL.
6.4. lodes reticulata King
Distribution: 42 MLY-PM.
6.5.1. lodes velutina King var. velutina
Distribution: 42 MLY-PM, MLY-SI.
Distribution: 42 BOR-SR, SUM.
6.6.1. odes vatesii Merr. var. vatesii
Distribution: 42 SUM.
6.6.2. lodes yatesii Merr. var. glabrescens (Ridl.) Sleumer
Distribution: 42 BOR-SR.
7. Merrilliodendron Kaneh.
7.1. Merrilliodendron megacarpum (Hemsl.) Sleumer
Distribution: 42 PHI, SUL; 43 BIS, NWG-IJ, NWG-PN, SOL-NO, SOL-SO.
Malesian Icacinaceae updated
8. Miquelia Meisn.
8.1. Miquelia caudata King
Distribution: 42 BOR-BR, BOR-SB, MLY-PM.
8.2. Miquelia celebica Blume
Distribution: 42 BOR-BR, BOR-SB, BOR-SR, PHI, SUL, SUM.
8.3. Miquelia philippinensis Merr.
Distribution: 42 PHI.
8.4. Miquelia reticulata Mert.
Distribution: 42 PHI.
9. Nothapodytes Blume
9.1. Nothapodytes foetida (Wight) Sleumer
Distribution: 42 PHI, SUM.
9.2. Nothapodytes montana Blume
Distribution: 42 JAW, LSI-LS, SUM.
10. Phytocrene Wall.
10.1. Phytocrene anomala Mert.
Distribution: 42 BOR-SB, BOR-SR.
10.2. Phytocrene borneensis Becc.
Distribution: 42 BOR-BR, BOR-KA, BOR-SB.
10.3. Phytocrene bracteata Wall.
Distribution: 42 BOR-BR, BOR-KA, BOR-SB, BOR-SR, MLY-PM, SUM.
10.4. Phytocrene hirsuta Blume
Distribution: 42 MOL, SUL.
10.5. Phytocrene interrupta Sleumer
Distribution: 43 NWG-PN.
10.6.1. Phytocrene macrophylla Blume var. macrophylla
Distribution: 42 JAW, PHI, SUM.
10.6.2. Phytocrene macrophylla Blume var. caudigera (Sleumer) Sleumer
Distribution: 42 BOR-SB.
10.6.3. Phytocrene macrophylla Blume var. dasycarpa (Miq.) Sleumer
Distribution: 42 SUL.
10.7. Phytocrene malacothrix Sleumer
Distribution: 43 NWG-PN.
10.8. Phytocrene oblonga Wall.
Distribution: 42 MLY-PM.
10.9. Phytocrene palmata Wall.
Distribution: 42 MLY-PM, SUM.
10.10. Phytocrene racemosa Sleumer
Distribution: 42 BOR-SR.
10.11. Phytocrene trichura Ridl.
Distribution: 42 MLY-PM.
111
112 Gard. Bull. Singapore 63(1 & 2) 2011
11. Platea Blume
11.1. Platea bullata Sleumer
Distribution: 42 BOR-SR.
11.2.1. Platea excelsa Blume var. excelsa
Distribution: 42 JAW, SUM.
Distribution: 42 BOR-BR, BOR-SB, BOR-SR, MLY-PM, MLY-SI, SUM.
11.2.3. Platea excelsa Blume var. microphylla (Sleumer) Sleumer
Distribution: 43 NWG-IJ, NWG-PN.
11.2.4. Platea excelsa Blume var. borneensis (Heine) Sleumer
Distribution: 42 BOR-BR, BOR-SB, BOR-SR, JAW, LSI-BA, LSI-LS, MOL, MLY-
PM, MLY-SI, PHI, SUL, SUM; 43 BIS, NWG-IJ, NWG-PN.
11.2.5. Platea excelsa Blume var. kinabaluensis (Sleumer) Sleumer
Distribution: 42 BOR-SB.
11.3. Platea latifolia Blume
Distribution: 42 BOR-BR, BOR-KA, BOR-SB, BOR-SR, JAW, MLY-PM, MLY-SI,
PHI, SUL, SUM; 43 BIS, NWG-IJ, NWG-PN.
11.4. Platea malayana Utteridge
Distribution: MLY-PM.
11.5. Platea sclerophylla Sleumer
Distribution: 42 BOR-SB.
12. Polyporandra Becc.
12.1. Polyporandra scandens Becc.
Distribution: 42 MOL; 43 NWG-IJ, NWG-PN, SOL-NO, SOL-SO.
13. Pyrenacantha Wight
13.1. Pyrenacantha repanda (Merr.) Mert.
Distribution: 42 PHI.
14. Rhyticaryum Becc.
14.1. Rhyticaryum elegans G.Schellenb.
Distribution: 43 NWG-IJ, NWG-PN.
14.2. Rhyticaryum fasciculatum Becc.
Distribution: 43 NWG-IJ.
14.3. Rhyticaryum gracile G.Schellenb.
Distribution: 43 NWG-PN.
14.4. Rhyticaryum longifolium K.Schum. & Lauterb.
Distribution: 43 NWG-IJ, NWG-PN.
14.5. Rhyticaryum lucidum G.Schellenb.
Distribution: 43 NWG-PN.
14.6. Rhyticaryum macrocarpum Becc.
Distribution: 43 NWG-IJ, NWG-PN.
Malesian Icacinaceae updated 113
14.7. Rhyticaryum novoguineense (Warb.) Sleumer
Distribution: 43 NWG-PN.
14.8. Rhyticaryum oleraceum Becc.
Distribution: 42 LSI-LS, MOL; 43 NWG-IJ.
14.9. Rhyticaryum oxycarpum K.Schum. & Lauterb.
Distribution: 43 NWG-PN.
14.10. Rhyticaryum purpurascens G.Schellenb.
Distribution: 43 NWG-PN.
14.11. Rhyticaryum racemosum Becc.
Distribution: 43 NWG-IJ.
14.12. Rhyticaryum rotundatum G.Schellenb.
Distribution: 43 NWG-PN.
15. Sarcostigma Wight & Arn.
15.1. Sarcostigma kleinii Wight & Arn.
Distribution: 42 BOR-SR, JAW, MLY-PM.
15.2. Sarcostigma paniculata Pierre
Distribution: 42 BOR-BR, BOR-SR, MLY-PM, PHI.
16. Sleumeria Utteridge, Nagam. & Teo
16.1. Sleumeria auriculata Utteridge, Nagam. & Teo
Distribution: 42 BOR-BR, BOR-KA, BOR-SB, BOR-SR.
STEMONURACEAE
Members of Stemonuraceae are trees or shrubs with falcate, naked terminal buds,
entire alternate leaves, and usually green young twigs. Inflorescences may be
axillary, terminal or leaf-opposed (rarely ramiflorous), with one to many flowers,
often umbelliform with cymose branches. The flowers are bisexual or functionally
unisexual, with a small cupular calyx and 4-5 free to connate petals that are often
inflexed at the apex. The stamens (and staminodes) are characteristically flattened
(except Codiocarpus), with clavate hairs on the filament. The fruits are usually white,
yellow, pink or red, and three genera have a prominent lateral appendage on one side
of the drupe. Recently Gomphandra and Medusanthera have been revised with new
species described (see Schori 2010, Schori & Utteridge 2010, and Utteridge 2011), and
a new species of Stemonurus has been described (Utteridge & Schori 2009).
17. Cantleya Ridl.
17.1. Cantleya corniculata (Becc.) R.A.Howard
Distribution: 42 BOR-BR, BOR-KA, BOR-SB, BOR-SR, MLY-PM, MLY-SI, SUM.
18. Codiocarpus R.A.Howard
18.1. Codiocarpus merrittii (Merr.) R.A.Howard
Distribution: 42 PHI.
114 Gard. Bull. Singapore 63(1 & 2) 2011
19. Gomphandra Wall. ex Lindl.
19.1. Gomphandra angustata Schori ined.
Distribution: 43 NWG-PN.
19.2. Gomphandra australiana F.Muell.
Distribution: 42 MOL; 43 NWG-PN.
19.3. Gomphandra borneensis Schori ined.
Distribution: 42 BOR-BR, BOR-SB, BOR-SR.
19.4. Gomphandra bracteata Schori ined.
Distribution: 42 PHI.
19.5. Gomphandra capitulata (Jungh. & de Vriese) Becc.
Distribution: 42 BOR-KA, MLY-PM, SUM.
19.6. Gomphandra coi Schori ined.
Distribution: 42 PHI.
19.7. Gomphandra chimaera Schori ined.
Distribution: 42 SUM.
19.8. Gomphandra conklinii Schori
Distribution: 42 PHI.
19.9. Gomphandra cumingiana (Miers) Fern.-Vill.
Distribution: 42 BOR-KA, BOR-SB, BOR-SR, PHI.
19.10. Gomphandra dinagatensis Schori ined.
Distribution: 42 PHI.
19.11. Gomphandra dolichocarpa Mert.
Distribution: 42 SUM.
19.12. Gomphandra engganensis Schori ined.
Distribution: 42 SUM.
19.13. Gomphandra fernandoi Schori & Utteridge ined.
Distribution: 42 PHI.
19.14. Gomphandra flavicarpa (Elmer) Merr.
Distribution: 42 PHI.
19.15. Gomphandra fuliginea (Elmer) Merr.
Distribution: 42 PHI.
19.16. Gomphandra fusiformis Sleumer
Distribution: 42 SUM.
19.17. Gomphandra halconensis Schori
Distribution: 42 PHI.
19.18. Gomphandra jacobsii Schori ined.
Distribution: 42 SUM.
19.20.1. Gomphandra javanica (Blume) Valeton subsp. javanica
Distribution: 42 JAW, LSI-BA, LSI-LS.
19.20.2. Gomphandra javanica (Blume) Valeton subsp. pseudojavanica (Sleumer)
Schori ined.
Distribution: 42 SUM.
19.21.1. Gomphandra kinabaluensis Schori ined. var. kinabaluensis
Distribution: 42 BOR-SB.
Malesian Icacinaceae updated 115
19.21.2. Gomphandra kinabaluensis Schori var. clemensiorum Schori ined.
Distribution: 42 BOR-SB.
19.22. Gomphandra lamanii Schori ined.
Distribution: 42 BOR-KA, BOR-SR.
19.23. Gomphandra lancifolia Mert.
Distribution: 42 PHI.
19.24. Gomphandra longipedunculata Schori ined.
Distribution: 42 BOR-SR.
19.25.1. Gomphandra luzoniensis (Merr.) Merr. subsp. /uzoniensis
Distribution: 42 PHI.
19.25.2. Gomphandra luzoniensis (Merr.) Merr. subsp. septentrionalis Schon &
Utteridge ined.
Distribution: 38 TAI; 42 PHI.
19.26. Gomphandra lysipetala Stapf
Distribution: 42 BOR-SB, BOR-SR.
19.27. Gomphandra macrosperma Schori ined.
Distribution: 42 BOR-SB.
19.28. Gomphandra mappioides Valeton
Distribution: 42 LSI-ET, LSI-LS, MOL, PHI, SUL.
19.29.1. Gomphandra melanesiensis Schori ined. subsp. melanesiensis
Distribution: 43 BIS, SOL-NO, SOL-SO.
19.29.2. Gomphandra melanesiensis Schori subsp. macrocarpa Schori ined.
Distribution: 43 SOL-SO.
19.30. Gomphandra microcarpa Schori ined.
Distribution: 42 MLY-PM.
19.31. Gomphandra montana (G. Schellenb.) Sleumer
Distribution: 43 NWG-PN.
19.32. Gomphandra muscosa Schori ined.
Distribution: 43 NWG-PN.
19.33. Gomphandra oblongifolia Merr.
Distribution: 42 PHI.
19.34. Gomphandra oligantha Sleumer
Distribution: 42 PHI.
19.35. Gomphandra palustris Schori
Distribution: 42 BOR-SR.
19.36. Gomphandra papuana (Becc.) Sleumer
Distribution: 43 NWG-IJ, NWG-PN.
19.37.1. Gomphandra parviflora (Blume) Valeton var. parviflora
Distribution: 42 SUM.
19.37.2. Gomphandra parviflora (Blume) Valeton var. magnifolia Schori ined.
Distribution: 42 SUM.
19.37.3. Gomphandra parviflora (Blume) Valeton var. paucibarbata Schori ined.
Distribution: 42 SUM.
116 Gard. Bull. Singapore 63(1 & 2) 2011
19.38. Gomphandra psilandra Schori ined.
Distribution: 42 PHI.
19.39. Gomphandra pseudoprasina Sleumer
Distribution: 43 NWG-PN.
19.40. Gomphandra puberula Rid.
Distribution: 42 MLY-PM.
19.41. Gomphandra quadrifida (Blume) Sleumer
Distribution: 42 MLY-PM, MLY-SI, SUM.
19.42. Gomphandra ramuensis (Lauterb.) Sleumer
Distribution: 43 NWG-IJ, NWG-PN.
19.43. Gomphandra rarinervis Schori
Distribution: 43 NWG-PN.
19.44. Gomphandra schoepfiifolia Sleumer
Distribution: 43 NWG-PN.
19.45. Gomphandra simalurensis Sleumer
Distribution: 42 SUM.
19.46. Gomphandra simulans Schori ined.
Distribution: 42 SUM.
19.47. Gomphandra subcordata Schori ined.
Distribution: 43 NWG-PN.
19.48. Gomphandra subrostrata Mert.
Distribution: 42 SUM.
19.49. Gomphandra tenuis Schori ined.
Distribution: 42 MLY-PM.
19.50. Gomphandra tomentella (Kurz) Mast.
Distribution: 41 THA; 42 MLY-PM.
19.51. Gomphandra ultramafiterrestris Schori ined.
Distribution: 42 PHI.
19.52. Gomphandra velutina Sleumer
Distribution: 42 SUL.
20. Hartleya Sleumer
20.1. Hartleya inopinata Sleumer
Distribution: 43 NWG-PN.
21. Medusanthera Seem.
21.1. Medusanthera gracilis (King) Sleumer
Distribution: 42 MLY-PM, SUM.
21.2. Medusanthera inaequalis Utteridge
Distribution: 43 NWG-IJ.
21.3. Medusanthera laxiflora (Miers) R.A.Howard
Distribution: 42 PHI; 43 NWG-IJ, NWG-PN, SOL-NO, SOL-SO.
21.4. Medusanthera malayana Utteridge
Distribution: 42 MLY-PM.
Malesian Icacinaceae updated ily
21.5. Medusanthera megistocarpa Utteridge
Distribution: 43 NWG-PN.
22. Stemonurus Blume
22.1. Stemonurus ammui (Kaneh.) Sleumer
Distribution: 43 BIS, NWG-PN, NWG-IJ, SOL-NO, SOL-SO.
22.2. Stemonurus celebicus Valeton
Distribution: 42 SUL.
22.3. Stemonurus corrugatus Utteridge & Schori
Distribution: 42 BOR-SR.
22.4. Stemonurus gitingensis (Elmer) Sleumer
Distribution: 42 PHI.
22.5. Stemonurus grandifolius Becc.
Distribution: 42 BOR-SB, BOR-SR, BOR-KA.
22.6. Stemonurus hallieri (Merr.) Merr.
Distribution: 42 PHI.
22.7. Stemonurus malaccensis (Mast.) Sleumer
Distribution: 42 BOR-KA, BOR-SB, BOR-SR, BOR-BR, MLY-PM.
22.8. Stemonurus monticola (G. Schellenb.) Sleumer
Distribution: 43 NWG-IJ, NWG-PN.
22.9. Stemonurus scorpioides Becc.
Distribution: 42 BOR-BR, BOR-SB, BOR-SR, JAW, MLY-PM, MLY-SI, SUM.
22.10. Stemonurus secundiflorus Blume
Distribution: 42 BOR-BR, BOR-KA, BOR-SB, BOR-SR, JAW, MLY-PM, SUM.
22.11. Stemonurus umbellatus Becc.
Distribution: 42 BOR-SR, BOR-KA, BOR-BR, BOR-SB, MLY-PM.
23. Whitmorea Sleumer
23.1. Whitmorea grandiflora Sleumer
Distribution: 43 SOL-NO, SOL-SO.
Summary
Family changes due to molecular and morphological data have altered the concept of
many families in the Flora Malesiana region. These family changes are summarised
in the APG publications (e.g., see Stevens 2001). The Icacinaceae is one such family
and has now been re-circumscribed into new groups. In 1971 the family was published
for the Flora Malesiana and, with the Cardiopteridaceae considered a separate family,
the group then included 100 species in 22 genera in two families (Sleumer 1971a,
1971b). The current concept of the group is now of three families: Icacinaceae s.str.
with 48 species in 12 genera; Cardiopteridaceae with 15 species in four genera; and
Stemonuraceae with 72 species in seven genera.
118 Gard. Bull. Singapore 63(1 & 2) 2011
References
Balgooy, M.M.J. van (1998) Malesian Seed Plants, vol. 2. Portraits of Tree Families.
Leiden: Rijksherbarium/Hortus Botanicus.
Brummitt, R.K. (2001) World Geographical Scheme for Recording Plant Distributions.
Ed. 2. International Working Group on Taxonomic Databases For Plant Sciences
(TDWG). Hunt Institute for Botanical Documentation.
Doleshy, F. (1966) Review of Sleumer ‘An account of Rhododendron in Malesia’. J.
Amer. Rhododendron Soc. 20(4).
Karehed, J. (2001) Multiple origin of the tropical forest tree family Icacinaceae. Amer.
J. Bot. 88: 2259-2274.
Karehed, J. (2002) Evolutionary studies in Asterids emphasising Euasterids II.
Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science
and Technology 761. Uppsala: Acta Universitatis Upsaliensis.
Mabberley, D.J. (1997) The Plant Book. Cambridge: Cambridge University Press.
Savolainen, V., Chase, M.W., Hoot, S.B., Morton, C.M., Soltis, D.E., Bayer, C., Fay,
M.F., Bruyn, A.Y. de, Sullivan, S. & Qiu, Y.-L. (2000) Phylogenetics of flowering
plants based upon a combined analysis of plastid atpB and rbcL gene sequences.
Syst. Biol. 49: 306-362.
Schori, M. (2010) A systematic revision of Gomphandra. Dissertation. Athens, Ohio,
USA: Ohio University.
Schori, M. & Utteridge, T.M.A. (2010) Three new species and a new name in Southeast
Asian Gomphandra (Stemonuraceae). Blumea 55: 189-195.
Sleumer, H.O. (1971a) Icacinaceae. Flora Malesiana Ser. I, 7: 1-87.
Sleumer, H.O. (1971b) Cardiopteridaceae. Flora Malesiana Ser. 1, 7: 93-96.
Soltis D.E., Soltis, P.S., Chase, M.W., Mort, M.E., Albach, D.C., Zanis, M., Savolainen,
V., Hahn, W.H., Hoot, S.B., Fay, M.F., Axtell, M., Swensen, S.M., Prince, L.M.,
Kress, W.J., Nixon, K.C. & Farris, J.S. (2000) Angiosperm phylogeny inferred
from 18S rDNA, rbcL, and atpB sequences. Bot. J. Linn. Soc. 133: 381-461.
Stevens, P.F. (2001 onwards) Angiosperm Phylogeny Website. Version 9, June 2008
[and updated since]. http://www.mobot.org/MOBOT/research/A Pweb/
Utteridge, T.M.A. (2001) A new species of Medusanthera Seem. (Icacinaceae) from
New Guinea: Medusanthera inaequalis Utteridge. Contributions to the flora of
Mt Jaya, IV. Kew Bull. 56: 233-237.
Utteridge, T.M.A. (2010) A new species of Platea (Icacinaceae) from Peninsular
Malaysia: Platea malayana Utteridge. Kew Bull. 65: 345-348.
Utteridge, T.M.A. (2011) A revision of the genus Medusanthera (Stemonuraceae,
Icacinaceae Ss.|.). Kew Bull. 66: 49-81.
Utteridge, T.M.A. & Schori, M. (2009) A new species of Stemonurus (Stemonuraceae,
Icacinaceae 8.|.) from Sarawak: Stemonurus corrugatus Utteridge & Schori. Kew
Bull. 64: 327-331.
Utteridge, T.M.A., Nagamasu, H., Teo, S.P., White, L.C. & Gasson, P. (2005) Sleumeria
(Icacinaceae): A new genus from northern Borneo. Syst. Bot. 30(3): 635-643.
Gardens’ Bulletin Singapore 63(1 & 2): 119-124. 2011 119
Clerodendrum confusion—redefinition of,
and new perspectives for, a large Labiate genus
J.A. Wearn! and D.J. Mabberley”
‘Herbarium, Library, Art and Archives,
Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3AE, U.K.
j-wearn@kew.org (corresponding author)
"Royal Botanic Gardens & Domain Trust, Sydney, New South Wales,
NSW 2000, Australia
ABSTRACT. Formerly referred to Verbenaceae s.l., Clerodendrum L. is one of the largest
genera within the Lamiaceae (Labiatae) s.l., and many of its species are of ecological and
commercial importance. However, confusion about species delimitation and identification
has reigned for many decades, resulting in large quantities of unidentified, or misidentified,
herbarium material. Results from recent molecular studies have provided a framework for
accurate placement of taxa. The revised concept of the genus is applied to taxa in Malesia in
order to produce a modern account for Flora Malesiana, which includes up-to-date descriptions
and much-needed keys. Progress made so far is reported.
Keywords. Clerodendrum, Labiatae, Lamiaceae, Malesia, Verbenaceae s.1.
Introduction
The long-standing Flora Malesiana project, first and foremost, aims to create “a
systematic account of the flora of Malesia, the plant-geographical unit spanning six
countries in Southeast Asia” (www.floramalesiana.org). With such high botanical
diversity in the region (an estimated 42,000 plant species: Roos 1993) and general
paucity of funds for research, why is Clerodendrum a genus worthy of particular note?
Clerodendrum is a large genus containing species that are important both
ecologically and commercially. The ecology of, often genus-specific, associations with
microfungi (Hosagoudar & Archana 2009, and see Minter 2010), and relationships
with ants (Maschwitz et al. 1994) and pollinators (Corner 1940, Yuan et al. 2010)
has been studied to a limited extent though further research is required in order to
understand these complex interactions. Some species are early successional colonisers
of degraded land (e.g., C. japonicum (Thunb.) Sweet) and could be used for habitat
restoration, but, others, like C. chinense (Osbeck) Mabb., can become pernicious
weeds, so that accurate identification and careful consideration must go hand-in-hand
before implementing restoration work. Many of the Malesian and other species are
highly ornamental, having large showy inflorescences and attractive foliage, and some
are already popular in the global horticultural trade (several, including the spectacular
orangey red and cream to pale yellow-flowered forms of C. paniculatum L., can be
120 Gard. Bull. Singapore 63(1 & 2) 2011
Fig. 1. Two colour forms of Clerodendrum paniculatum are popular in cultivation due to their
striking inflorescences and foliage (one form has cream to pale yellow corollas with green
inflorescence branches, and the other has pale orangey-red corollas with red inflorescence
branches). These plants are growing in the Singapore Botanic Gardens. Photos by J.A. Wearn,
2010.
Clerodendrum in Malesia 121
seen in the Singapore Botanic Gardens, for example; see Fig. 1). Clerodendrum species
have also been used medicinally for centuries in their countries of origin and rigorous
scientific trials are now underway to evaluate the potential of compounds extracted
from them, notably those with antipyretic and anti-inflammatory properties (examples
are cited in Shrivastava & Patel 2007), and antiviral activity (Kim et al. 2001).
Clerodendrum was long referred to Verbenaceae s.]. but its placement was
corrected following a series of molecular-based studies (Olmstead et al. 1993,
Wagstaff et al. 1998), such that it now falls into Lamiaceae (Labiatae). Molecular
studies during the 1990s lacked resolution below family level as only a few species
from each of the constituent genera were included. Although previously considered as
a pantropical and eastern temperate genus, 1t was recognised that Clerodendrum in this
broad sense was heterogenous and likely to be polyphyletic. The need for elucidation
of inter- and intrageneric relationships was addressed in a landmark paper by Yuan
et al. (2010), using cpDNA from 56 taxa (including 40 Clerodendrum species, sensu
Steane et al. 2004). They showed that the native American species were misplaced in
Clerodendrum and so those are now referred to other genera. Many of the African taxa
had already been excluded and placed in a revived Rotheca Raf., a more distantly allied
genus (Steane & Mabberley 1998, Steane et al. 2004). A combination of molecular
and morphological taxonomic studies necessitated reassessment of traditionally used
morphological characters, allowing the focus to be directed towards those characters
now seen as most relevant for taxonomic delimitation at the generic level and below.
Thus, a redefinition of Clerodendrum s.s. and its allies was completed (Yuan et al.
2010). This allowed us to undertake a reassessment of all Malesian taxa, previously
considered to be ‘Clerodendrum’.
Towards a reassessment of Malesian taxa
When Moldenke (1985: 310) wrote of Clerodendrum sensu lato, he referred to 584
taxa, the majority of which had one or more synonyms. Now numbering approximately
150-180 species, Clerodendrum is an exclusively Old World genus, its species
distributed largely within the tropics and subtropics with some found as far south as
Australia and as far north as central China and Japan. Outside Africa, the majority of
Clerodendrum species is found in the Malesian region but there has been no major
revisionary work on those for nearly a century. Thus, at the outset, we were largely
reliant upon the accounts of Schauer (1847) and Lam (1919). An account by Backer
& Bakhuizen van den Brink (1965) included taxa found in Java while a series of later
papers (Moldenke 1985-87) covered the genus only in part, with some questionable
infraspecific delimitation and other circumscriptions. One of us (DM) had prepared
a manuscript account of the genus and its allies for the Flora of Peninsular Malaysia
(Mabberley, in press) but it became obvious very quickly, from the literature alone, that
there were a multitude of names in use and abundant synonymy (varying depending
on the author) applied to taxa beyond the Malay Peninsula. However, it was not until
one of us (JW) began to trawl through the large numbers unidentified or misnamed
| Gard. Bull. Singapore 63(1 & 2) 2011
collections in herbaria that the confusion of species concepts and the full extent of the
task were realised. It was not uncommon for material to have been reidentified several
times—in some cases reaching a nomenclatural ‘full circle’, where the last botanist
viewing a particular collection had disagreed with those before him/her but agreed
with the original, contemporary identification! It soon became apparent that the Flora
Malesiana account would require a revision of monographic intensity.
Flora Malesiana and the future of Clerodendrum
As we near the end of our reassessment of Clerodendrum in Malesia, 210 names
have been considered since our project began in mid-2008 (Wearn & Mabberley, in
prep.). So far, 53 species are recognised, 13 names have been excluded from the genus
(referred to Faradaya F. Muell., Hosea Ridl., Rotheca and Volkameria L.) and two
have been excluded from the Flora Malesiana account due to incorrect understanding
of distributions (species now thought to be native or naturalised only outside Malesia).
A further eight names remain ‘insufficiently known’. The loss of type material and
lack of other collections aligning with the descriptions has meant that no progress
in such cases can yet be made. For example, Clerodendrum barba-felis Hallier f.
was described based on material deposited at WRSL, with a duplicate at PNH, now
both lost, with no conspecific material found elsewhere. This situation is by no
means peculiar to Clerodendrum as nearly every volume of Flora Malesiana contains
names excluded from accounts in this way. Unfortunately, it is a geographically and
taxonomically wide-ranging phenomenon, owed in part to the destruction of herbaria
or loss of sets through auctions, fires and so on, but perhaps also through the rarity of
the species described. Considering the rate of destruction of habitats in Malesia, the
species that have not been collected for over 50 years may now be extinct, particularly
as Clerodendrum species are by no means cryptic.
While many Clerodendrum species are not considered of conservation
concern, being commonly collected (e.g., C. disparifolium Blume), several have
not been seen in Malesia for 50 years or more (e.g., C. umbratile King & Gamble
(Mabberley, in press)), whether due to geographical sampling effort or species rarity.
A few species have very restricted geographical ranges and are considered vulnerable
(e.g., C. lankawiense King & Gamble, which is found only on certain islands off south-
west Peninsular Thailand and northwest Peninsular Malaysia (Mabberley, in press)).
Others, such as C. albiflos H.J. Lam from New Guinea and C. multibracteatum Mert.
from the Philippines, are known from only one or a few collections, and so have to
be categorised as ‘data deficient’ until more is known. Unfortunately, but perhaps not
surprisingly, the species that are in cultivation are those which are common naturally,
perhaps due as much to the ease with which they can be propagated and grown as it is
a result of their floral appeal. Currently, there is no ex sifu conservation resource.
Plants of the genus are commonly encountered during fieldwork but
frequently unidentified, or worse, misidentified so that they end up in completely the
wrong herbarium cupboard. We hope that our work for Flora Malesiana will clarify
123
i
Clerodendrum in Malesiz
Jiquitous confusion which has dominated this large and important
the seemingly ul ae
ee loin because our revision of Malesian Clerodendrum taxa has allowed
S 2w descriptions and keys (Wearn & Mabberley, in press, in prep.),
preparation of nc... : : ge
2quisite for accurate identification in the field and herbarium. These
which are a prer«. : :
RR ists as well as those undertaking conservation and restoration work.
will aid taxonom
a conc] UStO2> the importance of interdisciplinary collaboration (such as
nists, ethnobotanists and ecologists) and knowledge exchange must
between taxonon
is through this process much more can be understood about the taxa
2 eed the information that one seeks may have been already documented,
ent Supers in other disciplines are consulted, no-one may be able to make full
use of the ncomP ee knowledge. For example, during this work, Jw found that C.
Beaiacim de Vriese & Teijsm.. described from Ambon in Indonesia, was poorly
ex onceen botanists. An adequate account of the plant was created only as
Resa of Sci with an ethnobotanist (Roy Ellen at the University of Kent). RE had
slant several times during long-term (1970—present) research on the
ae ne tnd of Seram and was able to provide much additional information
g é e “raphs and notes on local uses, in addition to much-needed recent
including photog, Sie
material (Wearn « sia
References
ee CA & pakhuizen van den Brink, R.C. (1965) Clerodendrum L. Flora of Java
2: 607-611.
Comer, E.J.H. (;
es ce.
See & Archana, G.R. (2009) Host range of meliolaceous fungi in India.
Hf Threatendia Taxa 1: 269-282.
Kim Hd weg E-R» Shin, C.G., Hwang, D.J., Park, H. & Lee, YS. (2001)
HIV-1 intee@@S° inhibitory phenylpropanoid glycosides from Clerodendrum
“i IMSS Arch. Pharm. Res. 24: 286-291.
Seated Verbenaceae of the Malayan Archipelago. The Netherlands,
Lam, HJ. (191
M. de Waal.
mo (in press) Clerodendrum L. Flora of Peninsular Malaysia. Selangor:
Mabberley, D.J. |
ey Resettch Institute Malaysia.
~~ Fiala, B. & Linsenmair, K.E. (1994) Clerodendrum fistulosum
Maschwitz, U., é 3
(Ve a an unspecific myrmecophyte from Borneo with spontaneously
: ‘natia. Blumea 39: 143-150.
ae eee ; of Cybertruffle s Robigalia. Observations of fungi and their associated
: *" www.cybertruffle.org.uk/robigalia/eng (accessed 18 Aug 2010).
yo tray, (1985-1987) Notes on the genus Clerodendrum (Verbenaceae).
Moldenke, HN 5764 (series of papers)
Phytologia - a
1940) Wayside Trees of Malaya, vol. 1. Singapore: Government
124 Gard. Bull. Singapore 63(1 & 2) 2011
Olmstead, R.G., Bremer, B., Scott, K.M. & Palmer, J.D. (1993) A parsimony analysis
of the Asteridae sensu lato based on rbcL sequences. Ann. Missouri Bot. Gard.
80: 700-722.
Roos, M.C. (1993) State of affairs regarding Flora Malesiana: progress in revision
work and publication schedule. Flora Malesiana Bull. 11: 133-142.
Schauer, J.C. (1847) Clerodendron Linn. In: Candolle, A.P. de, Prodromus Systematis
Naturalis Regni Vegetabilis 11: 658-675. Paris: Sumptibus Sociorum Treuttel &
Wirtz.
Shrivastava, N. & Patel, T. (2007) Clerodendrum and healthcare: an overview. Med.
Aromat. Pl. Sci. Biotechnol. 1: 142-150.
Steane, D.A. & Mabberley, D.J. (1998) Rotheca (Lamiaceae) revived. Novon 8: 204—
206.
Steane, D.A., Kok, R.P.J. de & Olmstead, R.G. (2004) Phylogenetic relationships
between Clerodendrum (Lamiaceae) and other Ajugoid genera inferred from
nuclear and chloroplast DNA sequence data. Molec. Phylogen. Evol. 32: 39-4S.
Wagstaff, S.J., Hickerson, L., Spangler, R., Reeves, P.A. & Olmstead, R.G. (1998)
Phylogeny in Labiatae s.1., inferred from cpDNA sequences. P/. Syst. Evol. 209:
265-274.
Wearn, J.A. & Ellen, R.F. (in prep.) Dispelling the mystery of Clerodendrum
rumphianum (Lamiaceae) through interdisciplinary knowledge sharing.
Wearn, J.A. & Mabberley, D.J. (in press) Clerodendrum (Lamiaceae) in Borneo. Syst.
Bot. 36(4)
Wearn, J.A. & Mabberley, D.J. (in prep.) Lamiaceae. Clerodendrum L. Flora
Malesiana. Series I. Leiden: Nationaal Herbarium Nederland.
Yuan, Y.-W., Mabberley, D.J., Steane, D.A. & Olmstead, R.G. (2010) Further
disintegration and redefinition of Clerodendrum (Lamiaceae): implications for
the understanding of the evolution of an intriguing breeding strategy. Jaxon 59:
125-133.
Gardens’ Bulletin Singapore 63(1 & 2): 125-135. 2011 125
A synopsis of Coelostegia
(Bombacaceae/Malvaceae: Helicteroideae: Durioneae)
and new records from Borneo
I. Nadiah' and E. Soepadmo
Forest Research Institute Malaysia, 52109 Kepong, Selangor, Malaysia
‘nadiahidns@fnim.gov.my
ABSTRACT. A synoptic revision of Coe/ostegia Benth. (Bombacaceae/Malvaceae subfam.
Helicteroideae—Durioneae) in Borneo is given. Six species are recognised, of which four
(C. chartacea, C. kostermansii, C. montana and C. neesiocarpa) are endemic to Bomeo.
Coelostegia griffithii, previously recorded only from Peninsular Malaysia, Singapore, Java
and Sumatra, is now also found in Bomeo, while C. montana previously known only from
Sarawak and Kalimantan also occurs in Sabah. Gross morphological and micromorphological
characters show that the genus Coe/ostegia can be readily distinguished from other genera
in the Durioneae-group by the epicalyx being much shorter than the calyx, the induplicate-
saccate calyx character and the ovary being partly enclosed by the receptacle. The distinction is
also supported by micromorphological characters derived from trichomes, stomata, and pollen.
Nomenclatural (typification and synonymy) and taxonomic notes, ecology and geographical
distribution of the recognised species are provided.
Keywords. Borneo, Coelostegia, Durioneae, Helicteroideae, Malvaceae, taxonomy
Introduction
Bentham (1862) first described Coelostegia with one species, C. griffithii, from
Peninsular Malaysia. Beccari (1886) described two more species, from Sumatra (C.
sumatrana) and Borneo (C. borneensis), and was the first to describe the fruit and seed
of Coelostegia. Soegeng (1960) revised the genus and added three more species from
Borneo, and provided full descriptions, an identification key and illustrations of all
five species known to him. Sidiyasa (2001) described a new species, C. montana, from
East Kalimantan and Sarawak.
Prior to 1998, taxonomic and systematic studies based mainly on morphological
and anatomy characters carried out by various authors (e.g., Hutchinson 1959;
Cronquist 1968, 1981: Keng 1969; Takhtajan 1969) included Boschia, Coelostegia,
Cullenia, Durio, Kostermansia and Neesia in the tribe/section Durioneae of the
family Bombacaceae. From the late 1990°s, however, phylogenetic studies based on
chloroplast and nuclear ribosomal DNA (e.g., Alverson et al. 1998, 1999: Baum et al.
1998: Bayer et al. 1999: Nyffeler & Baum 2000; Bayer & Kubitzki 2003) strongly
suggested that the core Malvales families (Bombacaceae, Malvaceae, Sterculiaceae
and Tiliaceae) should be merged into an expanded family Malvaceae, and that nine
126 Gard. Bull. Singapore 63(1 & 2) 2011
subfamilies should be recognised, with the genera of the Durioneae-group to be included
in subfam. Helicteroideae-Durioneae. Cheek (2006, 2007), however, disagreed and
proposed placing the Durioneae genera in a separate family, the Durionaceae.
Synopsis of recognised taxa
Coelostegia Benth., Gen. Pl. 1 (1862) 213; Hooker f., Fl. Brit. India 1 (1875) 352;
Beccari, Malesia 3 (1889) 269; King, J. As. Soc. Beng. 60, 1 (1891) 56; Schumann in
Engler & Prantl., Nat. Pflanzenfam. 3, 6 (1895) 68; Ridley, Fl. Malay Penins. 1 (1922)
266; Bakhuizen f., Bull. Jard. Bot. Buitenz. 6, 3 (1924) 223; Soegeng, Reinwardtia
5, 3 (1960) 270; Hutchinson, Gen. Flow. Pl. 2 (1967) 526; Kochummen, Tree FI.
Malaya | (1972) 104; Cockburn, Trees of Sabah | (1976) 22; Ashton, Man. Non-Dipt.
Trees Sarawak 2 (1988) 53; Salma et al., Pl. Resources of South-East Asia 5, 2 (1995)
140; Coode et al. (eds), Checkl. Flow. Pl. Gymno. Brunei (1996) 41; Argent et al.
(eds), Man. Non-Dipt. Trees Centr. Kalimantan | (1997) 96; Beaman et al., Pl. Mount
Kinabalu 4 (2001) 164; Bayer & Kubitzki, Fam. Gen. Vasc. Pl. 5 (2003) 261. TYPE
SPECIES: Coelostegia griffithii Benth.
Distribution. Six species distributed in Sumatra (including Riau Archipelago),
Peninsular Malaysia, Java, Singapore and Borneo. In Borneo, four species are
endemics; Sarawak has five species (non endemic); Sabah three species (non endemic);
Brunei two species (non endemic) and Kalimantan five species (one endemic) (Fig 1).
Ecology. Lowland mixed dipterocarp and lower montane forest on clay-rich soils, to
c. 1450 m.
Notes. Soegeng (1959, 1960) pointed out that based on their overall vegetative and
reproductive characters, Coelostegia, Durio, Kostermansia and Neesia are distinct
genera but closely related to one another. Basing his conclusion on the anatomy of
vegetative parts, Baas (1972) fully supported Soegeng’s suggestion. Appendix A
summarises the macromorphological and micromorphological characters which can
be used to distinguish Coe/ostegia from the other genera.
1. Coelostegia borneensis Becc., Malesia 3 (1889) 272, Nelle Foreste Di Borneo
(1902) 572; Merrill, J. Str. Br. Roy. As. Soc., Spec. No. (1921) 377; Bakhuizen f.,
Bull. Jard. Bot. Buitenz. 6, 3 (1924) 224; Masamune, En. Phan. Born. (1942) 454;
Soegeng, Reinwardtia 5, 3 (1960) 272; Kochummen, Tree Fl. Malaya | (1972) 106;
Anderson, Check]. Trees Sarawak (1980) 153; Ashton, Man. Non-Dipt. Trees Sarawak
2 (1988) 54; Turner, Gard. Bull. Sing. 47, 1 (1995) 151; Salma et al., Pl. Resources
of South-East Asia 5, 2 (1995) 143; Argent et al. (eds), Man. Non-Dipt. Trees Centr.
Kalimantan | (1997) 97. TYPE: Beccari PB 2688, Borneo, Sarawak, Kuching district
(holo FI, iso BO! K!).
Coelostegia in Borneo 127
95°00°E 100°0'0"E 105°0'0"E 110°00°E 115°00°'E 120°0'0"E
10°0'0"N
10°0'0"N
PENINSULAR
©) MALAYSIA
S'0'0"N
5°0'0"N
>
0°0'0"
0°0'0"
tre nm
Ln 44
SUMATERA™. \ BORNEO
Legend
© C. bomeensis
C. chartacea
C. gnffithii
C. kostermansii
C. montana
C. neesiocarpa
S'0'0"S
5'0'0"S
10°0'0"S
10°0'0"S
95°0'0°E 100°0°0°E 105°0°0°E 110°00°E 11S5°00°E 120°00°E
Fig. 1. Distribution of six Coelostegia species in Peninsular Malaysia, Singapore, Borneo,
Sumatra and Java.
Distribution. Sumatra, Peninsular Malaysia and Borneo (Sarawak, Kalimantan,
Brunei).
Ecology. Lowland mixed dipterocarp, kerangas and swampy forest on slopes, river
banks, hillsides, low undulating country, on waterlogged soils and deep yellow sands
overlying tertiary clays at altitude 20-303 m.
Notes. In leaf surface, size, number of lateral veins, twig and types of fruit spines, C.
borneensis is closely related to C. chartacea but differs in the leaf texture (coriaceous
vs. chartaceous), leaf apex (short-acuminate, acumen 0.6 cm vs. long-acuminate,
acumen up to 1.5 cm), and fruit shape (globular vs. ellipsoid).
Specimens examined: PENINSULAR MALAYSIA. Johor: Lenggor FR, Whitmore
FRI 8651 (KEP); Gunung Sumalayang, Everett FRI 13875 (K, KEP, L, SAN, SING);
Panti FR, Pilus KEP 104507 (KEP); Labis FR, Yaacob KEP 104744 (KEP); Kluang FR,
Heaslett s.n. (KEP); ibid., Samsuri SA 391 (SING); Kg. Hubong, Kadim KN 284 (BO,
L, SING). Pahang: Aur FR, Whitmore FRI 3626 (KEP, L); Kuala Kemapan, Saw FRI
34187 (KEP): Menchali FR, Meijer KEP 94890 (K, KEP, L, SING). Selangor: Bukit
Lagong FR, Hamid KEP 81089 (KEP); Bukit Belata FR, Kochummen KEP 99372
(KEP); Terengganu: Dungun, Abdullah KEP 53363 (BO, KEP). -SUMATRA. North
128 Gard. Bull. Singapore 63(1 & 2) 2011
Sumatra: Langsa, Boschbouwproefstation bb. 2578 (BO, BZF, L). East Sumatra:
Pakan Baru, Tenajan River, Soepadmo 252 (BO). - BORNEO. Sarawak: Bako NP,
Yap 527 (KEP); ibid., Ashton S 24320 (A, BO, K, L, SAN, SAR, SING); ibid., Kuswata
401 (BO, K, L, SING); Similajau FR, Brunig S 8637 (BO, L, SAN); Telok Belian,
Ilias Paie S 35997 (KEP, L); Kuching, Beccari PB 2688 (BO, K) (type). Brunei:
Andulau FR, Ashton BRUN 586 (KEP, L, SING). Kalimantan: East Kalimantan, Ulu
Mahakam, Sidiyasa 1653 (BO, KEP, L, SAN, WAN). Central Kalimantan, Sampit
River, near Kuala Kuajan, Kostermans 8070 (BO).
2. Coelostegia chartacea Soegeng, Reinwardtia 5, 3 (1960) 273; Argent et al. (eds),
Man. Non-Dipt. Trees Centr. Kalimantan | (1997) 97. TYPE: Kostermans 5262,
Indonesia, East Kalimantan, East Kutei, Sangkulirang, Menubar R. (holo BO! iso A,
K, L! LAE, P, PNH, SING!).
Distribution. Endemic in Borneo (Sabah, Sarawak, Kalimantan).
Ecology. In primary forest, on hill and riversides at 25—606 m altitude.
Notes. A very distinct species that can be distinguished from the other species of
Coelostegia by its chartaceous leaves. Morphologically, C. chartacea 1s closely related
to C. borneensis but differs in its longer leaf acumen and the ellipsoid fruit (see note
on C. borneensis).
Specimens examined: BORNEO. Sabah: Lung Manis FR, Charington SAN 24731 (K,
SAN); Tankong, Lassan SAN 72805 (SAN); Sungai Beatrice, Cockburn SAN 84979 (K,
SAN); Sungai Bole, Lee SAN 96767 (SAN); Ulu Segama, Jamin SAN 98872 (SAN).
Sarawak: Samunsam Wildlife Sanctuary, Abang Mohtar S 52657 (KEP, SAR); Bako
NP, Nadiah et al. S 100582 (KEP). Kalimantan: East Kalimantan, Desa Gong Solok,
Arifin AA 3010 (BO, WAN); Sungai Menubar, Kostermans 5262 (BO, L, SING) (type);
Belajan River near Tabang, Kostermans 10679 (K); Tidoengsche Landen, bb. 17958
(BO, BZE, L).
3. Coelostegia griffithii Benth., Gen. Pl. 1 (1862) 213; Hooker f., Fl. Brit. India 1
(1875) 353; Masters, J. Linn. Soc. Bot. 14 (1875) 505; Beccari, Malesia 3 (1889)
270; King, J. As. Soc. Beng. 60, 1 (1891) 57; Schumann in Engler & Prantl., Nat.
Pflanzenfam. 3, 6 (1895) 68; Ridley, Fl. Malay Penins. 1 (1922) 266; Bakhuizen f.,
Bull. Jard. Bot. Buitenz. 6, 3 (1924) 224; Soegeng, Reinwardtia 5, 3 (1960) 274;
Kochummen, Tree Fl. Malaya | (1972) 106; Cockburn, Trees of Sabah 1 (1976) 23;
Anderson, Check]. Trees Sarawak (1980) 153; Ashton, Man. Non-Dipt. Trees Sarawak
2 (1988) 54; Turner, Gard. Bull. Sing. 47, 1 (1995) 151; Coode et al. (eds), Checkl.
Flow. Pl. Gymno. Brunei (1996) 41; Beaman et al., Pl. Mount Kinabalu 4 (2001) 164.
TYPE: Griffith 547, Malaya, Malacca (holo K! iso A, L! P).
Coelostegia in Borneo 129
Coelostegia sumatrana Becc., Malesia 3 (1889) 271: Bakhuizen f., Bull. Jard. Bot.
Buitenz. 6, 3 (1924) 224: Coelostegia griffithii Benth. forma suwmatrana (Becc.)
Bakhuizen f., Bull. Jard. Bot. Buitenz. 6, 3 (1924) 248. TYPE: Beccari PS 738, West
Sumatra, Padang Prov., Air Manchur (holo FI, iso BO! K! L!).
Distribution. Sumatra (including Riau Archipelago), Peninsular Malaysia, Singapore,
Java and Borneo (Sabah, Sarawak, Brunei) (Fig. 1).
Ecology. In mixed dipterocarp, kerangas and lower montane forests at 15—1393 m.
Notes. Soegeng (1960) cited C. griffithii as occurring only in Peninsular Malaysia,
Sumatra and Bangka. Detailed comparative study of specimens currently available at
BO, K, L, SAN and SAR herbaria show that the species also occurs in Borneo.
In Borneo, sterile specimens of C. griffithii can be easily confused with those
of C. kostermansii, C. neesiocarpa and C. montana. However, the fruit surface of C.
griffithii is typically covered with sharp conical spines compared to that of the other
three species which have a smooth or submuricate or muricate surface.
Specimens examined (* denotes new records in Borneo; ** denotes additional
localities in Peninsular Malaysia): PENINSULAR MALAYSIA. Johor: Bukit Paloh
Estate, Mohd Shah MS 395 (BO, SAR, SING); Labis FR, Whitmore FRI 3847 (KEP):
Banang FR, Suleiman KEP 70172 (KEP). **Kedah: Gunung Inas FR, Whitmore FRI
4694 (KEP). Kelantan: Kemahang FR, Chelliah FRI 6502 (K, KEP, L); Kuala Balak,
Suppiah FRI 28017 (K, KEP, L); Temangan, Baki KEP 68766 (KEP). Malacca: loc.
not. indicated, Derry 123 (SING); ibid., Griffith 547 (K, L) (type); Bukit China, Derry
95 (SING); Selandar, A/vins s.n. (SING). Negeri Sembilan: Senawang FR, Yakim FMS
518 (K, KEP, SING); Sendayan FR, FG Din 536 (BO, SING); Pasir Panjang, Yusop
FMS 4222 (KEP, SING); Gunung Angsi, Zainuddin FRI 14591 (K, KEP, L, SING);
Pasoh FR, Nadiah et al. FRI 53951 (A, K, KEP, L, SAN, SAR, SING). **Pahang:
Lesong FR, Whitmore FRI 15851 (KEP); Rompin, Ng FRI 22992 (KEP), Ng FRI
22921 (KEP). Perak: loc. not. indicated, Scortechini 1862 (SING); ibid., Scortechini
1863 (SING); Selama, Mat Said FMS 1250 (KEP): Chikus FR, Speldenwinde 5366
(KEP); Changkat Jong FR, Ng FRI 5644 (KEP); ibid., Ng FRI 5878 (KEP, L); Bubu
FR, Selvaraj FRI 11154 (KEP, L); ibid., Suppiah FRI 11675 (KEP): ibid., Abdul Rahim
KEP 86060 (KEP); Trong, Everett FRI 13987 (K, KEP, L, SAN, SING); Teluk Intan,
Mohd Haniff SFN 14315 (SING); Bintang Hyau FR, Kamarudin FRI 34556 (K, KEP,
SAN, SAR). Selangor: Sungai Buloh FR, Hamid FMS 1183 (KEP); ibid., Strugnell
FMS 7068 (KEP, SING); ibid., Kiai FMS 8387 (KEP); ibid., Foxworthy FMS 10213
(KEP); ibid., Jamaat FMS 1531] (KEP); ibid., DFO Klang FMS 18715 (KEP); ibid.,
Strugnell 23931 (KEP); ibid., Symington FMS 24445 (KEP); ibid., Strugnell FMS
27880 (KEP); ibid., Jamaat FMS 44944 (KEP); ibid., Jamaat FMS 45002 (KEP):
Bukit Cherakah FR, Abu Amin FMS 1872] (KEP); Forest Research Institute Malaysia,
Ng FRI 33540 (KEP); ibid., Motan KEP 94744 (K, KEP, L, SING). **Terengganu:
Gunung Tebu FR, Zainuddin FRI 17922 (K, KEP). SINGAPORE. Botanic Gardens,
130 Gard. Bull. Singapore 63(1 & 2) 2011
Ridley 3887 (K, SING); ibid., Mat s.n. (SING); Bukit Mandai, Corner s.n. (SING),
Bukit Timah, Ridley 4738 (SING); Mandai Rd., Kiah SFN 37112 (BO, KEP, SING).
JAVA. Jakarta, cultivated in garden, van Steenis 3105 (BO). — SUMATRA. North
Sumatra: Atjeh, Boschbouwproefstation bb. 8873 (BO). South Sumatra: Belinju,
Grashoff 48 (BO, L); Bajunglentjir, Endert 276 (BO, L); ibid., Grashoff 812 (BO, L);
ibid., Endert 85E. 1P. 754 (BO, BZF, K, L); 1bid., Boschbouwproefstation 1. PT. 788
(BO, L); Rawas, Grashoff 1110 (BO, L). East Sumatra: Indrapura, Volke 5 (BO, L);
Jambi, Roos TFB 2055 (L); Bandar Poelau, Yates 2586 (K, L); Badjalinggi, Lorzing
7397 (BO); Muarapantai, Mo/ 23859 (BO, BZF, L); Indragini, Buwalda bb. 30081
(BO, BZF, L); Sungai Missingit, Beguin 556 (BO, L). West Sumatra: Balaiselasa,
Boschbouwproefstation bb. 5969 (BO, L); Pariaman, Boschbouwproefstation bb.
6736 (BO, L); Ophir, Neth. Ind. For. Service bb. 19481 (BO, BZF, L, SING); ibid.,
Djabar bb. 19629 (BO, BZF, L); Pengkalan Tapus, de Haan bb. 29537 (BO, BZEF, L);
Malintang, Korthals s.n. (L); Between Bondjol-Lubuk Sikapang, 7Jeijsmann s.n. (BO);
Painan, Boschbouwproefstation S.W.K./1—32 (BO, BZF, L); Air Manchur, Beccari PS
620 (L); ibid., Beccari PS 738 (BO, K, L) (type of C. sumatrana Becc.). — *BORNEO.
Sabah: Kundasang, Singh SAN 27495 (L, SAN); ibid., Meijer SAN 37996 (SAN);
ibid., Fosberg SAN 44135 (L); Sosopodon, Lajangah SAN 33145 (SAN); ibid., Mikil
SAN 38516 (K, L, SAN); ibid., Mikil SAN 46782 (K, SAN); ibid., Sinanggul SAN
47979 (SAN); Sunsuron, Phillips SAN 89353 (SAN). Sarawak: Tg. Long Amok, Rena
George S 43060 (K, L, SAR); Lambir Hills NP, Nadiah et al. S 100573 (KEP, SAR,
SING). Brunei: River Ingei, Wong WKM 607 (K, KEP, L, SAN); Labi Hills FR, Coode
et al. 6826 (K); Bukit Teraja, Niga BRUN 15094 (SING); Pendayan FR, Wvatt-Smith
KEP 80130 (KEP); Bukit Biang, Ashton BRUN 5584A (BO, K, KEP, L, SAR, SING).
4. Coelostegia kostermansii Soegeng, Reinwardtia 5, 3 (1960) 277; Argent et al. (eds),
Man. Non-Dipt. Trees Centr. Kalimantan | (1997) 97. TYPE: Kostermans 12548,
Indonesia, East Kalimantan, West Kutei, Tudjung Plateau, Mt. Maranga (holo BO! iso
A, CANB; K! KEP! L! NY; P):
Distribution. Endemic in Borneo (confined to Kalimantan).
Ecology. Primary forest on sandy loam soil, at 100-250 m.
Notes. C. kostermansii is morphologically very similar to C. neesiocarpa but
consistently differs in having a rough-surface and distinctly 5-angled fruit (vs. smooth
and rounded), elongate-ovoid seed with a caruncle up to 0.7 cm long (vs. ovoid with a
caruncle up to 1.2 cm long), and slender slightly kneed petiole (vs. thick and strongly
kneed petiole).
Specimens examined: BORNEO. Kalimantan: West Kalimantan, Mt. Maranga,
Kostermans 12548 (BO, K, KEP, L) (type); Mt. Damus, Hallier 776 (BO). East
Coelostegia in Borneo 131
Kalimantan, Belajan River near Tabang, Kostermans 10583 (L); Tabang, Kostermans
10659 (L).
5. Coelostegia montana Sidiyasa, Blumea 46 (2001) 165. TYPE: Sidivasa & Arifin
1529, Indonesia, East Kalimantan, Bulungan District, Kayan Mentarang National
Park, Gunung Lunjut (holo WAN, iso BO!, K, L!).
Distribution. Endemic in Borneo (confined to Sabah, Sarawak and Kalimantan) (Fig.
1).
Ecology. In dipterocarp and submontane forests on well-drained ridges, on igneous
(andesitic) derived soils, at 884-1450 m altitude.
Notes. Sidiyasa (2001) described C. montana based on fruiting specimens from
Sarawak (Anderson S 2846/) and fruiting specimens with young flower buds from
Kalimantan (Sidiyasa & Arifin 1529). The recently collected specimens from Sabah
(Nadiah et al. SAN 149577) bearing matured fruits and fully develop flowers represent
a new record of this species for the state, thus extending its distribution in Borneo.
Coelostegia montana is closely related to C. kostermansii but can be
distinguished by its c. 7-8 pairs of leaf lateral veins (vs. c. 9-13 pairs), narrowly
obovate stipules, c. 6 mm long (vs. lanceolate stipules, c. 4 mm long), depressed
conical and apically lobed flower buds c. 3 mm in diameter (vs. apiculate flower buds,
up to 2 mm), and dark blue, subglobose fruits with rounded base (vs. yellowish green,
ovoid fruits that are distinctly 5-angled at base).
Specimens examined: BORNEO. Sabah: Tambunan district, Rafflesia trail, Nadiah
et al. SAN 149577 (KEP, SAN). Sarawak: Kapit, Sungai Balleh, Anderson S 28461
(BO, K, KEP, KLU, L, SAR). Kalimantan: East Kalimantan, Kayan Mentarang NP,
Gunung Lunjut, Sidivasa & Arifin 1529 (BO, K, L, WAN) (type).
6. Coelostegia neesiocarpa Soegeng, Reinwardtia 5, 3 (1960) 279; Anderson, Checkl.
Trees Sarawak (1980) 153; Ashton, Man. Non-Dipt. Trees Sarawak 2 (1988) 56; Argent
et al. (eds), Man. Non-Dipt. Trees Centr. Kalimantan | (1997) 97. TYPE: de Zwaan
bb. 11288, Indonesia, East Kalimantan, Bulungan, Rumah R. (holo BO! iso BZF!).
Distribution. Endemic in Borneo (confined to Sarawak and Kalimantan).
Ecology. In lowland forest at 100-300 m altitude, growing on dacite-derived alluvial
fans in damp sandy valleys.
Notes. Coelostegia neesiocarpa differs from the other species in the genus in having
an elliptic-ovate, coriaceous, concolorous leaves with rounded base; thick and strongly
132 Gard. Bull. Singapore 63(1 & 2) 2011
kneed petiole; subglobose fruits up to 14 cm long, 11 cm diameter, with a smooth
surface and rounded base; and ovoid seeds with a caruncle c. 1.2 cm long.
Specimens examined: BORNEO. Sarawak: Hose Mountain, Mujong, Batu Kapal,
Ashton S 21242 (BO, K, L, SAR, SING). Kalimantan: East Kalimantan, Salimbata,
Rumah R., de Zwaan bb. 11288 (BO, BZF) (type); Upper Mahakam, Henar bb. 20696
(BO, BZF).
ACKNOWLEDGEMENTS. We acknowledge the generosity of the directors, keepers and
curators of herbaria (BO, BZF, K, KEP, L, SAN, SAR, SING and the Kinabalu National Park)
for the loan of specimens and facilities rendered. This project was financially supported by
RM-9 grants (Vote. no. 20300202023).
References
Alverson, W.S., Karol, K.G., Baum, D.A., Chase, M.W., Swensen, S.M., McCourt,
R. & Systema, K.J. (1998) Circumscription of the Malvales and relationships to
other Rosidae: evidence from rbcL sequence data. Amer. J. Bot. 85(6): 876-887.
Alverson, W.S., Whitlock, B.A., Nyffeler, R., Bayer, C. & Baum, D.A. (1999)
Phylogeny of the core Malvales: evidence from ndhF sequence data. Amer. J.
Bot. 86(10): 1474-1486.
Baas, P. (1972) The vegetative anatomy of Kostermansia malayana Soegeng.
Reinwardtia 8(2): 335-344.
Baum, D.A., Alverson, W.S. & Nyffeler, R. (1998) A durian by any other name:
taxonomy and nomenclature of the core Malvales. Harvard Pap. Bot. 3(2): 315—
330.
Bayer, C. & Kubitzki, K. (2003) Malvaceae. In: Kubitzki, K. (ed) The Families and
Genera of Vascular Plants 5: 225-311.
Bayer, C., Fay, M.F., Bruijn, A.Y. de, Savolainen, V., Morton, C.M., Kubitzki, K.,
Alverson, W.S. & Chase, M.W. (1999) Support for an expanded family concept
of Malvaceae within a recircumscribed order Malvales: a combined analysis of
plastid afpB and rhcL DNA sequences. Bot. J. Linn. Soc. 129: 267-303.
Beccari, O. (1886) Le Bombacaceae Malesi descritte ed illustrate. Malesia 3(4): 202—
280, Tavola XII-XXXVI.
Bentham, G. & Hooker, J.D. (1862) Bombacaceae (Malvaceae Tribe Bombaceae).
Genera Plantarum |: 195-213.
Cheek, M. (2006) The validation of two new family names in Malvales: Durionaceae
and Brownlowiaceae. Kew Bull. 61: 443.
Cheek, M. (2007) Durionaceae. In: Heywood, V.H. et al. (eds) Flowering Plants
Families of the World, p. 134.
Cronquist, A. (1968) The Evolution and Classification of Flowering Plants. London
and Edinburgh: Thomas Nelson & Sons Ltd.
Coelostegia in Borneo 133
Cronquist, A. (1981) An Integrated System of Classification of Flowering Plants. New
York: Columbia University Press.
Hutchinson, J. (1959) The Families of Flowering Plants Volume 1—Dicotyledons.
London: Oxford University Press.
Keng, H. (1969) Orders and Families of Malayan Seed Plants. Singapore: Singapore
University Press.
Kostermans, A.J.G.H. (1958) The genus Durio Adans. (Bombac.). Reinwardtia 4(3):
47-153.
Masliya, M. (2008) Kajian Epidermis Daun ke Atas Beberapa Spesies Bombax,
Durio dan Kostermansia (Bombacaceae). B.Sc. Thesis, Universiti Kebangsaan
Malaysia.
Nilsson, S. & Robyns, A. (1986) Bombacaceae. World Pollen and Spore Flora 14.
Stockholm: The Almqvist & Wiksell Periodical Company.
Nyffeler, R. & Baum, D.A. (2000) Phylogenetic relationships of the durians
(Bombacaceae-Durioneae or /Malvaceae/Helicteroideae/Durioneae) based on
chloroplast nuclear ribosomal DNA sequences. Plant Syst. Evol. 224: 55-82.
Salma, I. (1999) The taxonomic significance of trichome morphology in the genus
Durio (Bombacaceae). Gard. Bull. Singapore 51: 55—70.
Salma, I. (2000) The significance of pollen morphology in the taxonomy of the genus
Durio (Bombacaceae). Gard. Bull. Singapore 52: 261-271.
Salmizawati, M.S. (2008) Kajian Anatomi dan Mikromorfologi Epidermis Daun
Terhadap Beberapa Spesies Neesia Bl. (Bombacaceae). B.Sc. Thesis, Universiti
Kebangsaan Malaysia.
Sidiyasa, K. (2001) Coelostegia montana, a new species of Bombacaceae from
Borneo. Blumea 46: 165-168.
Siti Fatimah, S. (2008) Kajian Morfologi Polen ke Atas Beberapa Spesies Daripada
Genus Durio dan Neesia (Bombacaeae). B.Sc. Thesis, Universiti Kebangsaan
Malaysia.
Soegeng, R.W. (1959) Kostermansia Soegeng, a new genus in Bombacaceae
(Durioneae). Reinwardtia 5(1): 1-9.
Soegeng, R.W. (1960) The genus Coelostegia Benth. (Bombac.). Reinwardtia 5(3):
269-291.
Soepadmo, E. (1961) A monograph of the genus Neesia Blume (Bombacaceae).
Reinwardtia 5(4): 481—508.
Takhtajan, A. (1969) Flowering Plants—Origin and Dispersal. Edinburgh: Oliver &
Boyd Ltd.
Tan, L.H. (2008) Kajian Epidermis Daun ke Atas Beberapa Spesies Daripada Genus
Coelostegia dan Ceiba (Bombacaceae). B.Sc. Thesis, Universiti Kebangsaan
Malaysia.
Webster, G.L., Del-Arco-Aguilar, M.J. & Smith, B.A. (1996) Systematic distribution
of foliar trichome types in Croton (Euphorbiaceae). Bot. J. Linn. Soc. 121: 41-57.
134 Gard. Bull. Singapore 63(1 & 2) 2011
Appendix A. Morphological characters distinguishing the genera of the Durioneae-group in
Borneo: Coelostegia, Durio, Kostermansia and Neesia. Sources: Kostermans 1958; Soegeng
1959, 1960; Soepadmo 1961; Baas 1972; Nilsson & Robyns 1986; Webster et al. 1996; Salma
1999, 2000; Sidiyasa 2001; Masliya 2008; Salmizawati 2008; Siti Fatimah 2008; Tan 2008.
Characters
Coelostegia
Durio
Kostermansia
Neesia
Buttresses
Leaf size (cm)
Lower leaf
surface:
pubescence
Tertiary/ inter-
costal veins
Midrib
Epicalyx
Calyx-lobes/
sepals
Corolla/ petals
Stamens (filament
& anther)
Ovary
Style
Stigma
large, thin,
convex, spreading
(4.5—)6—12(—14.8)
x (1.1—-)2-5(-6)
scales only
indistinct,
reticulate
evident, flat or
raised above
reduced in size,
subtending calyx,
3-lobed at anthesis
induplicate-
saccate
shorter than
calyx, calyptrate,
perigynous
longer than ovary;
topped by three
1-celled anthers
partly embedded
in receptacle,
covered by peltate
scales
conspicuous,
filiform
discoid, peltate,
conspicuous
rounded, straight
to concave, not
spreading
3-425 x 3-15
simple & stellate
hairs, also scales
generally not
prominent,
reticulate
sunken or
channelled above
completely
enveloping flower
bud, splitting into
2 lobes
not induplicate-
saccate
mostly showy,
longer than
calyx, free,
long-persistent,
hypogynous
longer than ovary;
each filament with
1—many unilocular
anthers
superior, covered
by peltate scales
& stellate hairs
well developed
small, capitellate
up to 7 m high,
plank-like,
spreading
(6—)9—-13(-19) x
(2-)4-6(-9.5)
scales only
prominent,
reticulate
strongly
prominent above
partly enveloping
flower bud,
splitting into 2
lobes
not induplicate-
saccate
shorter than
calyx, not showy,
free, caducous,
hypogynous
shorter than ovary;
topped by two
bean-shaped,
basifixed, 2-celled
anthers
superior, covered
by peltate scales
reduced, or very
short, thick
large, convex,
discoid, peltate
large, thin,
convex, spreading
6-60 = 3-25
stellate hairs;
rarely sparse
minute long-
fimbriate scales
distinct, reticulate
obscure, depressed
above
completely
enveloping flower
bud, splitting into
2-5 lobes
not induplicate-
saccate
shorter than calyx,
hypogynous,
calyptrate
longer than ovary;
topped by one
2-celled anther
superior, covered
by hirsute, stellate
hairs
short, conical or
filiform
small, capitellate,
round or
subpentagonous
Coelostegia in Borneo
Fruit surface
Fruit-valves inside
Fruit dehiscence
Aril or caruncle
Seeds
Cotyledons
Pollen type
Type of apertures
Pollen shape
Ornamentation of
exine
Trichomes (on
lower leaf surface)
Stomata type
Stomata on lower
leaf surface
spiny to muricate
or smooth
glabrous
dehiscing to c.
1/2-1/3 of its
length (valves
split while fruit
is still attached
to tree, becoming
erect or reflexed)
basal caruncle
present
smooth, in 2 rows
in each locule
thin, foliaceous,
covered by 2 flat-
convex lobes of
the endosperm
Durio-type
3-colporate, short
narrow colpus
prolate-spheroidal
microreticulate to
smooth
dentate-peltate
scales & glandular
hairs
anisocytic,
tetracytic
randomly arranged
covered with
slender or stout
spines
glabrous
some fruits do
not dehisce, or
dehisce to the very
base (generally
dehiscent only
after falling to the
ground)
aril absent or
present and
covering half to
whole seed
ellipsoid, in 2
rows in each
locule; large, pale
brown to reddish
/ black
thick, flat-convex,
fleshy; endosperm
absent
Durio-type
3-colporate, short
broad colpus
oblate-spheroidal
smooth
stellate & dentate
peltate scales;
appressed stellate
& simple hairs;
glandular hairs
paracytic
randomly arranged
densely spiny
glabrous
dehiscing to the
base (valves split
while fruit is still
attached to tree,
becoming erect or
reflexed)
no aril / caruncle
few, large, glossy,
dark brown (white
when fresh)
foliaceous, flat,
covered by 2
partite endosperm
Kostermansia-
type
3-colporate, long
narrow colpus
oblate-spheroidal
microreticulate
subentire peltate
scales & glandular
hairs
amfiparacytic
in circles around
trichome bases
135
muricate or short-
spiny
with dense
brownish, hirsute,
prurient hairs
dehiscing to
c. 1/2-1/3 of
its length to +
completely (valves
split while fruit
is still attached
to tree, becoming
erect or reflexed)
basal caruncle
present
ellipsoid, smooth
foliaceous,
enveloped by 2
flat-convex lobes
of the endosperm
Durio-type
3-colporate, short
narrow colpus
prolate-spheroidal
microreticulate
appressed stellate,
dendritic, simple
& glandular hairs
anisocytic,
tetracytic
randomly arranged
‘ei
‘
Gardens’ Bulletin Singapore 63(1 & 2): 137-144. 2011 137
A synopsis of Jarandersonia (Malvaceae:
Brownlowioideae)
H.S. Tan, R.C.K. Chung' and E. Soepadmo
Forest Research Institute Malaysia, 52109 Kepong, Selangor, Malaysia
'richard@frim.gov.my
ABSTRACT. A revision of Jarandersonia was conducted as part of a study of Malvaceae:
Brownlowioideae for the Tree Flora of Sabah and Sarawak Project. Six species of Jarandersonia
are recognised for Borneo, of which J. pentaceoides R.C.K.Chung & H.S.Tan, endemic
to central Kalimantan, is new to science. A complete list of exsiccatae, nomenclatural and
taxonomic notes, geographical distribution and conservation status of the recognised species,
are provided.
Keywords. Borneo, Brownlowioideae, Jarandersonia, Kalimantan, Malvaceae, tree flora
Introduction
Jarandersonia was first described by Kostermans in 1960 based on the Sarawak
species, J. paludosa (now a synonym of J. purseglovei (Kosterm.) Kosterm.). In
the past fifty years, the genus had been included in the family Tiliaceae. Recently,
molecular evidence has supported the inclusion of Jarandersonia in the subfamily
Brownlowioideae of the expanded family Malvaceae s./. based on ndhF, atpB and
rbcL data (Alverson et al. 1999; Bayer et al. 1999; Nyffeler and Baum 2000; Bayer and
Kubitzki 2003). In Malaysia, Brownlowioideae consists of five genera, namely Berrva,
Brownlowia, Diplodiscus, Jarandersonia and Pentace, and has about 55 species. The
genus Jarandersonia can be easily distinguished from Brownlowia, Diplodiscus and
Pentace by its elliptic or obovate leaves, densely hairy fruit spines and a combination
of micromorphological characters.
Jarandersonia, locally known as baru baran (Iban) in Sarawak, was named
after J.A.R. Anderson, a forest botanist working in Sarawak and Brunei from 1951
to 1970. Kostermans (1962, 1970) described three more species and transferred
Brownlowia clemensiae Burret to the genus, making a total of five known species of
Jarandersonia. In our recent revision of the genus for the Tree Flora of Sabah and
Sarawak Project, an additional new species, J. pentaceoides R.C.K.Chung & H.S.Tan,
from central Kalimantan was discovered. In an assessment of the conservation status
of the five species, one was found to be critically endangered, two endangered, one
vulnerable, and one of least concern.
138 Gard. Bull. Singapore 63(1 & 2) 2011
Synopsis of species
Jarandersonia Kosterm., Reinwardtia 5 (1960) 319, Reinwardtia 8 (1970) 17;
Hutchinson, Gen. Flow. Pl. 2 (1967) 491; Ashton, Man. Non-Dipt. Trees Sarawak
2 (1988) 448; Bayer & Kubitzki in Kubitzki (ed.), Fam. Gen. Vasc. Pl. 5 (2003)
258: LaFrankie, Trees of Tropical Asia (2010) 479. TYPE SPECIES: Jarandersonia
paludosa Kosterm. [= Jarandersonia purseglovei (Kosterm.) Kosterm.].
Vernacular name. Baru baran (Iban, Sarawak).
Distribution. The genus comprises six species endemic to Borneo (2 are in western
Sarawak, | Sabah, | Sarawak and Kalimantan, | Sarawak, Brunei and west Kalimantan,
1 central Kalimantan). (Fig. 1)
Ecology. Mainly found in mixed dipterocarp forest and peat swamp forest.
Notes. The terminology used for trichome types mainly follows Webster et al. (1996).
Three main trichome types were observed in Jarandersonia, namely, subentire-
lepidote (radii webbed 80—100%), dentate-lepidote (radii webbed 50-80%) and
stellate-lepidote (radii webbed 30-50%).
110°O'E 110°0'E 112°0'E 114°0'E 116°O'E 118°0'E
ex
SABAH {37%
2°0'N
2°0'N
0°0
110°0'E
"Legend
Jarandersonia clemensiae
2°0'S
Jarandersonia parvifolia
Jarandersonia pentaceoides
Jarandersonia purseglovei
a4 * DP
40'S
Jarandersonia rinoreoides
Jarandersonia spinulosa
110°0'E 112°0'E 114°0'E 116°0'E 118°0'E
Fig. 1. Known distribution of Jarandersonia species in Borneo (right). Species distributions in
the south-western part of Sarawak are shown in magnified view (left).
Jarandersonia in Borneo 139
1. Jarandersonia clemensiae (Burret) Kosterm., Reinwardtia 8 (1970) 18; Anderson,
Check]. Trees Sarawak (1980) 339; Ashton, Man. Non-Dipt. Trees Sarawak 2 (1988)
448; Whitmore et al., Tr. Fl. Indonesia Checkl. Kalimantan 2, 1 (1990) 356; Rantai
& Chai, New Checkl. Trees Sarawak (2007) 333. Brownlowia clemensiae Burret,
Notizbl. Bot. Gart. Berl.-Dahl. 13 (1936) 252: Kostermans, Communic. For. Res. Inst.
Bogor 73 (1961) 29, fig. 23; Ashton, Man. Non-Dipt. Trees Sarawak 2 (1988) 432.
TYPE: J. Clemens & M.S. Clemens 22202, Borneo, Sarawak, Mt Gading (holo B7; iso
BO, K, NY barcode 00415382, SAR). (Fig. 2A)
Vernacular names. Baru baran (Iban), nginjoaja (Bidayuh Padawan, Sarawak).
Distribution. Endemic to Borneo (Sarawak and Kalimantan). In Sarawak, known from
Lundu, Kuching, Bau, and Serian districts. Also occurring in Kalimantan.
Conservation status. Least concern. Most of the populations are found in protected
areas.
Ecology. In lowland mixed dipterocarp forest, in riparian and secondary forests. Very
local, usually along small stream banks on clay-rich alluvium, to 200 m altitude.
Notes. Jarandersonia clemensiae is closely related to J. purseglovei. The former,
however, differs from the latter in having a subcordate leaf base (vs. acute to rounded),
12—16 pairs of lateral veins (vs. 18—23 pairs), flattened lateral veins above (vs. sunken
above), rounded midrib (vs. square) and unbranched fruit spines with stellate-lepidote
and dentate-lepidote trichomes (vs. short-branched with branch tips bearing simple or
2-armed setose hairs).
2. Jarandersonia parvifolia Kosterm., Reinwardtia 8 (1970) 18; Anderson, Checkl.
Trees Sarawak (1980) 339; Ashton, Man. Non-Dipt. Trees Sarawak 2 (1988) 448;
Whitmore et al., Tr. Fl. Indonesia Check]. Kalimantan 2, | (1990) 356; Rantai & Chai,
New Checkl. Trees Sarawak (2007) 333. TYPE: /lias S 15561, Borneo, Sarawak,
Bintulu, Segan FR (holo BO; iso A, SAN, SAR, SING). (Fig. 2B)
Vernacular name. Baru baran mit (Iban, Sarawak).
Distribution. Endemic to Borneo (Sarawak, Brunei and Kalimantan). In Sarawak,
known from Bintulu, Tatau and Kapit districts. Also occurring in Brunei and
Kalimantan.
Conservation status. Critically endangered A2c. Two localities had been converted to
oil palm plantations. Details for the remaining localities are scanty.
140 Gard. Bull. Singapore 63(1 & 2) 2011
gay
Ecology. In mixed dipterocarp forest, on leached yellow clay soils, to 700 m altitude.
Notes. Jarandersonia parvifolia can be easily distinguished by its small elliptic to
broadly elliptic leaves, 4-11.5 x (1.4-)2—5 cm and square midrib in cross-section.
3. Jarandersonia pentaceoides R.C.K.Chung & H.S.Tan, Syst. Bot. (in press).
Proposed Type: J.K. Jarvie & A. Ruskandi 5769, Borneo, Kalimantan, Kalimantan
Tengah, Samba (holo KEP; iso A, BO, K, SAN, SING). (Fig. 2C)
Distribution. Known only from the type, recorded in Samba, central Kalimantan.
Ecology. Primary forest, hilly terrain on red clay soil with slope, at attitudes to 300 m.
Notes. Jarandersonia pentaceoides 1s most similar to /. rinoreoides, having cuneate leaf
base, rounded midrib, slender and unbranched fruit spines. However, J. pentaceoides
is distinct in having densely stellate-lepidote and dentate-lepidote trichomes along the
edges of tertiary veins and honeycomb-like quaternary veins on the lower leaf surface,
a kneed petiole that is swollen at both ends, and a sparsely stellate-lepidote and tufted
hairy seed coat, whereas in J. rinoreoides the quaternary veins are absent, the petiole
is straight and not swollen at both ends and the seed coat is glabrous. The leaves
of J. pentaceoides are similar to those of some common Pentace and Mallotus (1.e.,
M. leucodermis Hook.f. and M. muticus (Mill.Arg.) Airy Shaw) in shape, number
of secondary veins, and its kneed petiole that is swollen at both ends. However, the
morphological characters of J. pentaceoides do not match the features of both Pentace
and Mallotus, because the new species have distinct quaternary veins (vs. absent in
both Pentace and Mallotus) and spiny fruits (vs. winged fruits in Pentace and shortly
stiff spiny in Mal/otus). The fruit spines of J. pentaceoides are similar to Commersonia
bartramia (L.) Merr. but the former can be easily distinguished by its cuneate leaf base
(vs. cordate) and indehiscent fruit (vs. splitting open fruit into 5 valves).
4. Jarandersonia purseglovei (Kosterm.) Kosterm., Reinwardtia 6 (1962) 299;
Anderson, Checkl. Trees Sarawak (1980) 339; Ashton, Man. Non-Dipt. Trees Sarawak
2 (1988) 449: Whitmore et al., Tr. Fl. Indonesia Checkl. Kalimantan 2, 1 (1990) 356;
Ranta1 & Chai, New Checkl. Trees Sarawak (2007) 333. Brownlowia purseglovei
Kosterm., Gard. Bull. Singapore 17 (1958) 1, Communic. For. Res. Inst. Bogor 73
(1961) 28, fig. 22. TYPE: J.N. Purseglove P 4662, Borneo, Sarawak, Mt Pueh (holo
SING barcode 0050678; iso BO). (Fig. 2D)
Jarandersonia paludosa Kosterm., Reinwardtia 5 (1960) 319. TYPE: J.4.R. Anderson
S 6554, Borneo, Sarawak, Lundu district (holo K; iso BO, L, SAR, SING barcode
0050679).
Jarandersonia in Borneo 14]
Vernacular names. Baru baran kasar, baru barun daun kasar (Iban, Sarawak).
Distribution. Endemic to Borneo (W Sarawak). Recorded from Lundu, Kuching and
Bau districts.
Conservation status. Endangered A2c, due to loss of natural habitat and its population
size having been reduced by more than 50%.
Ecology. Lowland mixed dipterocarp forest, on shallow peat and ground-water podsols
near sea-level.
Notes. The leaves of Jarandersonia purseglovei are very similar to those of Durio
oblongus Mast. (Malvaceae: Helicterioideae/Durioneae). However, the species can be
distinguished from D. oblongus by the cuneate leaf base (vs. rounded), rounded midrib
(vs. ridged) and prominent intermediate veins (vs. inconspicuous). This species is also
closely related to J. clemensiae (see note under J. clemensiae).
5. Jarandersonia rinoreoides Kosterm., Reinwardtia 8 (1970) 17; Whitmore et al..
Tr. Fl. Indonesia Check]. Kalimantan 2, 1 (1990) 356. TYPE: W. Meijer SAN 27885,
Borneo, Sabah, Sandakan, Tabin, W of Sulap (holo SAR; iso K, KEP barcode 76031,
L, SAN). (Fig. 2E)
Distribution. Endemic to Borneo (Sabah). Recorded from Labuk Sugut, Sandakan,
Kinabatangan and Lahad Datu districts.
Conservation status. Vulnerable B lab(ii1). The species is found in less than 10 locations
that are not strictly protected.
Ecology. In mixed dipterocarp forest, usually near streams, to 700 m altitude.
Notes. Jarandersonia rinoreoides is closely related to J. spinulosa but differs in its
midrib and lateral veins (glabrous vs. pilose hairy above), and its fruit spines (slender,
tufted hairy intermixed with simple or 2(—3)-armed setose hairs, unbranched vs. stout,
glabrous, branched with branch tips bearing simple or 2-armed setose hairs).
6. Jarandersonia spinulosa Kosterm., Reinwardtia 6 (1962) 300; Anderson, Checkl.
Trees Sarawak (1980) 339; Ashton, Man. Non-Dipt. Trees Sarawak 2 (1988) 449:
Whitmore et al., Tr. Fl. Indonesia Checkl. Kalimantan 2, 1 (1990) 357: Rantai & Chai,
New Checkl. Trees Sarawak (2007) 333. TYPE: H.N. Ridley s.n., Borneo, Sarawak,
Mt Matang (holo K). (Fig. 2F)
142 Gard. Bull. Singapore 63(1 & 2) 2011
Distribution. Endemic to Borneo (Sarawak). In Sarawak, known from Kuching and
Bau districts.
Conservation status. Endangered B2ab(i11). The species is known from only two
specimens with the latest collection made in 1985. No populations are known to occur
in the network of protected areas and some of the forest reserves where this species
occurs are small and fragmented. No ecological details were available.
Ecology. Lowland mixed dipterocarp forest, along rivers, to 50 m altitude.
Note. This species is closely related to J. rinoreoides (See note under J. rinoreoides).
Identification list
Numbers after the collector numbers refer to the following Jarandersonia species:
1 = J. clemensiae; 2 = J. parvifolia, 3 = J. pentaceoides; 4 = J. purseglovei; 5 = J.
rinoreoides; 6 = J. spinulosa. When collection numbers are not available, dates are
given within brackets.
Aban SAN 97250: 5; Anderson S 25415: 1; S 26759: 1; S 6554: 4; Awang Enjah S
68042: 1. Buxton A 548: 1; Bojeng S 9355: 1. Clemens & Clemens 22202: 1. Dewol
SAN 99462: 5. Frodin & Othman 2027: 1; Fordin et al. 2119: 1; Fuchs 21376: 2.
Guijing SAN 45510: 5; SAN 45511: 5; SAN 45518: 5. Ilias S 15561: 2. Jacobs 5567: 1;
Jaamat & Tachun FMS 39640: 2; James S 29846: 1; Jarvie & Ruskandi 5769: 3; Jong
904: 1; Jugah S 51590: 1. Lakising SAN 70168: 5. Meijer SAN 27885: 5; SAN 31016:
5; SAN 51237: 5; SAN 53222: 5; Madani SAN 61097: 5; Munting S 54250: 1. Othman
S 37050: 1; S 37820: 4; S 40042: 1; Othman et al. S 49899: 1; S 63818: 1; Othman
& Munting S 54348: 2; S 54349: 2. Purseglove P 4662: 4. Rantai et al. S 68437: 1;
Reto et al. 506: 1; 508: 1; 511: 4; Ridley s.n. (1/1915): 6. Sigin & Lidah SAN 97197:
5; Sinanggul SAN 57101: 5; Suah SAN 37379: 5. Tukirin & Partomihardjo K 3319: 2.
Yahud et al. S 93215: 1;.S 932162 13 Sv932182 VS 932232455 93224: ES 93226
S 93227: 1; S 93228: 42S 93229: 4: S93231- 12S 932637 lassi simi. 969) 2a
Sol 2636:
ACKNOWLEDGEMENTS. We are deeply indebted to the Malaysian Government especially
the Ministry of Science, Technology and Innovation (MOSTI) and the Ministry of Natural
Resources and Environment (NRE) for their generous financial support. We gratefully
acknowledge the support of the Director-General of FRIM and the Directors of the Sabah and
Sarawak Forestry Departments for their continuous guidance and encouragement. We thank
Wendy S.Y. Yong for preparing the geographical distribution map. Thanks are also due to the
keepers and curators of following herbaria: A, BM, BO, E, K, KEP, L, NY, SAN, SAR, and
SING for the loan of specimens and facilities provided.
143
Jarandersonia in Borneo
Fig. 2. Fruit spines of Jarandersonia species. A. J. clemensiae, spines stout, stellate-lepidote
and dentate-lepidote, unbranched. B. J. parvifolia, spines stout, tufted hairy, short-branched
with the branch tips bearing simple or 2(—3)-armed setose hairs. C. J. pentaceoides, spines
slender, stellate-lepidote with simple setose hairs, unbranched. D. J. purseglovei, spines stout,
tufted hairy, short-branched with the branch tips bearing simple or 2-armed setose hairs. E. J.
rinoreoides, spines slender, tufted hairy intermixed with simple or 2(—3)-armed setose hairs,
unbranched. F. J. spinulosa, spines stout, glabrous, short-branched with the branch tips bearing
simple or 2-armed setose hairs.
144 Gard. Bull. Singapore 63(1 & 2) 2011
References
Alverson, W.S., Whitlock, B.A., Nyffeler, R., Bayer, C. & Baum, D.A. (1999)
Phylogeny of the core Malvales: Evidence from ndhF sequence data. Amer. J.
Bot. 86(10): 1474-1486.
Bayer, C. & Kubitzki, K. (2003) Malvaceae. In: Kubitzki, K. & Bayer, C. (eds) The
Families and Genera of Vascular Plants 5: 225-311.
Bayer, C., Fay, M.F., de Bruijn, A.Y., Savolainen, V., Morton, C.M., Kubitzki, K.,
Alverson, W.S. & Chase, M.W. (1999) Support for an expanded family concept
of Malvaceae within recircumscribed order Malvales: a combined analysis of
plastid afpB and rbcL DNA sequences. Bot. J. Linn. Soc. 129: 267-303.
Chung, R.C.K., Tan, H.S. & Soepadmo, E. (in press.) A remarkable new species of
~ Jarandersonia (Malvaceae - Brownlowioideae) from central Kalimantan,
Borneo. Syst. Bot.
Kostermans, A.J.G.H. (1960) Jarandersonia Kosterm. A new Bornean genus of
Tiliaceae - Brownlowieae. Reinwardtia 5(3): 319-321.
Kostermans, A.J.G.H. (1961) A monograph of the genus Brownlowia Roxb. (Tiliaceae).
Commun. Forest Res. Inst., Bogor 73: \—62.
Kostermans, A.J.G.H. (1962) Miscellaneous Botanical Notes 4. Reinwardtia 6(3):
281-325.
Kostermans, A.J.G.H. (1970) The Genus Jarandersonia Kosterm. Reinwardtia 8(1):
17-20.
Nyffeler R. & Baum, D.A. (2000) Phylogenetic relationships of the durians
(Bombacaceae - Durioneae or Malvaceae - Helicterioideae - Durioneae) based
on chloroplast and nuclear ribosomal DNA sequence. P/. Syst. Evol. 224: 55-82.
Webster, G.L., Del-Arco-Aguilar, M.J. & Smith, B.A. (1996) Systematic distribution
of foliar trichomes types in Croton (Euphorbiaceae). Bot. J. Linn. Soc. 121:
41-57.
Gardens’ Bulletin Singapore 63(1 & 2): 145-153. 2011 145
Towards an account of Sapotaceae for Flora Malesiana
Peter Wilkie
Royal Botanic Garden Edinburgh, 20a Inverleith Row, Edinburgh, EH3 SLR, U.K.
p.wilkie@rbge.ac.uk
ABSTRACT. An overview of the pan-tropical family Sapotaceae is provided with particular
focus on the Malesian region. Past and current taxonomic and phylogenetic research is
summarised and publications relating to the production of a Flora Malesiana Sapotaceae
account highlighted. Challenges to delivering a Flora Malesiana account are identified and
some potential solutions suggested.
Keywords. Flora Malesiana, Sapotaceae, systematics, taxonomy
Introduction
Sapotaceae is a pantropical family of trees and shrubs and is composed of about 50
genera and 1000 species. In Malesia there are an estimated 15 genera and 300 species.
The family is ecologically important with representatives of the family common in
the forests of Malesia. They occur from beach forests at sea level to mossy montane
forests at over 4000 m altitude. The family is economically important and produces
the important heavy hardwood timber, Bitis (mainly Madhuca utilis (Ridl.) H.J.Lam,
Palaquium ridleyi King & Gamble and Palaquium stellatum King & Gamble) and the
light to medium hardwood, Nyatoh, from many other species (Ng 1972). The family
also produces edible fruit with Manilkara zapota (L.) P.Royen (sapodilla plum, cikz),
and Chrysophyllum cainito L. (star apple), introduced from Central America, the most
widely cultivated. The latex produced from Palaquium gutta (Hook.) Burck (Gutta
Percha) has been used in the insulation of cables, golf balls and in root fillings in
dentistry (Burkill 1966, Boer & Ella 2000).
It is a family which has historically been acknowledged as being taxonomically
problematic (Pierre 1890; Baillon 1891; Dubard 1912, 1915; Lam 1939; Aubrévile
1964: Baehni 1965; Pennington 1991). Although species tend to be relatively well
defined, the genera are not. Estimates of number of genera range from 122 (Aubrévile
1964) to 53 (Pennington 1991). There is a high level of homoplasy in the family and
unique synapomorphies for genera are rare with most being distinguished on character-
state combinations.
The family can be quite readily identified in the field by the white exudate
(latex) produced from the cut bark and twigs, the spirally arranged leaves which are
usually crowded at the tips of twigs and the often coppery underside of leaves. The
flowers are in axillary fascicles usually behind the leaves and in fruit the calyx and
style are persistent. The calyx provides a good taxonomic character at generic and
146 Gard. Bull. Singapore 63(1 & 2) 2011
tribal levels and seed shape and the position and extent of the seed scar provide useful
characters at the generic and species level (Pennington 1991).
Monographic studies
Most monographic work on the family in Malesia was undertaken by researchers based
at the Rijksherbarium, Leiden from the 1930’s to early 1960’s. Over this period many
papers were published in Blumea as part of the series “Revision of the Sapotaceae of
The Malaysian Area in a Wider Sense”. \n these publications, the geographical area
covered was much larger than what we today term Malesia and included areas such as
the Solomon Islands, New Caledonia and Fiji.
The most important contributors include the director of the Rijksherbarium
between 1933 and 1962, Herman Johannes Lam. He published important revisions
of genera in the Sapotaceae including Manilkara (Lam 1941a) and Burckella (Lam
& Royen 1952a). He also produced family accounts for Sapotaceae of the Dutch
East Indies (Lam 1925, 1927), New Guinea (Lam 1932) and the Pacific Islands (Lam
1942), a paper on the Phylogenetics of Sapotaceae (Lam 1935) as well as several notes
on a range of other genera (Lam 1938, 1939, 1943a, 1943b, 1957). He also published
taxonomic accounts of the closely related family Sarcospermataceae including a
Revision of Sarcospermaceae (Lam & Varossieau 1938, 1939; Lam 1941b) and an
account of Sarcospemataceae for Flora Malesiana (Lam 1948).
A co-worker of Lam’s, Pieter van Royen, revised many large genera of
Sapotaceae including Planchonella (Royen 1957a), Palaquium (Royen 1960a) and
Madhuca(Royen 1960b). He also revised several other smaller genera such as Burckella
(Lam & Royen 1952a, 1957b; Royen 1959), Mimusops (Royen 1952), Manilkara
(Royen 1953, 1957c), Xantolis (Royen 1957d), Diploknema (Royen 1958a), Aulandra
(Royen 1958b), Eberhardtia (Royen 1960d) and Mastichodendron (Royen 1960c) and
compiled an account of Sapotaceae covering some 16 genera and 260 species for Flora
Malesiana which was never published (Royen, Unpub.).
Willem “Wim” Vink started to revise van Royen’s unpublished account but
due to the large number of new forestry collections coming from the Malesian region,
the many new species needing to be described and the many poorly known species
needing new description, this could not be brought to fruition within the year allocated
to the task. Several revisions of smaller genera such as Lepfostylis, Pycnandra and
Magodendron (Vink 1957) and Chrysophyllum (Vink 1958), however, were published
and important contributions continue to be made by him (Vink 1995, 2001, 2002). Other
Rijksherbarium researchers who published important works on Sapotaceae include
van den Assem who revised Ganua (Assem 1953, Assem & Kostermans 1954), Jeuken
who revised /sonandra (Jeuken 1952); Bruggen (1958a, 1958b) who revised Payena
and Aesandra, and Herrmann-Erlee who revised Krausella and Pouteria (Herrmann-
Erlee & Lam 1957, Herrmann-Erlee & Royen 1957).
One of the most important publications covering Sapotaceae throughout its
distribution is The Genera of Sapotaceae (Pennington 1991). Based on morphology,
Sapotaceae in the Malesian flora 147
this reviewed the whole family and brought much clarity to the genera. It recognised
53 genera and 5 tribes and considered Sarcosperma to be part of Sapotaceae (not
Sarcospermataceae). In Malesia, all 5 tribes are represented with Isonandreae containing
most Malesian genera. Based on this work, the World Checklist of Sapotaceae was
produced (Govaerts et al. 2001), as well as the related website (http://apps.kew.org/
wesp/home.do).
Molecular phylogenetic studies
The first large-scale molecular studies of Sapotaceae were produced by Anderberg
& Swenson (2003) and Swenson & Ardenberg (2005). Based on molecular and
morphological data, they proposed a new subfamily classification of Sapotaceae
with three subfamilies being recognised, Sarcospermatoideae, Sapotoideae and
Chrysophylloideae. Malesian genera are found in all three subfamilies. The sampling
from Malesia, however, was poor with only 5 taxa represented. The two largest genera
in Malesia, Madhuca and Palaquium had only a single species sampled and these were
from outside the Malesian region. Swenson and co-workers have published several
other important phylogenetic studies on the subfamily Chrysophylloideae especially
from New Caledonia (Bartish & Swenson 2005, Swenson & Bartish 2007, Swenson
& Munzinger 2007). Also researching the subfamily Chrysophylloideae, Triono et al.
(2007) produced a molecular phylogeny of Pouteria from Malesia and Australasia to
re-assesses the generic delimitation of Pouteria and its affinities with Planchonella.
This did not support the broad circumscription of Pouteria by Pennington (1991).
Smedmark & Anderberg (2006, 2007) published work on the subfamily
Sapotoideae and provided a useful backbone to research on the subfamily but again
few samples were included from the Malesian region. This is being addressed by the
author and co-workers at the Royal Botanic Garden Edinburgh (RBGE) who have
substantially increased sampling of genera from the Malesian region, in particular
from Pennington’s tribe Isonandreae (Wilkie et al., in prep.). This will be used to help
establish a robust generic and infrageneric framework to facilitate future monographic
research on the family in Malesia. Researchers at RBGE are also using this data to
investigate biogeographic patterns within Isonandreae (Richardson et al., in prep.).
Floristic studies
Several floristic accounts of the family have been produced in the Malesian region.
The Tree Flora of Malaya (Ng 1972) covered 11 genera and 76 species and kept
Sarcosperma in the Sarcospermaceae, the Manual of the Larger and More Important
non Dipterocarp Trees of Central Kalimantan (Argent et al. 1997) covered 5 genera
and 36 species, and the Tree Flora of Sabah and Sarawak (Chai & Y1i 2002) covered
11 genera and 120 species, including Sarcosperma. Two further flora accounts are
in preparation. The Sapotaceae account for the Flora of Peninsular Malaysia will
148 Gard. Bull. Singapore 63(1 & 2) 2011
cover an estimated 1] genera and more than 80 species and will have contributions
from a wide range of Malaysian taxonomists (Wilkie et al., in prep.) and the Flora of
Thailand Sapotaceae account covering 9 genera and 45 species (Chantaranothai 1999)
is due to be published by 2011 (Chantaranothai, in prep.).
Challenges to delivering a Flora Malesiana account
An account of Sapotaceae for Flora Malesiana with an estimated 15 genera and
300 species is clearly deliverable; however, there are several issues that need to be
addressed if a modern account is to be produced. The first is the development of a
robust generic framework within Sapotaceae. It has been recognised that generic limits
are problematic and that the circumscription of several genera are still not clarified
fully. Recent molecular phylogenetic studies are helping address this but much still
needs to be done, in particular increased sampling of taxa found in the Malesian
region. A second challenge is that previous sectional classifications of genera by
authors such as Dubard (1909), Lam (1925), van Royen (1960a) are not congruent
with recent molecular phylogenies (Wilkie et al., in prep). This makes producing
taxonomic accounts of large genera such as Madhuca and Palaquium difficult. It 1s
therefore important that new sectional classifications are developed for these genera
so that taxonomists can work with groups of manageable size. Finally, there is still a
lack of good fertile material for many Malesian taxa. If full taxonomic descriptions are
desired, a detailed collecting programme for the family is needed, in particular from
the under-collected areas of East Malesia.
Facilitating the production of a Flora Malesiana account
For taxonomists to contribute to the Flora Malesiana Sapotaceae account, they need
access to specimen information, access to the actual specimens and access to the
literature. Over the past two years, RBGE has been trying to address these issues largely
through the establishment of the Sapotaceae Resource Centre (www.sapotaceae. info).
This brings together Sapotaceae specimen data held by various institutes in Malesia
and Europe (E, K, KEP, L, SAN, SAR, SING) as well as from individuals who have
worked on Sapotaceae. To date, access to over 45,000 Sapotaceae specimen records
is available via the website. The data is constantly being updated and cleaned as it is
used for projects such as the Flora of Peninsular Malaysia.
A large collection of Sapotaceae herbarium specimen images has also been
gathered. In order that they are as accessible to as many people as possible, these
are being linked to specimen information on the website. However, with over 16,000
Sapotaceae herbarium specimen images from Malesia, this will take some time.
The most effective and efficient way to make these images available on the web is
constantly being reviewed.
A library of some 500 Sapotaceae reprints built up over 30 years by T.D.
Pennington and V. Wink has been converted to PDF format and the best way to share
this information (taking into account copyright issues and other international projects
such as the Biodiversity Heritage Library) is currently being investigated.
Sapotaceae in the Malesian flora 149
The delivery of a Flora Malesiana Sapotaceae account will require the
enthusiasm and commitment of taxonomists from the Malesian region, particularly
those at an early stage in their career. The Sapotaceae Resource Centre is designed to
stimulate interest in the family as well as link Sapotaceae researchers together.
Conservation assessments
Taxonomists should not just be in the business of documenting plants before they are
gone—they need to be doing more to make sure they don’t go in the first place. Integral
to the Flora Malesiana Sapotaceae account has to be the production of Conservation
Assessments (IUCN 2001) as these can lead to long-term monitoring of species and
their active conservation (e.g., Chan 2007; Chua et al. 2009, 2010).
Conclusion
Delivering an account of Sapotaceae is achievable in the medium term. However, this
will require the commitment of a substantial amount of time by researchers. To help
this happen, the objectives of Flora Malesiana need to be embedded and championed
by many more institutes. Taxonomists from Malesia also need to be much more
involved in producing accounts than has traditionally been the case. This will require
financial support. Flora Malesiana as the collective voice of Malesian taxonomy is
ideally placed to address these concerns and to help leverage funding from global
initiatives to achieve these aims.
ACKNOWLEDGEMENTS. The Forest Research Institute Malaysia is thanked for hosting
me for a year to study Sapotaceae (EPU permit 40/20019/2522). The European Union
SYNTHESYS Programme is acknowledged and thanked for its financial support (Ref. NL-
TAF-87) to visit the Leiden Herbarium and to study and photograph its Sapotaceae collections.
Dr. Terry Pennington and Dr. Willem Vink are thanked for making so much of their Sapotaceae
resources available to me and Dr. Georgina Stewart and James Stewart are thanked for
converting the Sapotaceae literature to PDF’s and for editing various manuscripts. Last but
by no means least, Dr. Martin Pullan and Dr. James Richardson are thanked for their help in
developing the Sapotaceae website.
References
Anderberg, A.A. & Swenson, U. (2003) Evolutionary lineages in Sapotaceae
(Ericales): A cladistic analysis based on ndhF sequence data. /nt. J. Plant Sci.
164(5): 763-773.
Argent, G., Saridan, A., Campbell, E.J.F. & Wilkie, P. (1997) Manual of the Larger
and More Important Non-dipterocarp Trees of Central Kalimantan. Two vols.
Forest Research Institute, Samarinda, Indonesia.
Assem, J.v.d. (1953) Revision of the Sapotaceae of the Malaysian area in a wider
sense. 4. Ganua Pierre ex Dubard. B/umea 7(2): 364-400.
150 Gard. Bull. Singapore 63(1 & 2) 2011
Assem, J.v.d. & Kostermans, A.J.G.H. (1954) Revision of the Sapotaceae of the
Malaysian area 1n a wider sense. 4a. Ganua Pierre ex Dubard (Additional Notes).
Blumea 7(3): 481-483.
Aubréville, A. (1964) Sapotacées. Adansonia, Memoire no. |: 1-157.
Baehni, C. (1965) Mémoires sur les Sapotacées. 3. Inventaire des genres. Boissiera
Lie
Baillon, H. (1891) Histoire des Plantes. (Sapotacées) 11: 255.
Bartish, I.V., Swenson, U., Munzinger, J. & Anderberg, A.A.(2005) Phylogenetic
relationships among New Caledonian Sapotaceae (Ericales): molecular evidence
for generic polyphyly and repeated dispersal. Amer. J. Bot. 92(4): 667-673.
Boer, E. & Ella, A.B. (eds) (2000) Plant Resources of South East Asia No. 18. Plants
Producing Exudates. Leiden: Backhuys Publishers.
Bruggen, A.C. van (1958a) Revision of the Sapotaceae of the Malaysian area in a
wider sense. 15. Payena A. De Candolle. Blumea 9(1): 89-138.
Bruggen, A.C. van (1958b) Revision of the Sapotaceae of the Malaysian area in a
wider sense. 16. Aesandra Pierre. Blumea 9(1): 139-142.
Burkill, I.H. (1966) 4 Dictionary of the Economic Products of the Malay Peninsula.
Vol. 2. Kuala Lumpur: Ministry of Agriculture and co-operatives. 1241-2444.
Chai, P.P.K. & Yui, P.-C. (2002) Sapotaceae. In: Soepadmo, E., Saw, L.G. & Chung,
R.C.K. (eds) Tree Flora of Sabah and Sarawak 4: 203-342. Kuala Lumpur:
Forestry Research Institute Malaysia.
Chan, Y.M. (2007) A ray of hope for Dipterocarpus semivestitus once thought extinct.
Conservation Malaysia 6: 4.
Chantaranothai, P. (1999) The Sapotaceae of Thailand. Thai Forest Bulletin 27: 139-166.
Chua, L.S.L., Kiew, R. & Chan, Y.M. (2009) Assessing conservation status of
Peninsular Malaysian Begonias (Begoniaceae). Blumea 54 (1-3): 94-98.
Chua, L.S.L., Suhaida, M., Hamidah, M. & Saw, L.G. (2010) Malaysia Plant Red List:
Peninsular Malaysian Dipterocarpaceae. Research Pamphlet No. 129, Forest
Research Institute Malaysia.
Dubard, M. (1909) Les Sapotacées, du groupe des Isonandrées. Revue Générale de
Botanique. Paris.
Dubard, M. (1912) Les Sapotacées de groupe des Sideroxylinées. Ann. Inst. Bot. Geol.
Colon. Marseille, Ser. 2, 10: 1.
Dubard, M. (1915) Les Sapotacées de groupe des Sideroxylinées-Mimusopées. Ann.
Inst. Bot. Geol. Colon. Marseille, Ser. 3, 3: 1.
Govaerts, R., Frodin, D.G. & Pennington, T.D. (2001) World Checklist and Bibliography
of Sapotaceae. Kew: Royal Botanic Gardens Kew.
Herrmann-Erlee, M.P.M. & Lam, H.J. (1957) Revision of the Sapotaceae of the
Malaysian area in a wider sense. 8. Krausella H.J.Lam. Blumea 8(2): 446-451.
Herrmann-Erlee, M.P.M. & Royen, P. van (1957) Revision of the Sapotaceae of the
Malaysian area in a wider sense. 9. Pouteria Aublet. Blumea 8(2): 452-509.
IUCN (2001) UCN Red List Categories and Criteria, version 3.1. Gland (Switzerland)
and Cambridge (U.K.): IUCN Species Survival Commission.
Sapotaceae in the Malesian flora PSil
Jeuken, S.J. (1952) Revision of the Sapotaceae of the Malaysian area in a wider sense.
1. A monograph of the genus /sonandra. Blumea 6(3): 547-615.
Lam, H.J. (1925) The Sapotaceae, Sarcospermaceae and Boerlagellaceae of the Dutch
East Indies and surrounding countries. Bull. Jard. Bot. Buitenzorg, Ser. 3, 7:
1-289.
Lam, H.J. (1927) Further studies on Malayan Sapotaceae. |. Bull. Jard. Bot. Buitenzorg,
Ser. 3, 8: 381-493.
Lam, H.J. (1932) Enumeration of Sapotaceae, thus far known from New Guinea. Nova
Guinea 14(4): 549-567.
Lam, H.J. (1935) Phylogeny of single features: as illustrated by a remarkable new
Sapotaceous tree from British Malaya (Madhuca Ridleyi, n. sp.). Gard. Bull.
Straits Settlem. 9: 98-112.
Lam, H.J. (1938) Aulandra cauliflora H.J.Lam. Hooker s Icones Plantarum 4(3): Plate
3360.
Lam, H.J. (1939) On the system of the Sapotaceae, with some remarks on taxonomical
methods. Recueil Trav. Bot. Néerl. 36: 509-525.
Lam, H.J. (1941a) Notes on the Sapotaceae-Mimusopoideae in general and on the far-
eastern Manilkara-allies in particular. Blumea 4(2): 323-358.
Lam, H.J. (1941b) Sarcosperma in New Guinea. B/umea 4(2): 322.
Lam, H.J. (1942) A tentative list of wild Pacific Sapotaceae, except those from New
Caledonia. Blumea 5(1): 146.
Lam, H.J. (1943a) Note on Pouteria (Sapotaceae). Blumea 5(2): 336-338.
Lam, H.J. (1943b) Trois noueaux Planchonella de la Nouvelle Guinée avec ine note
sur Krausella. Boissiera 7: 91-99.
Lam, H.J. (1948) Sarcospermataceae. Flora Malesiana Ser. 1, 4: 32-34.
Lam, H.J. (1957) Revision of the Sapotaceae of the Malaysian area in a wider sense.
4b. Ganua Pierre ex Dubard (Second Supplement) Blumea 8(2): 510-513.
Lam, H.J. & Royen, P. van (1952a) Revision of the Sapotaceae of the Malaysian area
in a wider sense. 2. Burckella Pierre. Blumea 6(3): 580-593.
Lam, H.J. & Royen, P. van (1952b) Concise revision of the Sarcospermataceae.
Blumea 7(1): 148-153.
Lam, H.J. & Varossieau, W.W. (1938) Revision of the Sarcospermataceae. B/umea
3(1): 183-200.
Lam, H.J. & Varossieau, W.W. (1939) Additional notes on Sarcosperma. Blumea 3(2):
262-262.
Ng, F.S.P. (1972) Sapotaceae. In: Whitmore, T.C. (ed) Tree Flora of Malaya |: 473.
Kuala Lumpur: Longman Malaysia.
Pennington, T.D. (1991) The Genera of Sapotaceae. Kew: Royal Botanic Gardens
Kew.
Pierre, J.B.L. (1890) Notes Botaniques: Sapotacées. Paris.
Royen, P. van (1952) Revision of the Sapotaceae of the Malaysian area in a wider
sense. 3. Mimusops L. Blumea 6(3): 594-595.
Royen, P. van (1953) Revision of the Sapotaceae of the Malaysian area in a wider
sense. 5. Manilkara Adanson em. Gilly in the Far East. B/umea 7(2): 401-412.
S52 Gard. Bull. Singapore 63(1 & 2) 2011
Royen, P. van (1957a) Revision of the Sapotaceae of the Malaysian area in a wider
sense. 7. Planchonella. Blumea 8(2): 235-445.
Royen, P. van (1957b) Revision of the Sapotaceae of the Malaysian area in a wider
sense. 2a. Additional notes on Burckella Pierre. Blumea 8(2): 201-203.
Royen, P. van (1957c) Revision of the Sapotaceae of the Malaysian area in a wider
sense. 5a. Additional notes on Manilkara Adanson. Blumea 8(2): 204-206.
Royen, P. van (1957d) Revision of the Sapotaceae of the Malaysian area in a wider
sense. 6. Xantolis Rafinesque. Blumea 8(2): 207-234.
Royen, P. van (1958a) Revision of the Sapotaceae of the Malaysian area in a wider
sense. 14. Diploknema Pierre. Blumea 9(1): 75-88.
Royen, P. van (1958b) Revision of the Sapotaceae of the Malaysian area in a wider
sense. 17. Aulandra H.J. Lam. Blumea, Suppl. 4. (H.J. Lam Jubilee volume):
263-267.
Royen, P. van (1959) Revision of the Sapotaceae of the Malaysian area in a wider
sense. 2b. Additional notes on Burckella Pierre. 9a. Additional notes on Pouteria
Ablet. 18. Cassidispermum Hemsley. 19. Chelonespermum Hemsley. Nova
Guinea 10(1):131—142.
Royen, P. van (1960a) Revision of the Sapotaceae of the Malaysian area in a wider
sense. 23. Palaquium Blanco. Blumea 10(2): 432-606.
Royen, P. van (1960b) Revision of the Sapotaceae of the Malaysian area in a wider
sense. 20. Madhuca Gmelin. Blumea 10(1): 1-117.
Royen, P. van (1960c) Revision of the Sapotaceae of the Malaysian area in a wider
sense. 22. Mastichodendron Cronquist. Blumea 10(1): 122-125.
Royen, P. van (1960d) Revision of the Sapotaceae of the Malaysian area in a wider
sense. 21. Eberhardtia Lecomte. Blumea 10(1): 118-121.
Smedmark, J.E.E. & Anderberg, A.A. (2006) Accounting for variation of substitution
rates through time in Bayesian phylogeny reconstruction of Sapotoideae
(Sapotaceae). Molec. Phylogenet. Evol. 39: 706-721.
Smedmark, J.E.E. & Anderberg, A.A. (2007) Boreotropical migration explains
hybridization between geographically distant lineages in the pantropical clade
Sideroxyleae (Sapotaceae). Amer. J. Bot. 94(9): 1491—1505.
Swenson, U. & Anderberg, A.A. (2005) Phylogeny, character evolution, and
classification of Sapotaceae (Ericales). Cladistics 21: 101-130.
Swenson, U., Bartish, I.V. & Munzinger, J. (2007) Phylogeny, diagnostic characters
and generic limitation of Australasian Chrysophylloideae (Sapotaceae, Ericales):
evidence from ITS sequence data and morphology. Cladistics 23: 201-228.
Swenson, U., Munzinger, J. & Bartish, I.V. (2007) Molecular phylogeny of
Planchonella (Sapotaceae) and eight new species from New Caledonia. Jaxon
56(2): 329-354.
Triono, T., Brown, A.H.D., West, J.G. & Crisp, M.D. (2007) A phylogeny of Pouteria
(Sapotaceae) from Malesia and Australasia. Austral. Syst. Bot. 20: 107-118.
Vink, W. (1957) Revision of the Sapotaceae of the Malaysian area in a wider sense. 10.
Leptostylis Bentham. 11. Pycnandra Bentham. 12. Magodendron Vink. Nova
Guinea 8(1): 87-128.
Sapotaceae in the Malesian flora 153
Vink, W. (1958) Revision of the Sapotaceae of the Malaysian area in a wider sense. 13.
Chrysophyllum L. Blumea 9(1): 21—74.
Vink, W. (1995) Revision of Magodendron (Sapotaceae) with observations on floral
development and morphology. B/umea 40(1): 91-107.
Vink, W. (2001) Three Bornean species of Madhuca (Sapotaceae). Blumea 46(1):
193-199.
Vink, W. (2002) Some Malesian species of Pouteria (Sapotaceae). Blumea 47(1):
95-147.
ex
a fl es aa? e =\(SC At,
~<¥
&
41
|
sdriégenp atl
éae® 8 ?
setts" diintaady
isco?
hi ott T4Pt ey
riley, a
a) bret Fives)
7a
air
-
Gardens’ Bulletin Singapore 63(1 & 2): 155-162. 2011 155
Precursor to flora account of Procris (Urticaceae)
in Peninsular Malaysia
Barry J. Conn! and Julisasi T. Hadiah?
"National Herbarium of New South Wales,
Mrs Macquaries Rd, Sydney NSW 2000, Australia
barry.conn@rbgsyd.nsw.gov
*Pusat Konservasi Tumbuhan, Kebun Raya Bogor,
Jl. Ir. H. Juanda 13, Bogor 16122, Indonesia
ABSTRACT. A review of Elatostema J.R.Forst. & G.Forst. and Procris Juss. (Urticaceae)
occurring in Peninsular Malaysia resulted in four taxa, currently classified in Elatostema,
being transferred to the genus Procris. Six species of Procris are recognised as occurring in
Peninsular Malaysia with the following new combinations provided here: Procris acaulis
(Hook.f.) B.J.Conn & J.T.Hadiah; Procris curtisii (Ridl.) B.J.Conn & J.T.Hadiah; and Procris
repens (Lour.) B.J.Conn & J.T.Hadiah. A modified description of Procris (including Pellionia)
and a key to the species occurring in Peninsular Malaysia are provided.
Keywords. Elatostema, Pellionia, Peninsular Malaysia, Procris, Urticaceae
Introduction
Elatostema is a very large genus that is considered to consist of approximately 300
herbaceous and sub-shrubby species (Friis 1993) that are characterised by having
female flowers arranged on a flattened discoid or lobed receptacle. It is the type genus
of the Tribe Elatostemateae (Conn & Hadiah 2009). Elatostema is widely distributed
throughout the tropical, subtropical and sub-temperate regions from the west to east
coast of Africa, Madagascar and Mascarene Islands, through Sri Lanka, southern
India, tropical Himalaya, Bangladesh, Myanmar to South-East Asia, Micronesia, then
throughout Papuasia, to eastern Australia, New Caledonia, northern New Zealand, and
Polynesia. It also occurs in subtropical and sub-temperate regions of China, Taiwan
and Japan. Since it was first described by Forster & Forster (1775), the taxonomic
circumscription of the genus and infrageneric taxa have been problematic. The generic
description includes features of what is now regarded as the separate genus Procris.
In the protologue of Elatostema, the species Elatostema pedunculatum and E. sessile
were described. Elatostema pedunculatum was circumscribed as having flowers
with 5 stamens (‘pentandrum’), whereas the flowers of E. sessile have 4 stamens
(‘tetrandrum’). The former species was later transferred to the genus Procris (Weddell
1856), whereas the latter remained in Elatostema. Furthermore, the taxonomic
distinction of the related taxa, E/atostematoides, Pellionia and Procris has continued
to be problematic.
156 Gard. Bull. Singapore 63(1 & 2) 2011
The first, and most complete account of the taxonomy of Elatostema was
provided by Schréter & Winkler (1935, 1936). They modified the subgeneric division
of previous workers (Hallier 1896, Winkler 1922) and proposed four subgenera,
namely, subg. E/atostema (as “Euelatostema’ ), subg. Elatostematoides, subg. Pellionia,
and subg. Weddellia. This division was primarily based on the nature of the leaves,
stipules, inflorescence and the presence and form of the receptacle. Some authors
maintain Pe//ionia as a distinct genus, for example, Wang (1980); Yahara (1984); Chen
et al. (2003). The latter authors circumscribed Pellionia as having cymose female
inflorescences, rarely with discoid receptacle and involucres, whereas the female
inflorescences of Procris are capitate or globose and lack an involucre. Furthermore,
the female flowers of Pe/lionia have (4 or) 5 perianth lobes, whereas female flowers
of Procris have 3 or 4 perianth lobes. Schréter & Winkler (1935, 1936) recognised
Procris as a separate genus. Based on chloroplast sequence data, preliminary analyses
of relationships within Elatostema do not support the recognition of the subgenus
Pellionia (Hadiah et al. 2003). Hadiah et al. (2008) concluded that there are four
genera within the Tribe Elatostemateae, namely Elatostema, Lecanthus, Pilea and
Procris, with Pellionia reduced to the synonymy of the last mentioned genus.
Materials and methods
Herbarium collections held at BO, K, KEP, NSW, SING were examined. Descriptive
data are managed by DeltaAccess 2.0 software (Hagedorn 2007 onwards).
An identification key to the genera of the tribe E/atostemateae occurring in
Peninsular Malaysia is provided. A brief description of Procris and a key to species
recognised for Peninsular Malaysia, together with a list of selected specimens
examined, are provided.
Distinguishing features of Elatostema and Procris
The following key summarises the diagnostic features of the genera of tribe
Elatostemateae that occur in Peninsular Malaysia, namely, Elatostema, Pilea and
Procris (including Pellonia).
Key to genera of tribe Elatostemateae
la. Leaves opposite, more or less isophyllous; lamina + equal basally; venation 3-pli-
MOL VEM vie wciccecscenes uatvn agpaients canne txdeesre cae eee reeee eee ee ee oe eee Pilea
lb. Leaves appearing alternate, anisophyllous, with nanophyll reduced, and usually
soon caduous; lamina very unequal basally; venation pinnate ................c:eeeee 2
Procris in Peninsular Malaysia 157
2a. Perianth lobes of female flowers usually 4 or 5, much shorter than ovary, or strongly
reduced, not corniculate at apex; achene 6—10-ribbed; male inflorescences usually
with receptacle, rarely cymose; receptacle of female inflorescences discoid ...........
«rps pascensisclzoyceeelelierr Jee Oa eee eee ee Elatostema
2b. Perianth lobes of female flowers 3—5, longer than ovary, usually corniculate
below apex; achene tuberculate or striate, rarely smooth, never ribbed; male
inflorescences cymose; receptacle of female inflorescences globose or head-like ..
ME PRR Sac 28 usc 5 diac bichn Vuveneted Sepsis vnvusbesvada CodddbenancaevetOuacdeaeess: Procris
Species of Procris in Peninsular Malaysia
In the account of the Urticaceae for the Malay Peninsula by Ridley (1924), four
species of Pellionia, five species of Elatostema and two species of Procris were
recognised. Turner (1995) provided an updated catalogue of species of Urticaceae
that were reported to occur in this region. In his catalogue, he listed 11 species of
Elatostema (including taxa previously recognised as belonging to Pellionia) and two
species of Procris. In this review of Turner’s list of species for these two genera, four
species listed as E/atostema are here transferred to Procris, resulting in six species
being recognised for this region.
Procris Commers. ex Juss., Genera Plantarum 403 (1789); Schréter, Repertorium
Specierum Novarum Regni Vegetabilis 45: 179-192 and 257-300 (1938), partly
revised.
TYPE SPECIES: Procris axillaris J.F.Gmel., Systema Naturae 2: 267 (1791).
Shrubs, self-supporting (erect/suberect), terrestrial or epiphytic, monoecious or
dioecious; branched and stinging hairs absent; internodes elongate, distinct. Stipules
caducous, free, axillary, intrapetiolar. Leaves subopposite to alternate, petiolate; lamina
not lobed; base oblique; margin toothed or entire; apex variable, leaf surface glabrous
or hairy; venation symmetric, secondary vein pinnate. Nanophylls present, often
not persistent. Cystoliths linear. Flowers unisexual. Male inflorescences distinctly
pedunculate, paniculate (usually openly branched), branching unordered; involucral
bracts absent. Male flowers actinomorphic; tepals 4 or 5, free; stamens 4 or 5, inflexed;
rudimentary ovary present. Female inflorescences pedunculate or sessile, head-like
(condensed); involucral bracts absent. Female flowers actinomorphic (or slightly
asymmetrical); tepals 3-5, equal, connate (at least in part); staminodes present; ovary
straight; style absent; stigma oblong, filiform to linear. Achene enclosed by perianth
(or almost so) or not enclosed by perianth, smooth or variously rough.
Distribution. Central western and eastern Africa, Madagascar to Sri Lanka and India,
Myanmar, southern China, Taiwan, throughout South-East Asia, Papua New Guinea,
Solomon Islands, New Caledonia, and east throughout south-west Pacific islands.
158 Gard. Bull. Singapore 63(1 & 2) 2011
Key to species of Procris in Peninsular Malaysia
la.» Leaf petiol ators: sacs cc eaci coe Caeser ee ce ee 2
Ib: heaf-sessile or subsessile:.c.x once etree eee p)
2a. Nanophyll appearing absent, not persistent, soon dehiscent ..............:cccccceeeeeeeees 3
2b.) (Nanophyllipresentandipersisicmiyeee eres tee eee Procris latifolia
3a. Leaf margin entire or sometimes slightly toothed or wavy near apex ................. 4
3b. Leaf margin toothed (dentate, serrate or crenate) .................06 Procris frutescens
4a. Leaves (megaphylls) ovate; venation asymmetric, 2- or 3-plinerved, secondary
veins faint butusually distinct: petiole haimy -ecoce-eeeste eet Procris acaulis
4b.. Leaves (megaphylls) oblong-elliptic; venation symmetric, pinnate, secondary
Veins inconspicuous; petiole: glabrous. 22. ee Procris pedunculata
5a. Nanophylls present and persistent; leaf (megaphyll) broadly ovate; margin with
rounded teeth; venation actinodromus; male inflorescence openly branched ........
DER BER A LR ee ee Procris repens
5b. Nanophylls absent; leaf (megaphyll) narrowly ovate-elliptic; margin obscurely
toothed; venation pinnate; male inflorescence more or less compact with closed
Enea Dat 22s 5 esa och See eo cee eee eee ee Procris curtisil
Note. The synonyms cited are only those that apply to Peninsular Malaysia.
1. Procris acaulis (Hook.f.) B.J.Conn & J.T.Hadiah, comb. nov.
Basionym: Pellionia acaulis Hook.f., Flora of British India 5: 562 (1888). Synonyms:
Pellionia javanica Wedd. var. acaulis (Hook.f.) Ridl., J. Straits Br. Roy. Asiat. Soc. 59:
187 (1911); Elatostema latifolium Blume var. acaule (Hook.f.) H.Schroeter, Repert.
Spec. Nov. Regni Veg. Beih. 83(2): 17 & 1935 (1936). TYPE: Malaysia, Penang, on
damp rocks, Kings Collectors 1659, May 1881 (K).
Distribution. Peninsular Malaysia, Cambodia.
Selected specimens examined: PENINSULAR MALAYSIA. Perak: Temango, Ridley
14560, Jul 1909 (SING). Penang: refer Type above.
2. Procris curtisii (Ridl.) B.J.Conn & J.T.Hadiah, comb. nov.
Basionym: Pellionia curtisii Ridl., J. Straits Br. Roy. Asiat. Soc. 82: 196 (1920).
Synonym: Elatostema curtisii (Ridl.) H.Schroeter, Repert. Spec. Nov. Regni Veg. Beih.
83(2): 35 (1936). TYPE: Malaysia, Perak, Gunung Bajong, Malacca am Kantaflu8,
Curtis s.n., Aug 1898 (SING).
Procris in Peninsular Malaysia 159
Distribution. Peninsular Malaysia, Thailand, Philippines, Indonesia, Papua New
Guinea, Solomon Islands.
Selected specimens examined: PENINSULAR MALAYSIA. Perak: Lenggong,
Ridley s.n., Aug 1909 (SING124283). Selangor: Gua Batu, Ridley 4717, 23 Jun 1889
(SING124284); 8/96, Dec 1898 (SING124286).
3. Procris frutescens Blume, Bijdr. 510 (1825-1826); H.Schroeter, Repert. Spec. Nov.
Regni Veg. 45: 272 (1938), descr. ampl. TYPE: Indonesia, Java, Blume 707, without
date (L).
Distribution. Peninsular Malaysia, Thailand, Philippines, Indonesia, Papua New
Guinea, Solomon Islands.
Selected specimens examined: PENINSULAR MALAYSIA. Perak: Larut and Matang,
Maxwell Hill, Burkhill 12686, 6 Mar 1924 (SING124847). Kelantan: Gua Musang,
Batu Papan, Kiew 2882 & Anthonysamy, 9 May 1990 (SING124843). Pahang:
Cameron Highlands, Boh Plantation, Mohd Nur 32746, 26 Apr 1937 (SING124838).
Johor: Mawai—Jemaluang road, Corner 32461, 11 Oct 1936 (SING124839 & 124840).
4. Procris latifolia Blume, Bijdr. 10: 509 (1825).
Pellionia latifolia (Blume) Boerlage, Handl. 3: 375 (1900). Elatostema latifolium
(Blume) H.Schroeter, Repert. Spec. Nov. Regni Veg. Beih. 83(2): 17 & 1935 (1936).
TYPE: Indonesia, Java, Res. Bantam, Sadjira, Blume s.n. (B, BO, BRSL, L).
Pellionia helferiana Wedd., Prodr. Syst. Nat. Regni Veg. 16(1): 170 (1869); Elatostema
helferianum (Wedd.) Hallier f., Ann. Jard. Bot. Buitenzorg 13: 316 (1896). TYPE:
Andaman, Helfer 4551 (K).
Pellionia javanica Wedd., Arch. Mus. Hist. Nat. 8: 288 (1856); Elatostema javanicum
(Wedd.) Hallier f., Ann. Jard. Bot. Buitenzorg 13: 316 (1896). TYPE: Indonesia, Java,
Lobb 283, anno 1846 (K, W).
Distribution. Peninsular Malaysia, Myanmar, Vietnam, Thailand, Indonesia.
Selected specimens examined: PENINSULAR MALAYSIA. Kedah: Langkawi, Selat
Panchor Henderson 29083, 23 Nov 1934 (SING124292). Penang: Plant House no. 1,
Mohd Nur s.n., 23 Sep 1918 (SING124293); Balik Pulau, Curtis 682, Mar 1886 (K,
SING124303 & 124304). Perak: Kuala Kangsar: Padang Rengas, Burkill 13580, 15
Jun 1924 (SING124295); Gunong Bubu, Chew 1220, 18 Aug 1966 (SING124298);
Tapah, Jor, Mohd Haniff 14253, 15 Sep 1924 (SING124312). Kelantan: Gua Musang,
Henderson 22693, 13 Aug 1929 (SING124307); Henderson 19503, 21 Oct 1927
(SING124311). Selangor: Kuala Lumpur, Batu Caves, Jeruyva 498, 20 Jan 1929
160 Gard. Bull. Singapore 63(1 & 2) 2011
(SING124291); Ridley 13370, Aug 1900 (SING124322). Pahang: Tioman, Juara Bay,
Burkill s.n., Jun 1915 (SING124296 & 124297); Jerantut: Gunung Tahan, Mohd Haniff
10197 & Mohd Nur, 21 Jun 1922 (SING124314). Malacca: Bukit Naning, Alvins
1105, 26 Mar 1885 (SING124294). Johor: Gunung Sumalayang, Chin 597, 3 Feb 1971
(SING124300); Sungai Segun, Corner 30694, 10 Apr 1936 (SING124301 & 124302);
Labis, Sungai Juasseh, Mohd Shah 2287 & Ahmad, 31 Jan 1971 (SING124317).
5. Procris pedunculata (J.R.Forst. & G.Forst.) Wedd., in DC. Prodr. 16: I, 191 (1869);
H.Schroeter, Repert. Spec. Nov. Regni Veg. 45: 259, (1938) (emend., as P. pedunculata
var. eupedunculata).
Basionym: Elatostema pedunculata J.R.Forst. & G.Forst., Characteres Generum
Plantarum, 53 (1775). LECTOTYPE (designated by Smith 1981): Society Islands
(Tahiti), J.R. Forster s.n., (K; isolectotype MW).
Distribution. Peninsular Malaysia, Marquesas Islands, Philippines, Indonesia, Papua
New Guinea, Solomon Islands, Christmas Islands; Cook Islands; Fiji; Society Islands.
Selected specimens examined: PENINSULAR MALAYSIA. Perlis: Titi Tinggi, Mata
Ayer Forest Reserve, Kiew 3702, 1 Jul 1993 (SING124852). Kelantan: Gua Musang,
Kiew 2992, 15 May 1990 (SING124851).
6. Procris repens (Lour.) B.J.Conn & J.T.Hadiah, comb. nov.
Basionym: Polychroa repens Lour., Fl. Cochinchinensis 2: 559 (1790). Synonyms:
Elatostema repens (Lour.) Hallier f., Ann. Jard. Bot. Buitenzorg 13: 316. (1896);
Pellionia repens (Lour.) Merr., Lingnan Sci. J. 6(4): 326 (1928). TYPE: unknown
— taxonomic concept applied here is based on the amplified concept of Schréter &
Winkler (1935).
Pellionia daveauana N.E.Br., Gard. Chron. 1880(2): 262 (1880). TYPE: Vietnam,
Phugnoe, G. Lebeuf s.n., without date (K).
Distribution. Peninsular Malaysia, Thailand, Philippines, Indonesia.
Notes. Schroéter (in Schréter & Winkler 1935) recognised three varieties within
‘Elatostema repens’; however, the taxonomic status of these taxa have not been
evaluated here.
Selected specimens examined: PENINSULAR MALAYSIA. Perlis: Gua Burma, Kiew
3655, 28 Jun 1993 (SING124791). Kedah: Baling, Kedah Peak, Curtis s.n., without
date (SING124785). Penang: Curtis s.n., without date (SING124782—124784);
Plant House no. 9, Mohd Nur s.n. 24 Sep 1918 (SING124798). Perak: Scortechini
485, without date (K); Hulu Perak, Grik, Burkill 12545 & Mohd Haniff, 19 Jun
1924 (SING124779); Kuala Kangsar, Kota Lama, Mohd Haniff 15528, 20 Oct 1924
Procris in Peninsular Malaysia 161
(SING124796). Kelantan: Gua Ninik, Henderson 19690, 26 Oct 1927 (SING124788):
Sungai Ketil, Henderson 22665, 12 Aug 1929 (SING124789); Sungai Bring, Kiew
2904, 11 May 1990 (SING124790). Selangor: Gombak, Batu Caves, Ridley 8186,
Dec 1896 (SING124800); Kepong, Symington 25169, 29 Jul 1931 (SING124806).
Pahang: Bentong, Burkill 16533 & Mohd Haniff, 6 Nov 1924 (SING124777); Raub,
Gali, Burkill 16839 & Mohd Haniff, 13 Nov 1924 (SING124778). Negeri Sembilan:
Bukit Sutu, A/vins 1962 1 Nov 1885 (SING124775); Tampin, Burkill 2575, 14 Jul
1917 (SING124776).
ACKNOWLEDGEMENTS. We acknowledge the generous support given to us by the
Singapore Botanic Gardens Visiting Research Fellowship in 2009.
References
Chen, J., Lin, Q., Friis, I., Wilmot-Dear, C.M. & Monro, A.K. (2003) Urticaceae. In:
Wu, Z.Y., Raven, P.H. & Hong, D.Y. (eds) Flora of China, vol. 5 (Ulmaceae
through Basellaceae), pp. 76-189. Beijing & St. Louis: Science Press & Missouri
Botanical Garden Press.
Conn, B.J. & Hadiah, J.T. (2009) Nomenclature of tribes within the Urticaceae. Kew
Bull. 64(2): 349-352.
Forster, J.R. & Forster, G. (1775) 53. Elatostema. Characteres Generum Plantarum.
London: B. White, T. Cadell & P. Elmsly.
Friis, I. (1993) Urticaceae. In: Kubitzki, K., Rohwer, J.G. & Bittrich, V. (eds), The
Families and Genera of Vascular Plants II, Flowering Plants — Dicotyledones.
Magnoliid, Hamamelid and Caryophyllid Families, vol. II, pp. 612—630. Berlin:
Springer-Verlag.
Hadiah, J.T., Conn, B.J. & Quinn, C.J. (2008) Infra-familial phylogeny of Urticaceae,
using chloroplast sequence data. Austral. Syst. Bot. 21(5): 375-385.
Hadiah, J.T., Quinn, C.J. & Conn, B.J. (2003) Phylogeny of Elatostema (Urticaceae)
using chloroplast DNA data. TJe/opea 10(1): 235-246
Hagedorn, G.M. (2007 onwards) DiversityDescriptions. Available at http://www.
diversityworkbench.net/Portal/DiversityDescriptions (verified 10 Mar 2011).
Hallier, H. (1896) Neue und bemerkenswerte pflanzen aus dem malaiischpapuanischen
inselmeer. Ann. Jard. Bot. Buitenzorg 13: 276-326.
Ridley, H.N. (1924) Urticaceae. In: The Flora of the Malay Peninsula, vol. 3 - Apetalae,
pp. 317-368. London: L. Reeve & Co.
Schréter, H. & Winkler, H. (1935) Monographie der gattung Elatostema s.1.:
Allgemeiner teil. Repert. Spec. Nov. Regni Veg. 83(1): 1-71.
Schroter, H. & Winkler, H. (1936) Monographie der gattung Elatostema s.1.: Spezieller
teil. Repert. Spec. Nov. Regni Veg. 83(2): 1-237.
162 Gard. Bull. Singapore 63(1 & 2) 2011
Smith, A.C. (1981) Urticaceae. In: Smith, A.C. (ed) Flora Vitiensis Nova, vol. 2,
Angiospermae: Dicotyledones, Families 44-116, pp. 209-251. Hawaii: Pacific
Tropical Botanical Garden.
Turner, I.M. (1995) A catalogue of the vascular plants of Malaya (Urticaceae). Gard.
Bull. Singapore 47(2): 491-493.
Wang, W.T. (1980) Classificatio specierum Pellionae (Urticaceae). Bull. Bot. Lab. N.
E. Forest. Inst., Harbin 6: 45-66.
Weddell, H.A. (1856) Monographie de la famille des Urticées. Nouv. Arch. Mus. Hist.
Nat. 9: 1400.
Winkler, H. (1922) Die Urticaceen Papuasiens. Beitr. Fl. Papuasien 8: 501-608.
Yahara, T. (1984) Pellionia and Elatostema in Thailand (Taxonomic studies of
Urticaceae II). J. Fac. Sci. Univ. Tokyo, Sect. 3, Bot. 13(4): 483-499.
Gardens’ Bulletin Singapore 63(1 & 2): 163-173. 2011 163
Towards a revision of Lejeunea (Lejeuneaceae) in Malaysia
G.E. Lee'*, S. Robbert Gradstein*?, A. Damanhuri! and A. Latiff'
‘School of Environmental and Natural Resource Sciences,
Faculty of Science and Technology,
Universiti Kebangsaan Malaysia, 43600, Selangor, Malaysia
*gaikee0808(@hotmail.com
> Muséum National d’Histoire Naturelle,
Department Systématique et Evolution UMS 7205,
Case Postale 39, 57 rue Cuvier.
75231 Paris cedex 05, France
ABSTRACT. As currently delimited, Lejewnea Lib. is characterised by the hyaline papilla at
the proximal side of the first tooth of the lobule: the usual absence of ocelli in the leaf lobe;
lobules occasionally reduced and with a single tooth; small, finely granular or homogeneous
oil bodies; underleaves without or with straight, upright lobes; thin-walled epidermal cells
that are larger than the medullary cells; branches of the Lejewnea-type; and gynoecia with
lejeuneoid innovations. In Malaysia, 29 species of Lejeunea have hitherto been recognised. In
the framework of an ongoing taxonomic revision, all characters used to circumscribe the genus
and the species from Malaysia have been critically assessed based on study of fresh material
collected during many field excursions throughout Malaysia (the Peninsular, Sabah, Sarawak)
as well as study of herbarium material. Some characters which were neglected previously,
such as the morphology of the first lobule tooth, superior central cells and female bracts and
bracteoles, are being critically evaluated. These characters may be useful for differentiating
Lejeunea from closely related genera, whereas characters of oil bodies, the perianth, the
lobule with a large disc cell, and underleaves with two large basal cells, are also useful for
distinguishing some species within the genus.
Keywords. Lejeunea, Lejeuneaceae, liverworts, Malaysia, taxonomy
Introduction
The pantropical genus Lejewnea Lib. is one of the largest genera in the liverwort
(Marchantiophyta) family Lejeuneaceae, containing some 150—200 species, almost
half of which occur in tropical Asia. The definition of the genus has long been
problematic and many species currently placed in Lejewnea were previously assigned
to other genera. It is also one of the most difficult and poorly understood hepatic genera.
Species of Lejewnea are small to medium-sized plants, delicate, translucent, with
shiny leaves in the field, frequently growing closely appressed on bark, living leaves
and other substrates. The highly variable species L. flava has an almost worldwide
distribution, occurring in Europe, North and South America, Africa, China, Japan,
Indomalesia, Australasia and the Pacific region, while other species are restricted only
to Asia, Europe, Africa or the Americas.
164 Gard. Bull. Singapore 63(1 & 2) 201]
In Malaysia, 29 species of Lejeunea have been reported, most of them from
Mount Kinabalu with 26 recorded species (Eifrig 1937; Mizutani 1963, 1966, 1970,
1978: Kodama 1976; Lee et al. 2010a). Eleven species are recorded from Peninsular
Malaysia, viz., L. anisophylla Mont. [= L. borneensis Steph.| (Kitagawa 1971), L. flava
(Sw.) Nees, L. patersonii (Steph.) Steph., L. cuculliflora (Steph.) Mizut. [=Taxilejeunea
cuculliflora Steph.], L. albescens (Steph.) Mizut. [= Taxilejeunea albescens Steph.]
(Inoue 1967), L. lumbricoides (Nees) Nees, L. umbilicata (Nees) Nees (Tixier 1980),
and L. patriciae Schaf.-Verw. (= L. pilifera Tixier) (Schafer-Verwimp 2001), L.
discreta, L. eifrigii, and L. sordida (Lee et al. 2010b). A revision of the genus Lejeunea
in Malaysia is currently being undertaken at Universiti Kebangsaan Malaysia by the
first author, based on examination of herbarium materials collected from this country,
mainly by Mizutani in the twentieth century (Mizutani 1966), and also specimens
from other herbaria such as SING, BOHR, HIRO, BO, CAL as well as fresh materials
coliected by the first author. Named specimens from adjacent countries are being
compared with the Malaysian taxa. Approximately 500 samples, including most of the
type specimens of Lejeunea occurring in this country, have so far been examined. The
type specimens are from the Conservatoire et Jardin botaniques de la Ville de Genéve
(G), Switzerland, Herbarium Haussknecht (JE), Germany, and the Hattori Botanical
Laboratory (NICH), Japan.
Characters delimiting the genus
Characters that differentiate the genus Lejewnea from others include the hyaline
papilla at the proximal side of the first tooth of the lobule; the usual absence of ocelli;
small, finely granular or homogeneous oil bodies; occasionally reduced lobules with
one tooth; underleaves that are entire or with 2 straight, upright lobes; branches of
the Lejeunea-type; and gynoecia with lejeuneoid innovations. Furthermore, there are
some additional characters that appear to be useful in defining the genus and which
were neglected previously, such as the morphology of the first lobule tooth, superior
central cells and female bracts and bracteoles.
First tooth. The first tooth in most species of Lejewnea is straight and the apex never
acuminate (Fig.1). The position of first tooth is usually upward-pointing, sometimes
pointing towards the stem or away from the stem. This character can be used to
distinguish Lejeunea from other genera in Lejeuneaceae such as Cheilolejeunea
(Spruce) Schiffn. and Drepanolejeunea (Spruce) Schiffn. where the first tooth is strongly
falcate and the apex occasionally acuminate, but seldom straight. Cheilolejeunea may
sometimes be confused in its habit with Lejewnea, especially in the field. The main
characters that differentiate these two genera are the position of the hyaline papilla
and the morphology of the oil bodies. However, the hyaline papilla is often collapsed
and the oil bodies rapidly evaporated in dried material. Therefore, the above described
difference in the shape of the first tooth may be used to separate these two genera, at
Lejeunea in Malaysia 165
Fig. 1. Lobule apex in some Malaysian species of Lejeunea (hyaline papilla shown in gray).
A. L. discreta Lindenb. from Damanhuri s.n. (UKMB). B. L. microloba Taylor from Kodama
40783 (NICH). C. L. sordida (Nees) Nees from G.E. Lee 1182 (UKMB). D. L. lumbricoides
(Nees) Nees from G.E. Lee 1155 (UKMB). E. L. albescens (Steph.) Mizut. from G.E. Lee 1157
(UKMB). ft: first tooth, de: disc cell.
least in Malaysia. The characters of the lobule tooth in Lejeuneaceae are discussed in
detail by He (1996).
Superior central cells. The anatomy of the underleaf base has been discussed as a
significant character in the taxonomy of the Lejeuneaceae (Bischler 1969, Winkler
1970, Gradstein 1975). The number of the superior central cells seems to be very
constant at the genus level and taxonomically relevant (Gradstein 1975). The superior
central cells can be perceived through transverse section of the underleaf base. The
origin of the superior central cells remains uncertain (Gradstein 1975); according
to Winkler (1970), they belong to the underleaf. According to Bischler (1969) and
Winkler (1970), there are two superior central cells in the genera of Lejeuneoideae
166 Gard. Bull. Singapore 63(1 & 2) 2011
(including Lejewnea). However, our studies show the presence of 4 superior central
cells in L. lumbricoides (Nees) Nees, L. eifrigii Mizut., L. discreta Lindenb. and L.
sordida (Nees) Nees (Fig. 2). This character can distinguish some Lejeunea species
from Drepanolejeunea, which has only two superior central cells (Gradstein 1975).
Female bracts and bracteoles. All the female bracts of Lejewnea have a rather short
and straight keel, and always without any wing on the keel. The margin of the bracts
and bracteoles are usually entire or slightly crenulate and seldom serrulate. The absence
of winged bracts in Lejewnea can be used to separate Microlejeunea where the latter
usually has winged or sinuate-dentate keels, with bracts and bracteoles occasionally
dentate (Schuster 1980).
100 pm
Fig. 2. Stem in cross section at the base of the underleaf, showing four superior central cells
(s). A. L. lumbricoides (Nees) Nees from G.E. Lee 1428 (UKMB). B. L. discreta Lindenb.
from Damanhuri s.n. (UKMB). C. L. sordida (Nees) Nees from G.E. Lee 1442 (UKMB). D. L.
eifrigii Mizut. from G.E. Lee 1194 (UKMB).
Lejeunea in Malaysia 167
Characters in species delimitation
Leaf habit and shape. Most of the species of Lejeunea have flat leaves when moist,
however some are with strongly convex leaves when moist, e.g., L. pectinella Mizut.,
L. umbilicata (Nees) Nees , L. lumbricoides (Nees) Nees, L. contracta Mizut. and L.
kinabalensis Mizut. (Fig. 3). The convex leaf in the moist condition only seems to
appear in more robust species of this genus. This character can be used to separate L.
Fig. 3. Leaf habit and shape. A. L. sordida (Nees) Nees from G.E. Lee 1154 (UKMB). B. L.
micholitzii Mizut. from Z. Iwatsuki 1383a (NICH). C. L. discreta Lindenb. from G.E. Lee
1146 (UKMB). D. L. alata Gott. from G.E. Lee 1199 (UKMB). E. from the holotype of L.
kinabalensis Mizut. (NICH). F. L. pectinella Mizut. from G.E. Lee 1037 (UKMB). G. L.
lumbricoides (Nees) Nees from G.E. Lee 1155 (UKMB). H. L. patriciae Schaf.-Verw. from
G.E. Lee 1099 (UKMB).
168 Gard. Bull. Singapore 63(1 & 2) 2011
pectinella Mizut. from L. discreta Lindenb., the former usually having convex leaves
when moist (Fig. 3). Leaf shape in Lejewnea varies from narrowly to broadly ovate to
rounded. The base is often gradually narrowed to the insertion which forms a long,
J-shaped outline, along 10—15 cells. The apex is usually acute to rounded (Fig. 4),
rarely apiculate except in L. eifrigii Mizut. and L. microloba Taylor, and without the
marginal rhizoids as in L. patriciae Schaf.-Verw., which has leaves with 5—10 short
rhizoids protruding from the apical margin.
Fig. 4. Leaf morphology. A. L. sordida (Nees) Nees from G.E. Lee 1154 (UKMB). B. L.
pectinella Mizut. from G.E. Lee 1037 (UKMB). C. L. wightii Lindenb. from G.E. Lee 1183
(UKMB). D. from the holotype of L. kinabalensis Mizut. (NICH). E. L. anisophylla Mont.
from G.E. Lee s.n. (UKMB). F. L. discreta Lindenb. from G.E. Lee 1146 (UKMB). G. L.
eifrigii Mizut. from G.E. Lee 1185 (UKMB). H. L. patriciae Schaf.-Verw. from G.E. Lee 1099
(UKMB).
Lejeunea in Malaysia 169
Oil bodies. Generally, Lejeunea has small, finely granular and homogenous oil bodies.
The number of oil bodies per cell and the shape of the oil bodies varies among the
species in this genus, e.g., L. /umbricoides (Nees) Nees has more than 8 oil bodies
per cell, usually ovoid, rarely ellipsoid, slightly glistening, whereas L. eifrigii Mizut.
has less than 8 oil bodies per cell, normally 3—S per cell, usually ellipsoid, sometimes
ovoid, and opaque (Fig. 5). However, most of the species, e.g., L. discreta Lindenb.
and L. patriciae Schaf.-Verw. have more than 8 oil bodies per cell, ovoid to ellipsoid,
which are glistening to somewhat opaque.
Perianth. The perianth in most species has 5 keels. However, perianths without keels
can also be found as in L. albescens (Steph.) Mizut., L. cuculliflora (Steph.) Mizut.,
L. microloba Taylor and L. umbilicata (Nees) Nees. The eplicate perianth is rather
a widespread character in Lejeunea. Reiner-Drehwald and Schafer-Verwimp (2008)
treated 13 species of Lejeunea with eplicate perianth, occurring in America, Europe
Fig. 5. Oil bodies. A. L. patriciae Schaf.-Verw. from G.E. Lee 1099 (UKMB). B. L.
lumbricoides (Nees) Nees from G.E. Lee 1184 (UKMB). C. L. discreta Lindenb. from G.E.
Lee 1423 (UKMB). D. L. eifrigii Mizut. from G.E. Lee 1194 (UKMB). E. Lejeunea sp. from
G.E. Lee 1424 (UKMB). F. L. wightii Lindenb. from G.E. Lee 1423 (UKMB). Scale bars: A,
B = 50 pm; C—F = 100 pm.
Gard. Bull. Singapore 63(1 & 2) 2011
170
Fig. 6. The perianth of Lejewnea. A. L. cuculliflora (Steph.) Mizut. from P. ¥. Wong with Kim
Wong 1604 (NICH). B. L. mimula Hiirl. from M. Mizutani 3239 (NICH). C. L. papilionacea
Steph. from Damanhuri s.n. (UKMB). D. L. eifrigii Mizut. from G.E. Lee 1185 (UKMB). E. L.
pectinella from Damanhuri s.n. (UKMB). F. L. anisophylla Mont. from G.E. Lee s.n. (UKMB).
Scale bars: A—F = 0.2 mm.
Lejeunea in Malaysia 171
and Africa. The apex of the perianth also plays an important role in separating some
of the taxa. A funnel-shaped apex is found only in L. mimu/da Hiirl. and an apex with a
5—7-cells-long beak occurs only in L. pectinella Mizut. (Fig. 6).
Lobule with a large disc cell. In some species a large rectangular cell, called the disc
cell by Mizutam (1970), very much larger than the first tooth, is situated below the first
tooth (Fig. 1). Species with this peculiar character are L. lumbricoides (Nees) Nees
and L. discreta Lmdenb. Most species, however, lack such a large rectangular cell and
sometimes this cell is of the same size as the first tooth, viz., in L. albescens (Steph.)
Mizut., L. microloba Taylor, and L. sordida (Nees) Nees.
Underleaf with two large basal cells. This feature has been used to separate the species
of Lepidolejeunea and Luteolejeunea by Piippo (1986). In the majority of Lejeunea
species these two large basal cells are present and easily distinguished (Fig. 7). The
presence of these two large basal ceils was found in L. lumbricoides (Nees) Nees,
L. microloba Taylor, L. umbilicata (Nees) Nees and L. eifrigii Mizut. but not in L.
sordida (Nees) Nees where the two large basal cells are undistinguishable.
Fig. 7. Underleaf with two large basal cells (bc). A. L. lumbricoides (Nees) Nees from G_E. Lee
1428 (UKMB). B. L. microloba Taylor from Kodama 40783 (NICH). C. L. umbilicata (Nees)
Nees from M. Mizutani 3769 (NICH). D. L. sordida (Nees) Nees from G.E. Lee 123] (UKMB).
E. L. eifrigii Mizut. from G_E_ Lee 1168 (UKMB).
7 Gard. Bull. Singapore 63(1 & 2) 2011
Outlook
Our taxonomic study suggests that some species of Lejeunea are very similar and
might lead to some new synonymy. For example, herbarium materials of L. pectinella
Mizut. and L. mizutanii Grolle from Malaysia are very similar, L. dipterocarpa E.W.
Jones from West Africa (Jones 1972), L. hui R.L. Zhu from China (Zhu & So 2001)
and L. kashyapii M. Dey, D.K. Singh & D. Singh from India (Dey et al. 2008) are
almost inseparable from L. papilionacea Steph. The last species was considered an
African species but Sdderstrém et al. (2010) have recently reported this species from
Java where it was described as Cardiolejeunea cadiantha Schust. & Kachroo. If the
above-mentioned species are indeed conspecific, it would seem that L. papilionacea
Steph. is actually widespread in Asia. Furthermore, the difference between L. wightii
Lindenb. and L. tuberculosa Steph., and between L. cuculliflora (Steph.) Mizut. and
L. umbilicata (Nees) Nees, 1s probably not sufficient to warrant species distinction
for these pairs of taxa. We are now trying to solve these problems by studying the
type specimens and additional materials. We also expect that further new additions
to Malaysia will be discovered in the future. Finally, we anticipate the separation
of the genus Microlejeunea from Lejeunea and the treatment of the Asiatic Lejeuna
punctiformis Taylor as a species of Microlejeunea. This species differs from Lejeunea
species in the stem which has only 3 medullary cells in transverse section, leaf lobe
with | or 2 ocelli at the base, a very large lobule, the first lobule tooth being rather
long and curved, the keel of the female bracts winged and occasionally dentate female
bracteoles. These characters also serve to separate Microlejeunea from Lejeunea (see,
e.g., Bischler et al. (1962) for a thorough discussion of their differences).
ACKNOWLEDGEMENTS. We would like to thank all the directors and curators of G, JE,
STR, NICH, SING, BO, CAL, HIRO and BORH for the loan of herbarium specimens. This
study is supported financially by the National Science Foundation (NSF), Malaysia and a Dana
Operasi UKM-OUP research fund awarded to Emer. Prof. Dato’ Abdul Latiff Mohamed.
References
Bischler, H. (1969) Notes sur ’'anatomie des amphigastres et sur le development du
paramphigastre et des rhizoids chez Drepano-, Rhaphido- et Lepto-lejeunea. Rev.
Bryol. Lichénol. 36: 45—S5.
Bischler, H., Miller, H.A. & Bonner, C.E.B. (1962) Studies in Lejeuneaceae IV: The
typification of the genus Microlejeunea. Nova Hedwigia 4: 173-187.
Dey, M., Singh, D.K. & Singh D. (2008) Two new species of Lejeunea Lib. (Hepaticae:
Lejeuneaceae) from Sikkim, India. J. Bryol. 30: 126—132.
Eifrig, H. (1937) Monografischen Studien tiber die Indomalayischen Arten von
Taxilejeunea. Ann. Bryol. 9: 73-114.
Lejeunea in Malaysia WS)
Gradstein, S.R. (1975) A taxonomic monograph of the genus Acrolejeunea. Bryophyt.
Biblioth. 4: 1-162.
He, X-L. (1996) On the taxonomic significance of lobule characters in the Leyjeuneaceae
(Hepaticae). Ann. Bot. Fennici 33: 311-316.
Inoue, H. (1967) Studies on oil bodies of some Malayan liverworts. J. Hattori Bot.
Lab. 30: 54-70.
Jones, E.W. (1972) African hepatics XXIII. Some species of Lejeunea. J. Brvol. 7:
23-45.
Kitagawa, N. (1971) Asmall collection of Hepaticae from Penang, Malaysia-continued.
Bull. Nara U. Educ. 20(2): 7-14.
Kodama, T. (1976) Three new species of Lejeuneaceae from Kinabalu. J. Hattori Bot.
Lab. 41: 381-387.
Lee, G.E., Damanhuri, A., Latiff, A. & Gradstein, S.R. (2010b) A taxonomic treatment
of Lejeunea discreta, L. eifrigii and L. sordida, new to Peninsular Malaysia. Acta
Biol. Pl. Agriensis 1: 37-S1.
Lee, G.E., Pécs, T., Damanhuri, A. & Latiff, A. (2010a) Lejeunea gradsteinii
(Lejeuneaceae), a new liverwort species from Mt. Kinabalu, Sabah. Acta Biol. Pl.
Agriensis 1: 29-36.
Mizutani, M. (1963) On some Indian species of the family Lejyeuneaceae described by
W. Mitten. J. Hattori Bot. Lab. 26: 171-184.
Mizutani, M. (1966) Epiphyllous species of Lejeuneaceae from Sabah (North Borneo).
J. Hattori Bot. Lab. 29: 153-170.
Mizutani, M. (1970) Lejeuneaceae, subfamilies Lejeuneoideae and Cololejeuneoideae
from Sabah (North Borneo). /. Hattori Bot. Lab. 33: 225-265.
Mizutani, M. (1978) Lejeuneaceae from Ishigaki and Iriomote Islands of Ryukyu
Archipelago. J. Hattori Bot. Lab. 44: 121-136.
Piippo, S. (1986) A monograph of the genera Lepidolejeunea and Luteolejeunea
(Lejeuneaceae, Hepaticae). Acta Bot. Fennica 132: 1-69.
Reiner-Drehwald, M.E. & Schafer-Verwimp, A. (2008) On /nflatolejeunea, Lejeunea
species with eplicate perianths and L. talamancensis sp. nov. from Costa Rica
(Lejeuneaceae). Nova Hedwigia 87: 387-420.
Schafer-Verwimp, A. (2001). On Lejewnea patriciae, nom. nov. for Lejeunea pilifera
Tixier. Candollea 56: 63-67.
Schuster, R.M. (1980) The Hepaticae and Anthocerotae of North America East of the
Hundreth Meridian, vol. 1V. New York: Columbia University Press.
Séderstr6m, L., Gradstein, S.R. & Hagborg, A. (2010) Checklist of hornworts and
liverworts of Java. Phytotaxa 9: 53-149.
Tixier, P. (1980) Contribution to the bryological knowledge of Malaysia. III. Cameron
Highlands-mosses and epiphyllous liverworts. Nova Hedwigia 32: 377-392.
Winkler, S. (1970) Zur Anatomie der Rhizoidplatten neotropischer Lejeuneaceen. Rev.
Bryol. Lichénol. 37: 47-SS.
Zhu, R.L. & So, M.L. (2001) Epiphyllous liverworts of China. Nova Hedwigia Beiheft
121: 126-162.
Gardens’ Bulletin Singapore 63(1 & 2): 175-188. 2011 N72)
The study of larger basidiomycetes, especially polypores,
in the Malesian region
and the role of the Singapore Botanic Gardens
Vincent Demoulin
Institut de Botanique, B.22, Université de Liege,
B-4000 Liege, Belgium
V.Demoulin@ulg.ac.be
ABSTRACT. The development of the study of larger basidiomycetes, especially polypores, in
the Malesian region is presented. The historical importance of the botanical gardens in Bogor
and Singapore is emphasised and an overview of the mycological collection in Singapore
is given. This includes several isotypes of taxa described by G.E. Massee, C.G. Lloyd and
N. Patouillard, as well as paratypes and holotypes of taxa described by E.J.H. Corner. The
problems linked to Corner’s material are discussed in the light of studies made in both Singapore
and Edinburgh. The polypore collection in Singapore is a valuable resource for studying any
geographical variation of fungal floras in the Malesian region and a unique tool for examining
any temporal change in this flora, given the continuity of collections in the island since H.N.
Ridley in 1892.
Keywords. Basidiomycetes, Malesia, polypores, Singapore Botanic Gardens
Introduction
The Malesian region in the sense of Flora Malesiana is one of the botanically most
interesting tropical regions. Progress in the study of its vascular flora has been
important but our knowledge of its mycological flora is still uneven, with some areas
and some groups better sampled than others. The most intensively studied area is
probably Peninsular Malaysia and Singapore and the best studied group is the larger
basidiomycetes. This is linked to the work of E.J.H. Corner, based in Singapore in the
thirties and forties. The importance of the mycological collection of the Singapore
Botanic Gardens (SING, here and elsewhere, acronyms designing herbaria follow
Holmgren et al. 1990) as well as that of E.J.H. Corner is emphasised. My experience 1s
based on several stays in the Singapore Botanic Gardens and a recent one in the Royal
Botanic Garden Edinburgh (E), the two institutions that host the majority of Corner’s
material. I also examined duplicates in the Royal Botanic Gardens Kew (K).
Mycology in the Malesian region
During the nineteenth and early twentieth centuries, mycological prospecting in the
Malesian region was, except for some expeditions or cruises, centred on the two
176 Gard. Bull. Singapore 63(1 & 2) 2011
gardens linked to the British and Dutch presence in the region, Singapore and Bogor.
However, no well-known specialist in mycology worked there until K.B. Boedijn
arrived in Sumatra in 1926, moving in 1928 to the Bogor (then Buitenzorg) garden,
and Corner in Singapore in 1929.
In mycological journals one will find biographies of those two scientists
by Donk (1965) for Boedijn, and Watling (2001b) and Watling & Ginns (1998) for
Corner. A detailed account of Corner’s contribution to Malaysian mycology has been
published by Watling in 2007.
Before the period of Boedijn and Corner, mycological collections were mostly
made by general botanists and referred to overseas specialists. From the time of H.N.
Ridley, directors or assistant directors of the Singapore Botanic Gardens have thus sent
many collections to Kew, but also to N. Patouillard and especially C.G. Lloyd. The
provided identifications form the basis of T.F. Chipp’s catalogue of fungi of the Malay
Peninsula (1921).
After the disruption of World War II, the situation diversified with more and
more research performed in local institutions. This is especially the case at the Bogor
herbarium with M. Rifai and the Forest Research Institute of Malaysia with S.S. Lee. In
both institutions, young researchers are now active and mycology is really developing
locally. This is accompanied by occasional involvement of foreign monographers, for
whom the speed of air travel makes short visits to the tropics easier. The publication of
nicely illustrated popular books like those of Tan (1990), Pegler (1997) and Zainuddin
et al. (2010) will certainly develop the interest of the general public for fungi.
The state of development can be found in publications like the recent book
edited by Jones, Hyde & Vikineswary (2007), the paper by Ahmad (1986) and the
Checklist of literature on Malaysian macrofungi by Lee et al., which is only available
on the internet (www.chm.frim.gov.my/Checklist_final.pdf).
The Singapore Botanic Gardens, which has been pivotal in those developments,
does not have at the moment a resident mycologist but gives excellent facilities to
visitors and curates meticulously a large mycological herbarium. Its importance for the
study of fungi in the Malaysian Region should be emphasised and will be the object
of the following section.
The mycological collection of the Singapore Botanic Gardens (SING)
The oldest fungal material in the herbarium of the Singapore Botanic Gardens goes
back to H.N. Ridley and dates as far back as 1892. After 1900 one sees collections by
I.H. Burkill and his wife E.M. Burkill, C.F. Baker and especially T.F. Chipp. Local
collectors were Kiah, Nur and Sappan. When he replaced Chipp, R.E. Holttum was
also active in mycological collecting. He mostly left this activity to Corner when the
latter joined the staff, but kept making collections until the early thirties.
Many of those collectors regularly sent part of their material to Kew and to
C.G. Lloyd. Massee, Lloyd and Patouillard have all described new species on the basis
of these collections. The holotypes are of course in the herbarium of the describing
Larger basidiomycetes in the Malesian Region 7
author (K for Massee, BPI for Lloyd, FH for Patouillard) but it is usually overlooked
that the isotype kept in Singapore can be a large collection from which a relatively
small part has been sent. This isotype can thus be precious for estimating the variability
of the type collection or in case of loss or damage to the holotype, as one can judge
from the illustrations of this paper.
An example of a work by Massee on Singapore fungi is his Fungi exotici
XVII (1914), entirely devoted to a mailing from Mrs. Burkill, with the description of
17 new species. A paper entirely devoted to Singapore boletes by Patouillard & Baker
(1918) describes 16 new species. Some are discussed by Singer (1981), who revised
their holotypes and a full discussion with type analysis is done by Watling (2001a).
An example of those species is Boletus spinifer Pat. & C.F.Baker, whose holotype was
also studied in detail by Heinemann & Rammeloo (1982). They noted a difference in
pores from the original description and synthesise the numerous discussions of this
unusual fungus. None of the authors who discussed this species seem to have thought
of revising Baker’s isotype in SING (no. 0036203). This could give more insight into
the variability of the collection on characters like the pores.
It was only during 1917 that Baker was assistant-director of the Singapore
Botanic Gardens, before moving to the Philippines. His collaboration with Patouillard
was nonetheless fruitful, for besides the boletes, Patouillard (1922) based two new
species of polypores on material sent by Baker. For one, Phellinus chaetoloma Pat.,
the isotype (SFN 5409) is important, for the holotype at FH has been determined as
Phellinus contiguus (Pers.:Fr.) Pat. by Ryvarden (1983), while the original description
does not fit that species, the setal structures in the hymenium being described as
obtuse. This was noted by Corner (1991, p. 119) who examined the Singapore isotype
and determined it as a resupinate Phellinus noxius (Corner) G.Cunn. I concur with this
determination. Corner suggested the specimens have been muddled. It would thus be
interesting to compare the two parts of the collection to confirm this. If Ph. chaetoloma
is really the same as Ph. noxius, it will be necessary to conserve this well known,
phytopathologically important name against Ph. chaetoloma, as well as if necessary
against Ph. sublamaensis Lloyd, which was considered as a prior name for Ph. noxius
by Ryvarden (1989).
The second species was Phaeolus iobaphus Pat., the isotype of which is also
worth detailed study. Here also there is a discrepancy between the original description,
that gives the spores as smooth, and the description by Ryvarden, that gives them
as finely asperulate. My preliminary impression is that the collection could be
heterogeneous (at least the SING part) and that in this SING part the asperulate spores
could be a contamination.
On the basis of his type examination, Ryvarden combined the epithet in
Wrightoporia. It 1s not clear whether Corner (1989a, p. 121) did look at the type
himself, or relied on Ryvarden’s description, but he made a further combination in
Stecchericium.
Several Singapore botanists, the last being R.E. Holttum, and the most prolific
T.F. Chipp, have sent material, especially of polypores to C.G. Lloyd. Lloyd described
several species on the basis of those collections, now housed in BPI. The polypore
178 Gard. Bull. Singapore 63(1 & 2) 2011
holotypes have been revised by Ryvarden (1989, 1990, 1992). The isotypes preserved
in SING are however worthwhile studying for there can be a large collection of which
a small part only was sent to Lloyd. Many collections which are not types are also
interesting for understanding Lloyd’s concepts and can be quite helpful for comparison
when one is working in the Singapore herbarium (Fig. 1).
I did spot 18 isotypes of Lloyd’s polypore species, some of which will be the
object of separate publications. As an example (see also Fig. 2-4) of the interest of this
material, I will discuss the case of Hexagonia angulata Lloyd and H. umbrosa Lloyd.
The two species were published at a short interval in 1920 as follows: H. umbrosa
Lloyd, Mycol. Writ. VI, Mycol. Not. 63 (May 1920) 957, with no figure; and H.
angulata Lloyd, Mycol Writ. VI, Mycol. Not. 64 (Sept. 1920) 1003, fig. 1831.
The type of H. umbrosa was not explicitly cited in the original description, but
the collection “Blakang Mati, 24 Dec. 1919. On dead mangrove below high tide mark,
T.F. Chipp. Singapore Field Number 5460”, present number 31697, is accompanied by
a note by Lloyd “5460 Hexagona umbrosa. This is the second collection you have sent.
The former was referred to Hexagona tenuis but this dark plant should have a name.”
This is a clear link to the publication, where it is stated: “This 1s the second collection
of this dark umber plant received from Mr Chipp...”
Te.
'
PWGAPORE BOTANIC GARDENS
ew Titi
iMG 0032631
Fig. 1. Echinoporia hydnophora (Berk. et Broome) Ryvarden sub Echinodia theobromae Pat..
Singapore, Botanic Gardens, on Quercus, 9 Dec 1919, Ahmad, S.F.N. 5143, new number SING
32631, is an example of an important historical collection in the Singapore herbarium, even
if not a type. Parts of this collection were sent by T.F. Chipp to Kew and to Lloyd. Patouillard
got a piece through Kew and sent to Singapore the letter reproduced here. Lloyd also reported
his comments (not reproduced here), which were published in Mycol. Writ. VI, Mycol. Not. 62
(Jan. 1920) 934-935, fig. 1704, 1705. This is the collection that convinced both Patouillard and
Lloyd that Echinodia theobromae (a name with a complex nomenclature outside the scope of
this paper) was indeed the conidial stage of a polypore.
Larger basidiomycetes in the Malesian Region 179
= ak at oS ae
Fig. 2. Isotype of Daedalea ridleyi Lloyd, Mycol Writ. VI, Mycol. Not. 62 (Jan. 1920) 930,
fig. 1689 at SING. The 10 fruitbodies have been mounted on two sheets. This is a single
collection (Singapore, Bukit Timah, 27.8.1900, H.N. Ridley, S.F.N. 4920) but later the two
sheets have been given different new numbers (31717 and 32628). One has been placed in the
type collection and not the other. They should be reunited in the type collection and worth a
detailed study. The holotype was identified “Lenzites acuta Berk.” by Ryvarden (1989), but
despite Lloyd’s comment that it is not related, I believe it is Daedalea sprucei Berk. or the
related (if distinct) D. Jangkawiensis Corner, of which the type is also present in SING.
The isotype of H. angulata was annotated “Pulau Penang, Waterfall Gardens,
23 jan. 1920, M. Noor, coffee brown on dead wood. Singapore Field Number 5604”,
present number 31696. It fits well the picture 1831 of Lloyd.
Those two collections were studied by Corner (1989a, pp. 20-21), while
Ryvarden published his revision of the holotypes stored in BPI, in the same year.
Ryvarden placed the two names in synonymy with Hexagonia tenuis (Hook.)
Fr., while Corner (1989b) synonymised H. angulata with H. umbrosa, which he
considered distinct from H. tenuis, treated in Trametes (invalid publication).
I personally agree that the two collections belong to a single species, distinct
from H. tenuis by being thicker and with bigger (about 1.5 mm) pores. I however do
not consider the species as far away from H. tenuis as Corner thought. He gave great
importance to the dimitic or trimitic nature of the context. There is however a large
variation in the abundance of ramifications of thin hyphae, which already appears in
Corner’s description and I feel he overemphasised the character.
180 Gard. Bull. Singapore 63(1 & 2) 2011
Wi era done free
Pht,
Singapore
Botanscal te vor
Locality ee aL
Elevation > 4 art
Fig. 3. Isotype of Fomes oroniger Lloyd, Mycol. Writ. VII, Mycol. Not. 73 (Oct 1924) 1330,
fig. 3048 (with Note 74) at SING. Malay Peninsula, Pahang, Fraser Hill, Elev. about 4000
ft., 12.1.1923, R.E. Holttum, SFN 11346 (later renumbered 71318). Lloyd already relied on
Holttum for the observation of the specimens retained in Singapore. The collection made up
of three nice big fruitbodies was well annotated by Y. Abe in 1990, but I am not aware of a
publication. The holotype was identified by Ryvarden (1989) as Phellinus pachyphloeus (Pat.)
Pat., but the present isotype was cited as Ph. melanodermus (Pat.) Fidalgo by Corner (1991, p.
112), who uses different characters to differentiate the two species.
Fig. 4. Isotype of Polystictus roseoporus Lloyd, Mycol. Writ. VI, Mycol. Not. 73 (Oct 1924)
1331, fig. 3053 (with Note 74) at SING. Malay Peninsula, State of Johore, Ulu Kahang, Elev.
450 ft., 2.6.1923, R.E. Holttum, SFN 10939. The identity with Microporus affinis (Blume &
Nees : Fr.) Kuntze proposed by Ryvarden (1992) after examination of the holotype is not
problematic, but one can see the abundance and variability of specimens in the isotype.
A problem that however remains, is whether H. umbrosa 1s distinct from H.
umbrinella Fr. The group of H. tenuis is certainly in need of revision and the well
preserved isotypes at SING should be part of it.
Corner’s mycological collections
As emphasised by Watling (especially 2007), intensive collecting by E.J.H. Corner has
been fundamental in the study of mycological diversity in the Malesian region, as well
as for mycological taxonomy as a whole.
Working with Corner’s material is, however, not straightforward, as I have
discovered after years of studies, initiated when R. Kiew, then curator of the Singapore
herbarium, asked me to assess the significance of numerous collections stored in the
Singapore herbarium, collections which had been overlooked for many years.
182 Gard. Bull. Singapore 63(1 & 2) 2011
I concentrated my work on polypores, a group for which I had the necessary
expertise, and especially well represented in the Singapore herbarium, which I visited
several times from 2004 to 2010. At that time, the aim was to reorganise the polypore
collection in an easily accessible way, made possible through two fellowships of
the Singapore Botanic Gardens, with the support of the Belgian National Science
Foundation (FNRS), as well as the general resources of my laboratory.
I have thus acquired a good idea of what material Corner had left in Singapore
and could compare it to the material in his personal herbarium kept at the Royal
Botanic Garden Edinburgh (E), as well as to the duplicates in Kew (K). Those visits to
Great Britain were made possible by a grant for a sabbatical year made by the FNRS.
Most authors have taken for granted that Corner’s types were systematically
to be found in his personal herbarium at E, as if when leaving Singapore after the war,
he had taken with him every interesting collection he had made. This is not the exact
situation, and if it is true that a majority of types are now hosted in E, several paratypes
and holotypes are to be found in SING. The great number of types kept in E makes
the revision by Hattori, whose publication started in 2000, of foremost interest. My
intention is to complement it with the Singapore types, through this paper and some
subsequent ones.
That some holotypes of Corner’s names should be in SING should be apparent
to an attentive reader of his “Ad Polyporaceas” (1983-1991), where the mention “herb.
Singapore” occurs beside the mention “herb. Corner”.
It also happens that the institution or herbarium where the type is conserved
was not indicated, and for types, this makes the new name invalid in the case of
the Xanthochroic Polypores, Ad Polyporaceas VII. This last volume of the series
was published in 1991, while Art. 37.7 of the ICBN (McNeill et al. 2006) makes
such indication of the localisation of the type of a new name mandatory after 1990.
An example of such an invalid name is Phellinus glaucescens (Petch) Ryvarden var.
minor Corner (1991, p.93). If anybody wanted to validate that name, which is not
my intention, the type is apparently the no. 28266 in SING, without corresponding
material in E.
One should note that, despite its title, the publication by Corner in 1993, does
not give indications on the status of its collections. It 1s a synthesis of the points of
polypore taxonomy on which he disagreed with contemporary mycologists.
Most of Corner’s collections in SING are from the years 1929-1932. This is,
however, not exclusively so. There are collections from 1929 in E and from 1941 in
SING. There is some difference in the proportion of material in E and SING by genera.
I have only seen a single Corner collection of Amauroderma in SING (A. rugosum,
Botanic gardens, 20/4/1932). This is not one of the ten bar-coded collections in E. The
same situation occurs with Zrametes, richly represented in E, but for which I have seen
only three collections in SING. Those are of 7: persoonii (Mont.) Pat. (later reduced
to synonymy of 77. scabrosa (Pers.) G. Cunn.), from 1929,1930 and 1941 and are not
recorded in E.
Larger basidiomycetes in the Malesian Region 183
In contrast, the Singapore herbarium is rich in Phellinus, under that name, or
more often Fomes. It is in that genus that Corner described his first polypores in 1932.
Those were:
— Fomes levigatus Corner, Trans. Br. Mycol. Soc. 17 (1932) 52. Later renamed
Phellinus leiomitus Corner, Beih. Nova Hedwigia 101 (Ad Polyporaceas VII)
(1991) 108, by reason of homonymy in the genus Phellinus.
— Fomes senex (Nees et Mont.) Imazeki var. bu/bosetosus Comer, ibid.: 75.
— Fomes senex (Nees et Mont.) Imazeki var. hamatus Corner, ibid.: 75.
— Fomes lamaensis (Murrill) Pat. var. secedens Corner. Gard. Bull. Straits Settlem.
5 (1932) 341.
— Fomes noxius Corner, ibid.: 342.
— Fomes pachyphloeus Pat. var. hispidus Corner, ibid.: 347.
The holotypes of all these taxa are in SING (see, e.g., Fig. 5 and 6) and I have not
located any material in E.
It may be that Corner had made enough progress in his study of Phellinus
(“Fomes’’) not to consider it necessary to take much material with him when he left
Singapore, especially given the fact that these are among the bulkiest polypores.
The type of Phellinus noxius (Corner) G.Cunn. is an especially noteworthy
collection. It seems to have been split with one part placed in the type collection,
receiving, in addition to the no. 25750, a general number 31727 (Fig. 6), while another
part of 25750 is in the general herbarium, despite the mention in Corner’s handwriting
“Fomes noxius Corner. Type”. The two parts should be reunited to better appreciate
the variability of the collection. Indeed the type material is more dimidiate than one
would guess from the original description. The importance of this collection comes
from the possible conflict with previous, less well known names, like Phellinus
chaetoloma Pat. and Fomes sublamaensis Lloyd. Phellinus noxius is a pathogen of
major economic importance in the tropics, with 741 hits in Google Scholar, against 4
for Ph. sublamaensis (Lloyd) Ryvarden. Even if the synonymy with Ph. sublamaensis
does not seem to be accepted anymore, sublamaensis being apparently a synonym of
Ph. lamaensis (Murrt.) Pat., conservation against Ph. chaetoloma might be necessary
as stated earlier.
Corner has mostly used the collections he had made as types of his new names;
however, he sometimes also used material of the Singapore herbarium collected by
other people. For example, the holotype of /nonotus perchocolatus Corner, Beih.
Nova Hedwigia 101 (Ad Polyporaceas VII) (1991) 123 was collected by Kiah. The
holotype “Singapore, Dalvey Road, 14 May 1920, leg. Kiah, Sing. F.N. 5714” has
received the new number 32830 and is what most authors would consider a Phellinus
in need of further examination and not an /nonotus. In the same publication, p. 124, is
also described a var. parvisporus of this species, with type “Malaya, Johore, leg. R.E.
Holttum s.n. Jul. 1931; herb. Singapore (ut “Poria 11a’)”. This is also present in SING
and has received the number 28254, probably later than the study by Corner.
184 Gard. Bull. Singapore 63(1 & 2) 2011
FLORA OF SINGAPORE.
REGISTERED
ENTERED IN
CARD-INDEX -
Fig. 5. Holotype of Fomes /evigatus Corner, Trans. Br. Mycol. Soc. 17 (1932) 52 at SING, later
renamed Phellinus leiomitus Corner, Beih. Nova Hedwigia 101 (1991), by reason of homonymy
in the genus Phellinus. This taxon is usually ignored in recent treatments of Phellinus.
Larger basidiomycetes in the Malesian Region 18
FLORA OF SINGAPORE,
, No. asgter
olen Fe, ome YO Comte
Where Catlected ts Bean <i. 3 See
E < Blooation Datei sae.” 1 Ha |
CARD-INDEX ie. SR nae |
OF : oe abate 20 Fy ow VRS filer :
K. he é Collector: F
pst Awe Gua 2 ws, ea ; ' :
Fig. 6. Holotype of Fomes noxius Corner, Gard. Bull. Straits Settlem. 5 (1932) 341 at SING
Phellinus noxius (Corner) G.Cunn., a species of major phytopathological importance in the
tropics. This is the part of the collection SFN 25750 which has received the new number 31727
and has been incorporated in the type collection. Another part of this collection is still in the
general herbarium.
nN
186 Gard. Bull. Singapore 63(1 & 2) 2011
Perspectives for the future
Further developments in the study of the fungal flora of the Malesian region will
certainly take place in several centres, especially the Bogor Herbarium and the Forest
Research Institute Malaysia. For Singapore, interesting results will shortly come from
the use of the rich herbarium.
The polypore collection in the Singapore herbarium, when reordered, will be
an excellent tool for a model study of the fungal flora of the Malesian region. This is
due to the large number of specimens that allow the study of the variability of species,
and the presence of reference material, illuminating the concepts of authors like Lloyd
and Corner.
It can be the basis of comparisons with distant areas, like Papua New Guinea,
for which the Belgian herbaria GENT and LG hold numerous collections (Quanten
1997). This comparison would tell us if the polypore flora varies between the
westernmost and easternmost parts of the Flora Malesiana territory.
It also presents a unique opportunity to follow floristic evolution in time. With
areas like the Gardens’ jungle or Bukit Timah regularly studied since Ridley’s time at
the end of the 19th century, one may get an indication of any change, linked to man’s
influence or climate, in the fungal flora of a tropical region. This, I believe, is unique
in the world.
ACKNOWLEDGEMENTS. My work in Singapore has been made possible by two fellowships
of the Singapore Botanic Gardens in 2005 and 2007 and the support of the Belgian Science
Foundation (FNRS, now FRS) for the stay in 2004 (contract 2.4551.99.F) and 2009 and 2010,
in the framework of a sabbatical year, that also allowed me to visit the herbaria of Kew and
Edinburgh. During my visits in Singapore I always found excellent facilities provided by the
curators, R. Kiew, who drew my interest to the Corner collections, and B. Tan, as well as
continuous help from J. Leong-Skornickova and the very efficient herbarium manager S. Lee.
In Kew (B. Spooner and B. Aguirre-Hudson) and Edinburgh (R. Watling and E. Haston) the
hospitality was also excellent and I wish I will keep visiting these three great herbaria.
References
Ahmad, N. (1986) Mycological research in Malaysia. Proc. Indian Acad. Sci., Pl. Sci.
96: 363-371.
Chipp, T.F. (1921) A list of the fungi of the Malay Peninsula. Gard. Bull. Straits
Settlem. 2: 311-418.
Corner, E.J.H. (1983) Ad Polyporaceas I. Beihefte zur Nova Hedwigia 75. Vaduz,
Liechtenstein: J. Cramer.
Corner, E.J.H. (1984) Ad Polyporaceas II & III. Beihefte zur Nova Hedwigia 78.
Vaduz, Liechtenstein: J. Cramer.
Corner, E.J.H. (1987) Ad Polyporaceas IV. Beithefte zur Nova Hedwigia 96. Berlin
and Stuttgart: J. Cramer.
Larger basidiomycetes in the Malesian Region 187
Corner, E.J.H. (1989a) Ad Polyporaceas V. Beihefte zur Nova Hedwigia 96. Berlin
and Stuttgart: J. Cramer.
Corner, E.J.H. (1989b) Ad Polyporaceas VI. Beihefte zur Nova Hedwigia 97. Berlin
and Stuttgart: J. Cramer.
Corner, E.J.H. (1991) Ad Polyporaceas VII. Beihefte zur Nova Hedwigia 101. Berlin
and Stuttgart: J. Cramer.
Corner, E.J.H. (1993) Ad Polyporaceas-Explicanda. Nova Hedwigia 57: 143-157.
Donk, M.A. (1965) The mycological publications of K. B. Boedijn. Persoonia 3: 325—
330, pl.4.
Hattori, T. (2000) Type studies of the polypores described by E. J. H. Corner from Asia
and the West Pacific I. Species described in Polyporus, Buglossoporus, Meripilus,
Daedalea, and Flabellophora. Mycoscience 41: 339-349.
Heinemann, P. & Rammeloo, J. (1982) Les bolets a sétules. Bull. Jard. Bot. Natl. Belg.
52: 477-482.
Holmgren, P.K., Holmgren, N.H., & Barnett, L.C. (1990) Index Herbariorum/ Part I:
The Herbaria of the World, Sth ed. New York: New York Botanical Garden.
Jones, E.B.G., Hyde, K.D. & Vikineswary, S. (2007) Malaysian Fungal Diversity.
Kuala Lumpur: Univ. of Malaya & Ministry of Natural Resources and Environment
Malaysia.
Massee, G.E. (1914) XII. Fungi exotici: XVII. Kew Bull. 1914: 72-76.
McNeill, J., Barrie, F.R., Burdet, M., Demoulin, V., Hawksworth, D.L., Marhold, K.,
Nicolson, D.H., Prado, J.P., Silva, P.C., Skog, J.E., Wiersema, J.H., N. Turland,
N.J. (eds) (2006) International Code of Botanical Nomenclature (Vienna Code)
adopted by the Seventeenth International Botanical Congress Vienna, July 2005.
Regnum Veg. 146. Ruggell, Liechtenstein: A.R.G. Gantner Verlag.
Patouillard, N. (1922) Quelques espéces nouvelles de Champignons. Bull. Trimestriel
Soc. Mycol. France 38: 83-87.
Patouillard, N. & Baker, C.F. (1918) Some Singapore Boletinae. J. Straits Branch Roy.
Asiat. Soc. 78: 67-72.
Pegler, D.N. (1997) The Larger Fungi of Borneo. Kota Kinabalu, Sabah: Natural
History Publications.
Quanten, E. (1997) The polypores (Polyporaceae s.|.) of Papua New Guinea. Opera
Bot. Belg. 11: 352.
Ryvarden, L. (1983) Type studies in the Polyporaceae 14. Species Described by N.
Patouillard, Either Alone or with other Mycologists. Occas. Pap. Farlow Herb.
Cryptog. Bot. 18: 39.
Ryvarden, L. (1989) Type studies in the Polyporaceae-21. Species described by C.G.
Lloyd in Cyclomyces, Daedalea, Favolus, Fomes, and Hexagonia. Mycotaxon
35: 229-236.
Ryvarden, L. (1990) Type studies in the Polyporaceae-22. Species described by C.G.
Lloyd in Polyporus. Mycotaxon 38: 83-102.
Ryvarden, L. (1992) Type studies in the Polyporaceae-23. Species described by C.G.
Lloyd in Lenzites, Polystictus, Poria, and Trametes. Mycotaxon 44: 127-136.
Singer, R. (1981) Notes on bolete taxonomy—III. Persoonia: 269-302.
188 Gard. Bull. Singapore 63(1 & 2) 2011
Tan, T.K. (1990) A Guide to Tropical Fungi. Singapore: Singapore Science Centre.
Watling, R. (2001a) (“2000”) Bresadola, Cesati and Patouillardd’s old world Boletes.
In: Papetti, C. & Consiglio, G. (eds) Micologia Duemile, pp. 577-599. Italy,
Trento: Associazione Micologica Bresadola.
Watling, R. (2001b) Edred John Henry Corner (1906-1996): a pioneer in tropical
mycology. Mycol. Res. 105: 1533-1536.
Watling, R. (2007) A mine of information and treasury of specimens: Corner’s
mycological legacy. In: Jones, E.B.G., Hyde, K.D. & Vikineswary, S. (eds)
Malaysian Fungal Diversity, pp. 25-39. University of Malaya & Ministry of
Natural Resources and Environment Malaysia.
Watling, R. & Ginns, J. (1998) E.J.H. Corner, 1906-1996. Mycologia: 732—737.
Zainuddin, N., Lee, S.S., Chan, H.T., Thi, B.K. (2010) A Guidebook to the Macrofungi
of Tasik Bera. Malaysia, Selangor Darul Ehsan: Forest Research Institute
Malaysia.
Gardens’ Bulletin Singapore 63(1 & 2): 189-195. 2011 189
e-Flora Malesiana: state of the art and perspectives
M.C. Roos'*, W.G. Berendsohn’*, S. Dessein*, T. Hamann’,
N. Hoffmann*, P. Hovenkamp'’, T. Janssen*, D. Kirkup’,
R. de Kok°, S.E.C. Sierra’, E. Smets!”’,
C. Webb® and P.C. van Welzen!
"Netherlands Centre for Biodiversity — Naturalis,
Section National Herbarium of The Netherlands, Leiden University,
PO Box 9514, 2300 RA Leiden, The Netherlands
*roos@nhn.leidenuniv.nl (corresponding author)
*Freie Universittat Berlin - Botanic Garden and Botanical Museum Berlin-Dahlem
K6nigin-Luise-Str. 6-8, DE-14195 Berlin, Germany
*The National Botanic Garden of Belgium
National Botanic Garden of Belgium, Domein van Bouchout, BE-1860 Meise, Belgium
"Royal Botanic Gardens, Kew
Richmond, Surrey TW9 3AE, U.K.
*The Arnold Arboretum, c/o Harvard University Herbaria
125 Arborway, Boston, MA, 02130-3500, United States of America
"Laboratory of Plant Systematics, K.U.Leuven
Kasteelpark Arenberg 31 box 2437, BE-3001 Leuven, Belgium
ABSTRACT. An overview is presented of available e-taxonomic products and ongoing
projects contributing to Flora Malesiana. This is presented in the context of a strong plea to
strengthen the implementation of state-of-the-art e-taxonomy tools to speed up the generation
and publication of Flora Malesiana information.
Keywords. EDIT, effective collaboration, European e-Floras Initiative, Flora Malesiana
Introduction
The increasing array of electronic taxonomy tools available for the elaboration and
dissemination of floristic information has brought many advantages. It enables a
shift from the traditional Flora concept as a static, printed account to a dynamic and
interactive format, allowing for rapid updating and multiple uses of information. It
provides taxonomists with means for instant interactive and remote co-operation,
including continuous processes for evaluating preliminary results by peers and
updating existing information. It also allows taxonomists to forge better links with
their user-communities by making the products of their research more tailor made and
accessible via internet (e.g., identification lists, specimen databases to monographs,
biodiversity and Non-Timber Forest Products assessments, and analyses of spatial
patterns of biodiversity).
190 Gard. Bull. Singapore 63(1 & 2) 2011
In order to strengthen and speed up Flora Malesiana activities, Roos &
Hovenkamp (2009) suggested that the Flora Malesiana community needed to adopt
a more pragmatic and flexible attitude—flexible in terms of formats and publication
strategy, and standards of robustness and confidence in the results.
The Flora Malesiana project has recently gained new momentum by starting a
new website (www.floramalesiana.org), by adopting e-taxonomy tools and by joining
a broadly supported initiative that wants to promote the adoption of richly interactive
and truly collaborative systems for the production and presentation of floras: The
European e-Floras (enhanced-Floras) initiative.
The European e-Floras initiative
Based on the results of an EDIT e-Flora Platform Workshop organised in Leiden
(January 2010) the Board of the Foundation Flora Malesiana has decided during
the most recent board meeting in Singapore to work towards an e-Flora Malesiana
platform, to be developed in co-operation with the European e-Floras initiative.
Participating institutions of the EDIT workshop in January also discussed the
possibility of starting collaboration activities between Flora projects and the EDIT
Platform for Cybertaxonomy by means of taxonomic exemplar groups (http://www.e-
taxonomy.eu/node/748).
During a second workshop in Berlin (March 2010) three exemplary groups
were set up: Flore d’ Afrique centrale (National Botanic Garden of Belgium; Dessein et
al., 2010; http://www.br.fgov.be/RESEARCH/DATABASES/FOCA/index.php), Flora
Malesiana (Netherlands Centre for Biodiversity; http://160.45.63.201/dataportal/
preview/flora-malesiana) and Kew’s African Floras (Royal Botanic Gardens Kew).
During a third workshop organised in Brussels (September 2010) participants
agreed to establish an “European e-Floras Initiative” to enhance communication
between (e-)Flora authors, e-taxonomy initiatives and users of content and thus to
increase research efficiency, reduce redundant efforts and speed up the preparation
of up-to-date, high-quality content on plant biodiversity in formats permitting rapid
updates and multiple uses (http://www.e-taxonomy.eu/node/859).
At present, the European e-Floras initiative is supported by 23 leading
institutes from Europe, Asia, Africa and Australia involved in Flora production.
Why an e-Flora Malesiana?
The Flora Malesiana progress has lost some of its momentum since the beginning
of the century. The overall progress in terms of the number of species covered is
too slow (about 75% of all species in Malesiana still need to be treated) to make a
meaningful contribution towards sustainability and conservation. Reasons for this are
(1) the decreasing number of taxon experts that are available and devoted to write
taxonomic treatments; (11) the scattered generation of information and its particular
e-Flora Malesiana 191
usage (i.e., just for one specific project and format) in differently oriented projects: (111)
duplication of efforts; and (iv) different formats. In order to use the available expertise
and resources effectively, a shift in the conventional work processes is needed.
Since 1950, Flora Malesiana has published its products as family treatments
(in total 204 up to now). At present, several volumes are no longer available as hard
copy (Series I Volume 1, 4-6, 9 Part 2—3 and Series II Volume | Part 1-3). Only
the most recent volumes are available in electronic formats that can be more easily
imported to a database than printed texts:
Text: Series I Volume 13;
Colour images: Volume 14:
Digital files: Volume 15 onwards, for Series II only Volume 3 completely available
as digital files;
CD-ROMs: Leguminosae: Caesalpinioideae and Mimosoideae, Orchids of New
Guinea Vol. I-VI, Orchids of the Philippines Vol. I.
An e-Flora will complement printed formats with electronic editions,
offer unlimited access and instant updates, and increase the cost-effectiveness
substantially by streamlining the production work flow. It will allow structured data
entry, interactivity, multimedia, and enhanced accessibility. Moreover, it will also
allow collaboration between several e-Flora projects (e.g., consortium members of
the European e-Floras Initiative), users of taxonomic data, and other databases. This
increases the possibilities for innovative scientific co-operation, also between other
research fields (ecology, ethnobotany, etc.) and attracts worldwide contributors using
services offered by the e-Flora.
How can FM work be strengthened?
Flora Malesiana (FM) work can be strengthened by increasing its global accessibility,
facilitating efficient remote collaboration, making use of databases to safeguard data,
changing the work flow in data preparation and presentation, and creating institutional
commitment.
1. Increasing FM global accessibility
FM Website. The visibility of FM on the World Wide Web will be increased by
bringing available information and data online via its website (www.floramalesiana.
org). The current website will gradually be brought up to date both in design and
contents, and will be established as the information exchange portal for FM.
It will feature information on the latest FM meetings (including web feed
formats used to publish frequently updated works), and the FM Editorial Committee,
collaborators, and other contributors. Furthermore, it will provide access to the FM
e-Flora, and also include links to other relevant e-Floras, e-initiatives and/or databases,
e.g., Biodiversity Heritage Library (BHL), Cyclopedia of plant collectors, European
Distributed Institute of Taxonomy (EDIT), Creating a Taxonomic e-Science (CATE).
192 Gard. Bull. Singapore 63(1 & 2) 2011
Journal Storage (JSTOR), specimen databases (e.g., plants.jstor.org), etc. FM printed
volumes and CD-ROMs will be available through the website.
Interactive Key to the Malesian Seed Plants. The Interactive Key to the Malesian
Seed Plants is a user-friendly electronic identification aid (DELTA key) for the plants
in SE Asia. The first version was published on the web (www.kew.org/herbarium/keys/
fm) and in CD-ROM format in 2004. It covers all seed plant families of the Malesian
region, and is supported by c. 1000 pictures and family portraits. In the second version,
the plan is to enlarge the key by including the plant families from Thailand and Indo-
China, by including all genera done in the Flora Malesiana as possible answers, and by
linking the key to the electronic version of the Flora.
List of remaining families. It is imperative that contributors and users can have access
to available FM information and data. An inventory of the families still not allotted to
(teams of) specialists and work that has been done and still needs to be done for Flora
Malesiana (in terms of genera treatments) is being made available through the FM
website. It will provide an overview of the different contributors, the working teams
and the taxa they work on (with an indication how well verified or preliminary they
are). The inventory will be periodically updated by the different contributors via the
website contact (roos(@nhn.leidenuniv.nl).
Checklist. For missing treatments a checklist of data will be made available by
extracting the necessary data from the Kew World Checklist. This will form a backbone
by incorporating missing parts of the FM area, provide an overview of missing taxa,
connect information, and encourage people to start working on those taxa. It will also
give an overview of the resources on the FM area that are still needed.
FM Bulletin. The FM Bulletin (http://www.nationaalherbarium.nl/fmbull/biblio.htm)
contains the bibliography on Malesian botany, fieldwork and other field trips, etc. It
was started in 1947 by C.G.G.J. van Steenis and through its indices is a goldmine of
information on SE Asian floristics and taxonomy. At present the Bullletin has been
reduced to the bibliographies on mosses and vascular plants and is only available
online. These contain only the bibliographies published from volume 11 (6) 1995 to
2009, but not those of 13(4) through 14(3) as yet.
Cyclopedia of plant collectors. The Cyclopedia of plant collectors (http://www.
nationaalherbarium.nl/fmcollectors) contains data on collectors in the Southeast Asian
Archipelago, also known as Malesia (comprising Brunei, Indonesia, Malaysia, Papua
New Guinea, Phillippines, Singapore, and Timor Leste). The data were collected by
Mrs. Van Steenis-Kruseman and they are digitised from FM ser. 1, part 1, 5, and 8.
Pictures of collectors and samples of handwriting and signatures are often included. The
data on collecting trips are more or less complete up to 1974. Later data on collectors
and their trips can be obtained from the FM Bulletin. The website is especially useful
when one digitises hand written specimen labels from the earlier collectors.
e-Flora Malesiana 193
2. Facilitating efficient remote collaboration
Collaboration can have different forms. A measure taken in 1989 to speed up and
revitalise the FM project was to establish family teams. At present, the most common
way is to divide work (either taxonomically or geographically), finish each part
independently and perhaps build a common (set of linked) data base(s) (see, e.g.. an
early example, the Euphorbiaceae: www-nationaalherbarium.nl/euphorbs) and compile
an overall treatment at the end. However, to make full use of the e-possibilities, it is
imperative to exploit the potential of internet communication tools that allow instant
communication and sharing of data, such as online fora, Internet Relay Chat (IRC),
Cloud computing services (e.g.. Google Docs), and social networking tools, during the
whole process of generating data and publishing information.
By making use of community e-tools for taxonomy (1.e., Scratchpads, EDIT
Platform for Cybertaxonomy) the efficiency of the taxonomic work processes (data
preparation and publication) could be increased. Scratchpads (http://scratchpads.
eu) is a social networking application that enables communities of researchers to
manage, share and publish taxonomic data online. It helps to increase visibility of
ongoing projects, and creates interaction and synergy between remote working groups.
The Taxonomic Editor - EDITor is part of the EDIT Platform for Cybertaxonomy
(Berendsohn 2010). It is a desktop application that can be used to edit data stored in a
standards-based community store (CDM-Store). It edits data in either a remote source,
or a local data source embedded in the application. The EDITor allows collaborators to
manage their data. Other CDM-based applications of the Platform allow the production
of printed versions in flexible format or direct and up-date output to a website (CDM
Data Portal). The latest version can be found at http://wp5.e-taxonomy.eu/cdm-setups
taxonomic _ editor).
3. Making use of e-databases to safeguard FM data
The defragmentation of taxonomic data and the use of common standards will increase
the sustainability of FM work. The Common Data Model (CDM) is a data format
for every conceivable type of data produced by taxonomists in the course of their
work (http://wp5.e-taxonomy.eu). It enables professionalised taxonomic software
development and allows for common standards that create sustainability. Furthermore,
it can be used to exchange information with other taxonomic databases such as
BRAHMS (hitp://dps.plants.ox.ac-uk/bol), and also non-taxonomic databases like
TRY (http://www.try-db.org). The CDM-based EDIT Platform for Cybertaxonomy
will facilitate the generation of species lists without generic contradictions (with the
advantage that other kinds of projects could also use it), but it also allows to store
alternative taxonomic classifications, e.g. entire monographs and flora treatments. This
will greatly assist the consolidation of the taxonomic research results in problem areas.
We envisage that the FM e-Flora will become available in several different
formats: (i) for online use on personal computers and mobile phones: (ii) as stand-
alone versions on CD/DVD-ROM or as down-loadable applications for smart phones,
tablets, and other mobile devices; and (111) as printed volumes using a print-on-demand
system that will always use the latest version of the e-Flora.
194 Gard. Bull. Singapore 63(1 & 2) 2011
4. Changing the work flow in data preparation and presentation
The use of e-tools facilitates the production of new content for printed and electronic
publications (instead of using printed sources to produce digitised content). However,
to port FM data, which is currently only available in print or in text format, to the
online tools, markup is a prerequisite, 1.e., the insertion of markers designating specific
content types in the text (e.g., a generic name, a distribution record, etc.). FM floristic
information is at present being digitised using XML (eXtensible Markup Language;
see http://www.w3.org/standards/xml). FM volumes that are already available in a
digital format can be marked up straightaway, whereas earlier volumes need to be
scanned first in high quality.
Mark-up is generally performed in Microsoft Word, using a combination
of automated procedures and manual corrections. Automation (e.g., Word-macros)
may be used to speed up the mark-up process of highly structured texts. Manual
corrections are required, e.g., when taxonomists have used various types of shorthand
notations to combine similar species names into one paragraph in printed floras, or
when typographical or text-recognition-errors interfere with the automated mark-up.
The resulting XML-files are imported into the EDIT CDM using a specific import
scheme. Figures are prepared for use with the marked-up files and the CDM, but
are located on a separate image server. The mark-up process and preparation of the
images takes roughly a month per volume to complete (25 MS pages/day). A preview
of the future FM e-Flora portal as a CDM Dataportal implementation can be found
at http://160.45.63.201/dataportal/preview/flora-malesiana. A finalised version of this
portal will be made available through the FM website in 2011.
5. Creating institutional commitment
Institutional commitment from Flora Malesiana institutions and support of the European
e-Floras Initiative activities is highly desirable and needed to speed up activities of the
Flora Malesiana project. The institutional commitment that is necessary to achieve
this is not limited to providing dedicated staff, but should also include the necessary
infrastructure, including connections to available high-throughput internet facilities:
TEIN3 and TEIN4 (http://www.tein3.net). Staff evaluation criteria should include
also contributions towards data base maintenance, conforming to the recent MoU on
Evaluation Criteria for Taxonomic Work as adopted by EDIT institutions.
References
Berendsohn, W.G. (2010) Devising the EDIT Platform for Cybertaxonomy. In: Nimis,
P.L. & Vignes-Lebbe, R. (eds) Jools for Identifying Biodiversity: Progress and
Problems, pp. 1-6. Italy, Trieste: Edizioni Universita di Trieste.
Dessein, S., Janssen, T., Groom, Q., Robbrecht, E., Roos, M. & Sierra, S. (2010)
E-Floras for Africa: state of the art and perspectives. In: Jeannoda, V.H.,
Razafimandimbision, S.G. & Block, P. de (eds) Abstracts. XIXth AETFAT
e-Flora Malesiana 195
Congress - Madagascar, 25—30 April, 2010, p. 133. Scripta Botanica Belgica 46,
Meise: National Botanic Garden of Belgium.
Roos, M. & Hovenkamp, P. (2009) Flora Malesiana in the coming decade. Blumea 54:
3-5.
http://wp5.e-taxonomy.eu/cdm-setups/taxonomic_ editor
http://dev.e-taxonomy.eu/trac/wiki/TaxonomicEditor
http://dps.plants.ox.ac.uk/bol
http://scratchpads.eu
http://wp5.e-taxonomy.eu
http://www.br.fgov.be/RESEARCH/DATABASES/FOCA/index.php
http://www.e-taxonomy.eu/node/748
http://www.e-taxonomy.eu/node/859
http://www. floramalesiana.org
http://www.kew.org/herbarium/keys/fm
http://www.nationaalherbarium.nl/fmbull/biblio.htm
http://www.nationaalherbarium.nl/fmcollectors
http://www.tein3.net
http://www.try-db.org
http://www.w3.org/standards/xml
http://160.45.63.201/dataportal/preview/flora-malesiana
Gardens’ Bulletin Singapore 63(1 & 2): 197-204. 2011 197
A summary of the total vascular plant flora of Singapore
K.Y. Chong’, Hugh T.W. Tan and Richard T. Corlett
Department of Biological Sciences,
National University of Singapore,
14 Science Drive 4, Singapore 117543
'kwek@nus.edu.sg
ABSTRACT. The last analysis of the vascular plant flora of Singapore was published more
than a decade ago. Since then, the conservation statuses of all native species have been assessed
and more exotic species have been recognised as naturalised. We present a holistic view of
the family compositions and life forms of the total flora, including many exotic species that
are found in cultivation only and not yet escaped or naturalised. Excluding extinct species,
exotic species now outnumber native species. Horticultural introductions have also strongly
influenced family compositions: legumes and palms are now the most species-rich families in
the total flora. Legumes are also a dominant family among all naturalised life forms. We briefly
discuss these implications for local conservation ecology.
Keywords. Cultivation, exotics, extinctions, natives, naturalisations, total flora, vascular plants
Introduction
Singapore’s vascular plant flora is relatively well-documented for a_ tropical
country. The first compilation was the Flora of Singapore by Ridley (1900) and his
supplementary notes that shortly followed (Ridley 1901). The checklist by Turner
et al. (1990) was the first, comprehensive, published update to Ridley (1900), and
incorporated an unpublished Flora of Syonan compiled by the staff of the Singapore
Botanic Gardens’ Herbarium during the Japanese Occupation from 1942 to 1945, as
well as Corlett’s (1988) list of naturalised plant species. Turner et al.’s checklist was
subsequently expanded into a fully-referenced list of vascular plant names (Turner
1993). Keng’s (1973-1987) annotated lists of seed plants of Singapore were developed
into the two volumes of the Concise Flora of Singapore (Keng 1990, Keng et al.
1998). These publications formed the backbone of subsequent updates and additions,
and included native species, naturalised exotics, known escapes from cultivation and
the most commonly cultivated species.
Two rounds of conservation assessments were conducted for the vascular
plants and published as the first and second editions of the Singapore Red Data
Book respectively (Ng & Wee 1994, Davison et al. 2008). With the second round
of assessments, every native vascular plant species, with the exception of those few
overlooked by past publications, has been assigned a conservation status. In addition,
two editions of the /00/ Garden Plants in Singapore (Boo et al. 2006) listed many
more vascular plant species that are cultivated in Singapore. Although not a strictly
botanical work, this book advanced our knowledge of the large numbers of plants
198 Gard. Bull. Singapore 63(1 & 2) 2011
that have been introduced into Singapore via horticulture. The chronology of these
publications is summarised in a timeline in Fig. 1.
Although past work has already described the taxonomic composition and life-
forms of the Singapore vascular plant flora, these have focused on various components
such as the naturalised exotics (Corlett 1988) or the native species (Turner 1994,
Turner et al. 1994). None of the studies, however, have taken a holistic approach to the
floristic composition from a total flora perspective, i.e., one that takes into account all
species found in the country, including exotics found in cultivation only.
We recently compiled a checklist of the total vascular plant flora of Singapore
(Chong et al. 2009). In this paper, we summarise our findings on the taxonomic
composition and life forms from a total flora perspective.
Keng's Concise Flora
(Gymnosperms and Dicots)
Keng's Annotated list of seed — Keng et al.'s Concise Flora
plants of Singapore (Parts I = | (Monocots)
xI
) 2008 Singapore Red Data
| Book
: ; = i Red Da
+ Ridley's Flora of Singapore Turner et all's Checklistof | 7 = i Singepbre el Data
Singapore Vascular Plants |
0S a ES SS OER EES SN ER,
1890 1910 1930 1950 1970 1990, 2010
Corlett’s Naturalized Flora of |
Unpublished Flora of Syeran Singapore 1001 6arden Plants 2nd ed.
L Ridley's Supplementary — Boo et al.'s 1001 Garden
Notes Plants ist ed.
Turner's "Names used for
Singapore Plants”
Fig. 1. Timeline of the major publications used in the compilation of a total flora of Singapore.
Materials and methods
Key reference sources and the methodology employed in the compilation of the flora
used in this analysis is given in Chong et al. (2009). In summary, all plant names given
in the local floristic literature reviewed above were compiled into a spreadsheet and
checked for synonyms. Species that were inferred to have been present in Singapore
before human-mediated introductions are considered “native”. Native species that have
not been collected or seen by botanists in the last 30 years are considered “extinct”,
as defined in Davison et al. (2008). Species whose presence is a result of human
involvement are considered “exotic”. Exotic species that have established outside of
cultivation but rely on repeated introductions of propagules for persistence in the wild
are termed “casual”. Exotic species that maintain self-replacing populations in the
wild independent of new human introductions are termed “naturalised”. Remaining
exotic species that can only be found in cultivation are termed “cultivated-only”. For
those weedy species that can only be found in human-modified habitats in Singapore,
Total vascular plant flora of Smgapore 199
for which we lack knowledge of their original biogeographic range, we apply the term
“cryptogenic” following the terminology by Carlton (1996).
Since the publication of our checklist, we have been continually updating and
correcting our database upon encountering new literature and information provided
by users of the checklist. Family circumscriptions of seed plants follow that of the
Angiosperm Phylogeny Group (APGIII 2009, Stevens 2001 onwards), while non-
seed plants follow Smith et al. (2006). The results presented here represent the most
updated version of this database.
Results and discussion
In our checklist, we reported relative proportions of native, exotic and cryptogenic
species. Here we report almost identical figures: there are 2141 native, 1822 exotic, and
210 cryptogenic species, constituting 51%, 44% and 5% of the total flora, respectively.
The number of exotic species is an underestimate, as we are likely to have left out
many other less commonly cultivated species in private gardens and nurseries.
Extinctions and introductions have had a major impact on the representation
of families in the total flora (Table 1). The Orchidaceae was the largest native family,
but massive extinctions, small numbers of species in cultivation and the absence of
spontaneous exotics has reduced its rank to eighteenth in the total extant flora. The
Rubiaceae, as the second largest native family, sustained far fewer extinctions and
remains dominant as the largest extant native family and the third largest in the total
extant flora. The Fabaceae is now the largest family overall as a result of high numbers
of both spontaneous and cultivated exotics, followed by the Arecaceae, which has the
largest number of cultivated species.
Trees are the most common life form for both natives and exotics (Fig. 2).
Among the extant woody natives, the Rubiaceae are the largest family, with 17 extant
tree species and 41 extant shrub species. These dominate the forest understorey.
Among the native trees, the Meliaceae has sustained the largest number of extinctions
(14 species; 35%), larger in proportion than other major families of trees. This may
be attributed to the declines and extinctions of large frugivores that disperse the large-
fruited Meliaceae, resulting in lack of recruitment and consequently decreased chances
of population persistence (Corlett 2007).
Epiphytes are the only life form where most of the species have become
extinct (Fig. 2a). Turner et al. (1994) noted that most of the epiphyte extinctions
are accounted for by the epiphytic habit of the orchids, and that ferns, the second
largest group of epiphytes, sustained proportionally far less extinctions. Here we
repeat this observation but also show that the Apocynaceae and the Loranthaceae, the
third and fourth largest families of epiphytes respectively, have suffered more than
50% extinctions (Table 2). The Apocynaceae epiphytes consist of only two genera:
Hoya and Dischidia. The Loranthaceae consist solely of hemi-parasitic mistletoes.
The susceptibility of epiphytes to extinction therefore appears to be consistent at least
among flowering plant families.
Gard. Bull. Singapore 63(1 & 2) 2011
200
0 ce = t i OS JRIOEILUIOISE]II\|
Sb 97 fii Z 0 61 0 GG iy SI Ip avadVI[IIN]
9 cE 97 s € © OI a SY © vl ¢ I 9 DvIIVIIISV
$9 06 E % SE 0 0 Se 0 O€ € €€ sevaoedivs01diq
09 97 67 5 S 3 6€ Lie rl 6 €7 avaorIwe]
€L as O€ r 9 3 § a (IP p 87 Z O€ aevaovIV
89 s 6S 8 I 0 6 0 7 6S 8 « LO aeaovuouUYy
$9 9 CS ZI I € 91 0 Pa OV ae tay eng aeoovyjue||Ayd
@ oe GC. 9S 9 v 9 9 + 99 a 9 II S 91 aeaovyURoV
6L 3 OS 97 s € r Ol 6€ b OE WAC 8b aevaoeriqioydng
or acs oy cS €T eee I I¢ I « OS 01 . 09 avadVIOW|
9 C6 ae ac hz 9 v € 9 S auell I¢ 9 LE avaoVATLI|
9 © 2 HS Ss 0 G ib 1 tb 6. BW Ol o IS avaoesodA)
9 © o Sb 9 Sb © € . OS 0 3 cP « ST « L9 avaoeudsody
s bil Cyan ee 169 9 € I omen 0 e Wy? & Ul @ SS avaovyA
y CEI < 86 iE el! 2 €C a JES 7 CE 1 €Y 9 6b aeaovod
¢ SEI 1 VOL o PE 0 s 6 6 &Y r Ol 7 5 7 6£ z vl aeaovigny
z II@ Ib 1 891 G € 2 ELI 0 S6 = = Ol 6 Ss aeIWVIATY
ES 6 OF 9 I 0 L I 9 SY 1 18 Siea9@e avaoeplyO
TSO = ae VOL. 2 rel - Ol 1 OF 1 061 ; 8 © OS oa II s 19 aeaouqef
JUSSI PITM AjUo jense) posryeinyen SONOXY stuasojdA1y = ue} Xq yOUNX| SOARN ATU]
IIV HV = -poweanyng IIV IIV
‘sardads Jo JOquINU ay} Jo 1YBLI ot) 0}
Jdiossodns ul USAIS d1e AIOSI}LO YORD UIYIIAM SOIL] JSaBIL] CQ JO syueyY ‘osodeSurg jo esoy yuLyd JepNoseA oy} Ul SoITUIe] CZ ISASIeT “YT 9quy
Total vascular plant
shrub- 384 (21%)
herb - 488 (27%)
climber - 146
(3%)
strangier - 5 _J
epiphyte - 2 100 -
Table 2. Largest five families of epiphytes in Singapore.
201
Family All Extinct Extant
eee lo+ . 142 22
Polypodiaceae 19 3 16
Apocynaceae 17 9 8
Loranthaceae i+ 8 6
Hymenophyllaceae 13 - 9
202 Gard. Bull. Singapore 63(1 & 2) 2011
Of the exotic growth habits, the herbs have the largest proportion of naturalised
species although more tree species were introduced for cultivation (Fig. 2b). Among the
naturalised exotic species, the five largest families of climbers are Fabaceae (7 species),
Convolvulaceae (4 species), Passifloraceae (3 species), Acanthaceae (2 species) and
Cucurbitaceae (2 species). The five largest families for naturalised, non-climbing woody
species are Fabaceae (29 species), Euphorbiaceae (5 species), Verbenaceae (5 species),
Urticaceae (4 species) and Solanaceae (3 species). Finally, the 10 largest families of
naturalised, non-climbing herbs are the Poaceae (23 species), Asteraceae (15 species),
Fabaceae (10 species), Rubiaceae (8 species), Euphorbiaceae (5 species), Acanthaceae
(4 species), Araceae (4 species), Amaranthaceae (4 species) and Cleomaceae (4
species). Many of these naturalised herbs are garden weeds or open wasteland ruderals
and do not pose a threat to native forests. The woody naturalised species may be more
of a cause for concern: some of the legumes form exotic-dominated woodlands where
recruitment and regeneration of native species are slow, while other species such as
Cecropia pachystachya Trécul (Urticaceae), Ptyvchosperma macarthurii (H.Wendl. ex
H.J.Veitch.) H.Wendl. ex Hook.f. (Arecaceae) and Syngonium podophyllum Schott
(Araceae) have been found in native forests edges and gaps (Lok et al. 2010).
Our compilation of a total vascular flora is the first for the tropics, and includes
both the remnants of the original tropical rainforest cover as well as the elements of
introduced flora from urbanisation. After taking extinctions into account, exotic species
richness now exceeds native species richness. Considering that many of the native
species are endangered and have smali population sizes, while some exotic species
have been planted in high densities throughout Singapore, the relative abundance
of exotics to natives is also likely to reflect this. Although a large number of exotic
species have been introduced, only a fraction has become naturalised, but more studies
are needed to evaluate the impacts of these naturalisations on the local flora and fauna.
Non-naturalised exotics in the urban environment also interact with native and exotic
animals, providing nesting sites and food. Given the influence that cultivation can
have on the floristic composition, replacing exotic horticultural species with native
plants may be a strategy for conservation. Functional diversity lost from extinctions
can be regained by reintroductions of extinct species, and populations of rare species
can be augmented with horticultural plantings, while reducing the risks of invasive
species introductions.
ACKNOWLEDGEMENTS. Weare grateful to the following people who have provided updates
and given feedback on mistakes found in our checklist: Yeo Chow Khoon, Alvin Lok, Ang Wee
Foong, Alex Yee, Ng Pei Xin, Teo Siyang, Tan Siu Yueh, Beatrice Ng, Suen Si Min, Nghiem
Thi Phuong Le, Ng Ting Hui, Lee Yen-Ling, Robert Teo and Patricia Yap. We also thank Dr.
Benito Tan and the organisers of the 8th Flora Malesiana Symposium for their invitation and
financial assistance that enabled C.K.Y. to give an oral presentation at the Special Session on
Singapore Plant Studies. This paper is dedicated to the late Dr. Hsuan Keng, who passed away
in 2009. Without his efforts on the flora of Singapore, this work would not have been possible.
Total vascular plant flora of Singapore 203
References
APG III (2009) An update of the Angiosperm Phylogeny Group classification for the
orders and families of flowering plants: APG III. Bor. J. Linn. Soc. 161: 105-121.
Boo, C.M., Kartini Omar-Hor & Ou-Yang, C.L. (2006) 1001 Garden Plants of
Singapore. 2nd ed. Singapore: National Parks Board.
Carlton, J.T. (1996) Biological invasions and cryptogenic species. Ecology 77:1653—
1655.
Chong, K.Y., Tan, H.T.W. & Corlett, R.T. (2009) A Checklist of the Total Vascular Plant
Flora of Singapore — Native, Naturalized and Cultivated Species. Singapore:
Raffles Museum of Biodiversity Research, National University of Singapore.
(http://rmbr.nus.edu.sg/raffles_museum_pub/flora_of_singapore_tc.pdf),
uploaded 12 Nov 2009.
Corlett, R.T. (1998) The naturalized flora of Singapore. J. Biogeogr. 15: 657-663.
Corlett, R.T. (2007) The impact of hunting on the mammalian fauna of tropical Asian
forests. Biotropica 39: 292-303.
Davison, G.W.H., Ng, P.K.L. & Ho, H.C. (2008) The Singapore Red Data Book
— Threatened Plants and Animals of Singapore. Singapore: Nature Society
(Singapore).
Keng, H. (1973) Annotated list of seed plants of Singapore (I). Gard. Bull. Singapore
26: 233-237.
Keng, H. (1974a) Annotated list of seed plants of Singapore (II). Gard. Bull. Singapore
27: 67-83.
Keng, H. (1974b) Annotated list of seed plants of Singapore (III). Gard. Bull. Singapore
27: 247-266.
Keng, H. (1976) Annotated list of seed plants of Singapore (IV). Gard. Bull. Singapore
28: 237-258.
Keng, H. (1978) Annotated list of seed plants of Singapore (V). Gard. Bull. Singapore
31: 84-113.
Keng, H. (1980) Annotated list of seed plants of Singapore (VI). Gard. Bull. Singapore
33: 329-367.
Keng, H. (1982) Annotated list of seed plants of Singapore (VII). Gard. Bull. Singapore
35: 83-103.
Keng, H. (1983) Annotated list of seed plants of Singapore (VIII). Gard. Bull.
Singapore 36: 103-124.
Keng, H. (1985) Annotated list of seed plants of Singapore (IX). Gard. Bull. Singapore
38: 149-174.
Keng, H. (1986) Annotated list of seed plants of Singapore (X). Gard. Bull. Singapore
39: 67-94.
Keng, H. (1987) Annotated list of seed plants of Singapore (XI). Gard. Bull. Singapore
40: 113-132.
Keng, H. (1990) The Concise Flora of Singapore: Gymnosperms and Dicotyledons.
Singapore: Singapore University Press.
204 Gard. Bull. Singapore 63(1 & 2) 2011
Keng, H., Chin, S.C. & Tan, H.T.W. (1998) The Concise Flora of Singapore Volume
II: Monocotyledons. Singapore: Singapore University Press.
Lok, A.F.S.L., Chong, K.Y., Tan, K.-X. & H.T.W. Tan (2010) A checklist of the
spontaneous exotic vascular plant flora of Singapore. COSMOS 6: 57-83.
Ng, P.K.L. & Wee, Y.C. (1994) The Singapore Red Data Book — Threatened Plants and
Animals of Singapore. Singapore: Nature Society (Singapore).
Ridley, H.N. (1900) The flora of Singapore. J. Straits Branch Roy. Asiat. Soc. 33:
27-196.
Ridley, H.N. (1901) Supplementary notes on the flora of Singapore. J. Straits Branch
Roy. Asiat. Soc. 35: 84-90.
Smith, A.R., Pryer, K.M., Schuettpelz, E., Korall, P., Schneider, H. & Wolf, P.G. (2006)
A classification for extant ferns. Jaxon 55: 705—731.
Turner, I.M. (1993) The names used for Singapore plants since 1900. Gard. Bull.
- Singapore 45: 1-287.
Turner, I.M. (1994) The taxonomy and ecology of the vascular plant flora of Singapore:
a Statistical analysis. Bot. J. Linn. Soc. 114: 215-227.
Turner, I.M., Chua, K.S. & Tan, H.T.W. (1990) A checklist of the native and naturalized
vascular plants of the Republic of Singapore. J. Singapore Natl. Acad. Sci. 18 &
19: 58-88.
Turner, I.M., Tan, H.T.W., Wee, Y.C., Ali bin Ibrahim, Chew, P.T. & Corlett, R.T. (1994)
A study of plant species extinction in Singapore: lessons for the conservation of
tropical biodiversity. Conservation Biol. 8: 705—712.
Gardens’ Bulletin Singapore 63(1 & 2): 205-212. 2011 205
The vegetation of Singapore
—an updated map
A.T.K. Yee'”, Richard T. Corlett!, S.C. Liew’ and Hugh T.W. Tan!
'Department of Biological Sciences, National University of Singapore
14 Science Drive 4, Singapore 117543
*haloriyee@gmail.com (corresponding author)
‘Centre for Remote Imaging, Sensing and Processing,
National University of Singapore,
10 Lower Kent Ridge Road, Singapore 119076
ABSTRACT. The primeval vegetation of Singapore was largely lowland dipterocarp forest,
with mangrove forest lining much of the coast and freshwater swamp forest found further
inland adjacent to the streams and rivers. After colonization by the British in 1819, almost all the
primeval vegetation was cleared for agriculture and other land uses. The most comprehensive
vegetation map of Singapore was made in the 1970s and has not been updated since. Here
we present an updated vegetation map of Singapore using information from satellite images,
published works, and extensive ground-truthing. Vegetation covers 56% of Singapore’s total
land area: 27% is actively managed (parks, gardens, lawns, etc.) and 29% is spontaneous
vegetation. Primary lowland dipterocarp forest and freshwater swamp forest cover only 0.28%
and is confined to the Bukit Timah and Central Catchment Nature Reserves. The majority of
the non-managed vegetation is secondary forest of various kinds, dominated by native or alien
trees. The managed vegetation and alien-dominated secondary vegetation are understudied and
deserve more research attention. The vegetation of Singapore should be re-mapped at regular
intervals in order to better understand the changes.
Keywords. Forest, map, Singapore, vegetation
Introduction
Most of primeval Singapore was covered by forest, while open habitats were largely
confined to beaches and coastal cliffs (Corlett 1991, 1992a). According to Corlett
(1991), 13% of the primeval vegetation was mangrove forest, 5% was freshwater
swamp forest, and the rest was mainly lowland dipterocarp forest (Fig. 1). Although
the island had been continuously inhabited for several centuries, rapid deforestation
occurred only after colonization by the British in 1819. By 1900, 90% of the primeval
forest had been cleared, mainly for agriculture and by 1935, rubber plantations
occupied 40% of Singapore’s area (Corlett 1991).
After the independence of Singapore in 1959, land use change was driven by
urbanization and modernization. Active large-scale land reclamation has been carried
out since the 1960s (Wong 1985). By 1973, less than 30% of Singapore’s land area was
covered by spontaneous vegetation while the rest was plantations, and suburban and
206 Gard. Bull. Singapore 63(1 & 2) 2011
Legend
Primeval Vegetation
IY) Freshwater Swamp Forest
| ~ oN sae GR Mangrove Forest
| Q
ge% HE Lowland Dipterocarp Forest
0 25 5 10 Kilometers
—————E—E——Ee
| Source: Corlett RT, 1991. Vegetation. In LS Chia, A Rahman & DBH Tay (eds.)
The Biophysical Environment of Singapore. Singapore University Press, Singapore
Fig. 1. Vegetation map of primeval Singapore.
urban areas (Hill 1977). By 1990, more than half of Singapore was urbanized, most
of the plantations were abandoned and more than 99% of the original forest had been
cleared (Corlett 1991, 1992a).
The vegetation of Singapore was partly mapped in the nineteenth century. For
example, Coleman’s and Thomson’s maps showed vegetation around the Singapore
Town (see Corlett 1992b). Vegetation can also be inferred from the topographic maps
of Singapore. R.D. Hill (1977) produced the first extensive vegetation map based
on aerial photographs from 1959 and 1969, and ground truthing from 1972 to 1973.
Subsequently, only maps of specific areas or vegetation types were produced, such
as the mangrove vegetation map of Hilton & Manning (1995), and the map of the
vegetation of the Bukit Timah and Central Catchment Nature Reserves (see Corlett
1997). Therefore, we aim to produce a recent vegetation map for the entire Singapore.
Materials and methods
Ground truthing was carried out in Singapore from June 2009 to March 2010. The
vegetation types were noted down in Singapore Street Directory, 2009 edition (Mighty
Minds® 2009) if the area was bounded by roads, or in a GPS receiver (Garmin
Vegetation map of Singapore 207
GPSMAP® 60CSx) with at most +7 m error if it was not bounded by roads. From the
ground truthing data, 405 regions of interest (ROIs) were created.
Two SPOT 5 satellite images from 08 Mar. 2006 and 31 Aug. 2007, and
the 405 ROIs were used to create a vegetation map of Singapore by a supervised
classification technique, using the maximum likelihood method, which is available
in the software package ENVI version 4.4 (ITT Visual Information Solution 2007). A
total of 85 out of the 405 ROIs were used as the training data. The remaining 320 ROIs
were used to assess the accuracy of the supervised classified vegetation map via the
‘confusion matrix’ function in ENVI version 4.4. Both satellite images underwent the
same procedures. We adopted the standard of Thomson et al. (1999), which suggested
at least 85% for overall accuracy, and at least 70% accuracy for individual classes.
When desirable accuracy was obtained (Tables | & 2), the supervised classified
vegetation map based on the 31 Aug. 2007 satellite image was overlain with that based
on the 08 Mar. 2006 satellite image (Fig. 2). Subsequently, persisting clouded areas
were filled in using information from Google Earth 5.1.3533.1731 (Google Inc. 2009).
Table 1. Confusion matrix computed for the automated vegetation map of year 2006. Values are
reported in percentages. The overall accuracy is 86.30%, and the Kappa coefficient is 0.7907.
Ground truthed
: cs eae a ueesiand Forest el
Water 927713 3.32 0.38 0.59 3.42
Non-vegetated 5.10 80.04 9.14 3.4] 4.08
Grassland 0.12 12.49 75.28 6.81 0.02
Forest 1.82 3.63 [S15 88.33 LPS
Mangrove Forest 0.23 0.52 0.05 0.87 81.33
Table 2. Confusion matrix computed for the automated vegetation map of 2007. Values are
reported in percentages. The overall accuracy is 90.61%, and the Kappa coefficient is 0.8621.
Ground truthed
clasineation 8" vegerateg Grassland Forest rege
Water 97.43 0.08 0.08 0.00 1.60
Non-vegetated 2.40 88.09 U2 0.59 3.08
Grassland 0.00 9.81 74.94 5.84 0.07
Forest 0.00 1.68 16.53 oF 13 8.84
Mangrove Forest 0.17 0.35 0.46 0.45 86.41
208 Gard. Bull. Singapore 63(1 & 2) 2011
Legend
HB Water
HBB Non-vegetated
MP Scrubland
HB Mangrove Forest
GE Forest
GR Clouded Area
0 5 10 20 Kilometers
L
Fig. 2. The final automated vegetation map of Singapore produced by filling in the clouded
areas of the 2007 map with patches from the 2006 map.
The map was then overlain with layers showing the managed area in Singapore, and
the Singapore Greenery Map (see Tee et al. 2009). Lastly, it was overlain with a layer
showing areas of primary forest, old secondary forest, freshwater swamp forest, and
mangrove forest, using information from personal communications and observations,
the topographic map of Singapore (Singapore Mapping Unit 2006), the vegetation of
the Bukit Timah and Central Catchment Nature Reserves (Corlett 1997), map of the
freshwater swamp at Nee Soon (Tuner et al. 2006), and the mangrove vegetation of
Singapore (Yee et al. 2010). As intermediate vegetation types exist, especially between
young and old secondary forests, this study adopted a more conservative approach in
classifying such vegetation. Hence, for example, forest that was intermediate between
young and old secondary forest would be classified as young secondary forest.
Results and Discussion
Table 3 lists the spatial extent of each vegetation class and the updated vegetation map
of Singapore is shown in Fig. 3. Vegetation covers 56% of Singapore’s total land area.
Actively managed vegetation occupies 27% of the total landmass while 29% of the
area 1s covered by spontaneous vegetation, which includes scrubland, lowland forest,
freshwater swamp forest, freshwater marsh, and mangrove forest. Primary forest only
Vegetation map of Singapore 209
Table 3. Area, proportion, and number of patches for each vegetation type. The total land area
of Singapore taken here is 72,574.68 ha.
Number of
Vegetation types Area (ha) Proportion (%) panes
Non-vegetated _ 28,270.43 38.85 20075
Managed vegetation 19,972.96 27.45 29075
Scrubland 4,307.54 5:92 8340
Young secondary Forest 14,288.48 19.64 2920
Old secondary forest 994.68 11.337) 42
Primary lowland dipterocarp forest 118.34 0.16 15
Mangrove forest 662.43 0.91 491
Freshwater marsh 76.6 0.11 227),
Freshwater swamp forest 283.12 0.39 125
= a xs = _
Legend
ea Water aay Scrubland
. oa = Non-vegetated = Young Secondary Forest
aa. bo : Ny ee Managed Vegetation fae! Old Secondary Forest
ae) Primary Forest
“ == Mangrove Forest
we Freshwater Swamp Forest
0 5 10 20 km
i I t | (Sod) Freshwater Marsh
Fig. 3. The manually edited vegetation map of Singapore.
210 Gard. Bull. Singapore 63(1 & 2) 2011
occupies 0.28% of Singapore’s total landmass. A total of 118 ha of primary lowland
dipterocarp forest can be found in the Bukit Timah Nature Reserve and in patches
scattered throughout the Central Catchment Nature Reserve. Primary freshwater
swamp forest, which is estimated to be 87 ha, can only be found in the Nee Soon
swamp forest, which is located in the Central Catchment Nature Reserve (Turner et
al. 1996).
Most of the spontaneous vegetation in Singapore belongs to lowland young
secondary forest. The lowland young secondary forest can be further subdivided
into native-dominated forest [e.g., young secondary forest dominated by Adinandra
dumosa (Holttum 1954; Sim et al. 1992)], abandoned plantations, and open woodlands.
However, these subtypes are not reflected in the map because they could only be
poorly separated in the satellite images used. Moreover, extensive ground truthing
in this forest type was not possible as some areas are inaccessible. Nonetheless, we
observed that open woodlands, which are usually dominated by exotics like Acacia
auriculiformis and Leucaena leucocephala, are common in recently cleared or
reclaimed land. In the Nature Reserves in the centre of the island, native-dominated
young secondary forests have been replaced through forest succession by taller, more
species-rich, native-dominated old secondary forests (Corlett 1997), but the eventual
fate of the exotic-dominated secondary forests outside the Reserves is unclear.
The ecology of managed vegetation in Singapore 1s also not well-understood,
despite it constituting 48% of the total vegetation cover. As the major land use type
in Singapore, managed vegetation has significant conservation potential and it has
received increasing attention from the Singapore government lately. The planted
trees can provide foods for birds and mammals. For example, the common palm civet
(Paradoxurus hermaphroditus) has been found to feed on fruits of the rain tree (A/bizia
saman), the most widely planted street tree in Singapore (Xu 2010).
There are some limitations with the map. Firstly, it is based on satellite images
from the years 2006 and 2007, but ground truthing was carried out in 2009 and 2010.
This is likely to have affected the regions-of-interest (ROIs) drawn, hence affecting
the supervised classification, and the accuracy of the map. Secondly, the maximum
likelihood produces a hard (all or nothing) classification, assuming the whole pixel is
homogenous. In reality pixels are rarely homogenous: for example, a pixel classified
as forest might actually not be fully forested. Lastly, intermediate vegetation types
exist, and this would once again affect the ROIs drawn. Despite these caveats, this
map is still a good approximation to the vegetation of Singapore in 2006—2010. Fine-
tuning could be done by using higher resolution satellite images and more ground
truthing. We recommend a long-term follow-up study to document the changes in the
vegetation of Singapore.
ACKNOWLEDGEMENTS. The Centre for Remote Imaging, Sensing and Processing
(CRISP) provided the first author materials and a workplace. The third author acknowledges
funding from the Agency for Science, Technology and Research (A*STAR) awarded to CRISP.
Vegetation map of Singapore 20k
Members of the Plant Systematics Lab of National University of Singapore provided great
help during ground truthing and information of land use types. Special thanks to Outward
Bound Singapore, National Parks Board, Singapore and Public Relations Branch of Ministry
of Defence, Singapore.
References
Corlett, R.T. (1991) Vegetation. In: Chia, L., Ausafur, R. & Tay, D. (eds) The Biophysical
Environment of Singapore, pp. 134—154. Singapore: Singapore University Press.
Corlett, R.T. (1992a) The ecological transformation of Singapore, 1819-1990. J.
Biogeogr. 19: 411-420.
Corlett, R.T. (1992b) The changing urban vegetation. In: Gupta, A. & Pitts, J. (eds)
Physical adjustments in a changing landscape: the Singapore story, pp. 190-214.
Singapore: Singapore University Press.
Corlett, R.T. (1997) The vegetation in the nature reserves in Singapore. Gard. Bull.
Singapore 49: 147-159.
Google Inc. (2009) Google Earth 5.1.3533.1731. California: Google Inc.
Hill, R.D. (1977) The vegetation map of Singapore: A first approximation. J. Trop.
Geogr. 45: 26-33.
Hilton, M.J. & Manning, S.S. (1995) Conversion of coastal habitats in Singapore:
Indications of unsustainable development. Environm. Conservation 22: 307-322.
Holttum, R.E. (1954) A succession of vegetation from bare ground on Singapore
Island. Malayan J. Trop. Geogr. 3: 27-32.
ITT Visual Information Solutions (2007) ENVI version 4.4. Colorado: ITT Visual
Information Solutions.
Mighty Minds (2009) Street Directory: Latest 2009 Edition. Singapore: Mighty Minds
Publishing.
Sim, J.W.S., Tan, H.T.W. & Turner, I.M. (1992) Adinandra be/ukar: an anthropogenic
heath forest in Singapore. Vegetatio 102: 125-137
Singapore Mapping Unit (2006) /:50,000 Topographic Map. Singapore: Mapping
Unit, Ministry of Defence.
Tee, S.P., Yong, W.K., Ng, J.S.C. & Ong, G.Z.M. (eds) (2009) Trees of our Garden
City. A Guide to the Common Trees of Singapore, 2nd ed., pp. 14-15. Singapore:
National Parks Board.
Thomlinson, J.R., Bolstad, P.V. & Cohen, W.B. (1999) Coordinating methodologies
for scaling landcover classifications from site-specific to global: Steps toward
validating global pap products. Remote Sensing Environm. 70: 16-28.
Turner, I.M., Boo, C.M., Wong, Y.K., Chew P.T. & Ali, I. (1996) Freshwater swamp
forest in Singapore, with particular reference to that found around the Nee Soon
Firing Ranges. Gard. Bull. Singapore 48: 129-157.
Wong, P.P. (1985) Artificial coastlines: The example of Singapore. Z. Geomorphol. 57:
175-192.
DVD Gard. Bull. Singapore 63(1 & 2) 2011
Xu, W.T. 2010. Autecology and Public Awareness of the Paradoxurus hermaphroditus
(Pallas, 1777) in an Urban Environment in Singapore. B.Sc. Honours’ thesis.
Singapore: National University of Singapore.
Yee, A.T.K., Ang, W.F., Teo, S.Y., Liew, S.C. & Tan, H.T.W. (2010) The present extent
of mangrove forest in Singapore. Nature in Singapore 3: 147-152.
Gardens’ Bulletin Singapore 63(1 & 2): 213-218. 2011 DAS
Notes on Araceae of Kuala Koh,
Kelantan, Peninsular Malaysia
Zulhazman Hamzah', Mashhor Mansor’ and P.C. Boyce?
'Earth Science Department, Faculty of Agro Industry & Natural Resources,
Universiti Malaysia Kelantan, Locked Bag 39,
16100 Kota Bharu, Kelantan, Malaysia
zulhazman@umk.edu.my
School of Biological Sciences, Universiti Sains Malaysia,
11800 USM, Pulau Pinang, Malaysia
ABSTRACT. A total of 32 species from 11 genera of aroids were collected from Kuala Koh,
Gua Musang, Kelantan. This represents about 23% out of an estimated 140 species and 39% of
the 28 genera of aroids reported for Peninsular Malaysia. These include 24 species that are new
records for Kelantan, including the recently described Homalomena kualakohensis Zulhazman,
Mashhor & P.C.Boyce, and the very rare Rhaphidophora corneri P.C.Boyce, refound after 75
years.
Keywords. Araceae, checklist, Kelantan, Peninsular Malaysia
Introduction
The state of Kelantan is located in the northeast of Peninsular Malaysia and is fortunate
to have remaining extensive areas of lowland forest and thus a rich and diverse
lowland biodiversity, notably in the in the Kelantan delta. Together with extensive
hill dipterocarp forests at Kuala Koh, the limestone hill and montane forests in south
Kelantan and large water bodies at Pergau, resulting in a characteristic distribution of
vegetation and especially composition and diversity of the aroids.
The earliest comprehensive account of the aroids of Peninsular Malaysia and
Singapore is Hooker’s Flora of British India (Hooker 1893), listing 89 species from
18 genera in Peninsular Malaysia. Thirty-two years later, Ridley (1925) conducted a
comprehensive study on aroids of Peninsular Malaysia and recorded 123 species from
23 genera. The latest listing of aroids for Peninsular Malaysia (Mashhor et al. 2011)
documents 140 species in 28 genera, which 25 species endemic. Studies on specific
genera that relate to Peninsular Malaysia are those of Furtado (1939) on Homalonema
Schott; Nicolson (1969) on Aglaonema Schott; Nicolson & Sivadasan (1981) on
Typhonium (Schott); Hay (1996a) on Colocasia Schott; Boyce (1998) on Epipremnum
Schott; Hay (1998) on Alocasia (Schott) G.Don; Boyce (1999) on Rhaphidophora
Hassk; Nguyen & Boyce (1999) on Amydrium Schott; Hay (1996b) and Hay &
Yuzammi (2000) on Schismatoglottis Zoll. & Moritzi; Bogner & Hay (2000) on
Piptospatha N.E. Br.; Boyce & Hay (2001) on Pothos L.; Sofiman et al. (2009) on
Cryptocoryne Fisch. ex Wydler., and Sofiman et al. (2010) on Scindapsus Schott.
214 Gard. Bull. Singapore 63(1 & 2) 2011
To date, there is no comprehensive record of the species of Araceae in Kelantan.
Previous studies on aroids related to Peninsular Malaysia, from Hooker (1893) and up-
dated by Mashhor et al. (2011) showed that only 42 species from 15 genera recorded
were collected from more than one location in Kelantan. Although there are a few
surveys on plants in Kelantan, none has focussed specifically on aroids. For example,
Chee et al. (2005) recorded just three species of aroids (A/gaomena nitidum (Jack)
Kunth, Homalomena humilis (Jack) Hook.f. and Scindapsus scortechinii Hook.f.) in
their checklist survey on plant species of Gunung Stong Forest Reserve. Shamsul et al.
(2005) also noted the same three species of aroids at different localities in the Gunung
Stong Forest Reserve during their survey of seed plants. A recent survey by the first
author of aroids in the granite area of the Jelawang Waterfall, Gunung Stong, revealed
another novel species of aroid in Peninsular Malaysia, Homalomena_ stongensis
Zulhazman, P.C.Boyce & Mashhor, ined. (Zulhazman et al. in press). The listing
offered here is the first attempt to compile an inventory of the aroids for Kelantan.
Materials and methods
The study area is located in Kuala Koh at the southern part of Kelantan in the Gua
Musang District, 180 km from the capital city of Kota Bharu. This area is covered
with lowland moist perhumid dipterocarp forest at an average altitude of 100 m a.s.l.
The surveyed area is at the confluence of two rivers, Sungai Lebir and Sungai Koh.
Sungai Lebir is the main river that joins the Sungai Galas to the Sungai Kelantan at
Kuala Krai.
Aroids were collected from Kuala Koh during field trips on 26-30 March and
31 May-—2 June, 2010. Detailed samplings were made along a 3-km distance along the
Rentis Ara. Specimens were collected with data on species identifications, habitats,
elevation and location (longitude and altitude). The specimens were later brought to
the Universiti Malaysia Kelantan and dried at 60°C. The dried material was processed
as herbarium specimens and incorporated. The specimens were deposited to the Her-
barium of Universiti Malaysia Kelantan, Malaysia. Appendix A shows the herbarium
number for each specimen collected. The living specimens were planted at the Agro-
Park, UMK as a pool genetic collection. The living collections are a vital resource for
Araceae research. Access to a well curated living collection enables plants collected
sterile to be flowered in cultivation. It facilitates crucially important enrichment of
herbarium collections by enabling preparation of photographs and alcohol-preserved
collections, etc. It also allows collection of fresh leaf samples for molecular data, and
other materials for anatomical and developmental research.
Results and discussion
Thirty-two species from 11 genera of aroids were recorded from the study area. This
represents about 23% of recorded species, and 39% of recorded genera for Peninsular
Araceae at Kuala Koh, Peninsular Malaysia DAS
Malaysia. Appendix A lists the aroids recorded from Kuala Koh. This includes 24
species (75% of 32 species of aroids collected) which are new records for Kelantan,
and one species new to science.
Six species of Homolamena Schott including one newly described species,
H. kualakohensis Zulhazman, P.C.Boyce & Mashhor (Zulhazman et al. 2011) were
recorded from this area. Other species are H. pontederiifolia Griff. ex Hook.f., H.
griffithii (Schott) Hook.f., H. wallichii Schott., H. rostrata Griff. and an unidentified
species of the Chamaecladon Supergroup.
Fifteen species from five genera of climbing aroids were collected from the
area. Rhaphidophora corneri P.C.Boyce is one of the most remarkable species found
at Kuala Koh (Boyce et al. in press). A small population of the species was located
on sandy soil on ridge-tops and flat open areas. The Type and hitherto only known
collection was collected by E.J.H. Corner in late 1935 from Kemaman, Terengganu
(Boyce 1999).
A few aroid species were found to be significantly restricted to streams and
associated gallery forests at Kuala Koh, as follows: Schismatoglottis wallichii Hook.f.,
S. calyptrata (Roxb.) Zoll. & Moritzi, S. brevicuspis Hook.f., Apoballis brevipes
(Hook.f.) S.Y.Wong & P.C.Boyce and 4. mutata (Hook.f.) S.Y.Wong & P.C.Boyce.
Scindapsus pictus Hassk. was found on sloping and hilly areas. Alocasia puber
(Hassk.) Schott, a species hitherto considered very rare in Peninsular Malaysia (Hay
1998) was noted to occur in inundated areas close to the stream.
Overall, most of the aroids found are restricted to the forest area, even though
a few species such as the Alocasia longiloba Mig. Complex, Amorphophallus prainii
Hook. f., and Colocasia esculenta (L.) Schott. can also be seen both within forest
and in settlement areas. The last named is not native to Peninsular Malaysia. The
most abundant species noted from this area is Ag/aonema nitidum (Jack) Kunth. This
species is distributed all over the area, especially on dry ridges.
Conclusions
This preliminary study lists 32 species in 11 genera of Araceae from the Kuala
Koh, Kelantan. Among the collection, there are 24 species new for Kelantan and an
undescribed species. The findings indicated that this area is relatively rich in aroids.
The area should repay further study.
ACKNOWLEDGEMENTS. This study forms part of the first author’s doctoral research
on phytogeographic studies of Araceae in Kelantan, Malaysia. The authors would like
to acknowledge the Department of Wildlife and National Parks Peninsular Malaysia
(PERHILITAN) for allowing them to conduct the study in the Kuala Koh National Park. Special
thanks to Mr. Nik Yuszrin Yusof, Ms. Naziah Zaid and Ms. Norzielawati Salleh for their kind
assistance. The first author’s project is funded by Universiti Malaysia Kelantan and through the
short-term research grant R/SGJP/A03.00/00279A/001/2009/000021 via the Faculty of Agro
216 Gard. Bull. Singapore 63(1 & 2) 2011
Industry and Natural Resources. The second and third authors’ fieldwork was supported by
USM Grant No: 1001/ JNC/ AUPRMO01.
References
Bogner, J. & Hay, A. (2000) Schismatoglottideae (Araceae) in Malesia II — Aridarum,
Bucephalandra, Phymatarum and Piptospatha. Telopea 9(1): 2000.
Boyce, P.C. (1998) The genus Epipremnum Schott (Araceae-Monsteroideae-
Monstereae) in West and Central Malesia. Blumea 43: 201.
Boyce, P.C. (1999) The genus Rhaphidophora Hassk. (Araceae-Monsteroideae-
Monstereae) in Peninsular Malaysia and Singapore. Gard. Bull. Singapore 51:
183-256.
Boyce, P.C. & Hay, A. (2001) A taxonomic revision of Araceae tribe Potheae (Pothos,
Pothoidium and Pedicellarum) for Malesia, Australia and the Tropical Western
Pacific. Telopea 9(3): 449-571.
Boyce, P.C., Zulhazman, H. & Sofiman, O.A. Studies on Monstereae (Araceae) of
Peninsular Malaysia IV: The enigmatic Rhaphidophora corneri retound after 75
years. Gard. Bull. Singapore (in press).
Chee, B.J., Lim, C.K., Kamarudin, S., Markandan, M., Shamsul, K., Manap,
T.A., Lim, K.H., Ku Yahaya, K.H., Abdul Razak, A. & Latiff, A. (2005) A
Preliminary Checklist of Plant Species of the Gunung Stong Forest Reserve.
In: Shaharuddin, M.I., Dahalan, T., Abdullah, S.S., Jalil, M.S., Faridah-Hanum,
I. & Latiff, A. (eds) Taman Negeri Gunung Stong, Kelantan: Pengurusan,
Persekitaran Fizikal, Biologi dan Sosio-ekonomi, pp. 319-340. Kuala Lumpur:
Jabatan Perhutanan Semenanjung Malaysia.
Furtado, C.X. (1939) Araceae Malesicae II. Notes on some Indo-Malaysian
Homalomena Species. Gard. Bull. Straits Settlem. 10: 183-238.
Hay, A. (1996a) A new Bornean species of Colocasia Schott with a synopsis of the
genus in Malesia and Australia. Sandakania 7: 31-48
Hay, A. (1996b) The genus Schismatoglottis Zoll. & Moritzi (Araceae:
Schismatoglottideae) in Peninsular Malaysia and Singapore. Sandakania 7:
1-30.
Hay, A. (1998) The genus Alocasia (Araceae-Colocasieae) in West Malesia and
Sulawesi. Gard. Bull. Singapore 50: 221-334.
Hay, A. & Yuzammi (2000) Schismatoglottiseae (Araceae) in Malesia 1:
Schismatoglottis. Telopea 9: 1-177.
Hooker, J.D. (1893) Aroideae: Flora of British India 6: 490-558. Dehra Dun, India:
Bishen Singh Mahendra Pal Singh.
Mashhor M., Boyce, P.C., Sofiman, O.A. & Baharuddin, S. (2011) The Araceae of
Peninsular Malaysia — A Preliminary Checklist, and Keys to the Higher Taxa.
Penang: Universiti Sains Malaysia.
Nguyen, V.D. & Boyce, P.C. (1999) The genus Amydrium (Araceae: Monsteroideae:
Monstereae) with particular reference to Thailand and Indochina. Kew Bull. 54:
379-393.
Araceae at Kuala Koh, Peninsular Malaysia 217
Nicolson, D.H. (1969) A Revision of the Genus Aglonema (Araceae), p. 69. Washington:
Smithsonian Institution Press.
Nicolson, D.H. & Sivadasan, M. (1981) Four frequently confused species of Typhonium
Schott. B/lumea 27: 483-497.
Ridley, H.N. (1925) The Flora of the Malay Peninsula 5:470. London: L. Reeve &
Co. Ltd.
Shamsul, K., Manap, T.A., Kamarudin, S., Markandan, M., Chee, B.J., Faridah-
Hanum, I., Lim, K.H., Ku Yahaya, K.H., Abdul Razak, A. & Latiff, A. (2005) An
additional annotated checklist of seed plants of Gunung Stong Forest Reserve.
In: Shaharuddin, M.I., Dahalan, T., Abdullah, S.S., Jalil, M.S., Faridah-Hanum,
I. & Latiff, A. (eds) Taman Negeri Gunung Stong, Kelantan: Pengurusan,
Persekitaran Fizikal, Biologi dan Sosio-ekonomi, pp. 341-382. Kuala Lumpur:
Jabatan Perhutanan Semenanjung Malaysia.
Sofiman, O.A., Boyce, P.C. & Chan, L.K. (2010) Studies on Monstereae (Araceae) of
Peninsular Malaysia III: Scindapsus lucens, a new record for Malaysia. Gard.
Bull. Singapore 62(1): 9-15.
Sofiman, O.A., Jacobsen, N. & Mashhor, M. (2009) Cryptocoryne of Peninsular
Malaysia, p. 102. Penang: Universiti Sains Malaysia.
Zulhazman, H., Boyce, P.-C. & Mashhor, M. (2011) Studies on Homalomeneae (Ara-
ceae) of Peninsular Malaysia IV: Homalomena kualakohensis, A new species
from Kelantan, Malaysia. Acta Phytotax. Geobot. 61(1): 37-41.
Zulhazman, H., Boyce, P.C. & Mashhor, M. Studies on Homalomeneae (Araceae) of
Peninsular Malaysia III: Homalomena stongensis, A remarkable new species
endemic to Gunung Stong, Kelantan. Gard. Bull. Singapore (in press).
Appendix A. List of aroids found in Kuala Koh, Gua Musang, Kelantan. E = endemic to
Peninsular Malaysia; LC = Living collection; NR = new record for Kelantan; R = rare.
Genus Species Herbarium No. Habitat Remarks
Aglaonema nitidum (Jack) Kunth UMK 87 Dry ridges
simplex Blume LC Dry ridges NR
Alocasia longiloba Migq. UMK 141 Ridges, open sites, dry spots,
often within shrub areas
puber (Hassk.) Schott UMK 139 Streambanks, wet sites NR, R
Amorphophallus prainii Hook.f. EC Dry ridges, often in shrub NR
areas
Amydrium medium (Zoll. & UMK 92 Open canopy area, oftenon NR
Moritzi) Nicolson big trees
Anadendrum microstachyum de Ver UMK 123 Shady wet areas, often on NR
& Becker small trees
Apoballis brevipes (Hook.f.) UMK 110 Stream gallery, open areas NR
S.Y. Wong & P.C.
Boyce
mutata (Hook.f.) S.Y. . UMK 104 Stream gallery, open areas NR
Wong & P.C. Boyce
218
Epipremnum
Homalomena
Rhaphidophora
Schismatoglottis
Scindapsus
giganteum (Roxb.)
Schott
griffithii (Schott)
Hook.f.
pontederiifolia Griff.
ex. Hook.f.
wallichii Schott
rostrata Griff.
kualakohensis
H. Zulhazman,
M.Mashhor &
P.C.Boyce
Chamaecladon
supergroup (Sp.1)
beccarii (Engl.) Engl.
corneri P.C.Boyce
falcata Ridl.
korthalsii Schott
lobbii Schott
maingayvi Hook.f.
brevicuspis Hook.f.
calyptrata (Roxb.)
Zoll. & Moritzi
scortechinii Hook.f.
wallichii Hook.f.
hederaceus Schott
perakensis Hook.f.
pictus Hassk.
treubii Engl.
sp. 1
sp.2
Gard. Bull. Singapore 63(1 & 2) 2011
UMK 128 Open canopy and dry areas, NR
high ridges, often on big
trees
UMK 1 Slopes, ridges
IMS Slopes, ridges NR
UMK 3 Slopes, ridges NR
ILC Slopes, ridges NR
UMK 6 Slopes, ridges NR, E
LC Slopes, wet, stream margin, NR
clay soil
LEC Streambanks, on rock and
soil, wet and shady area
UMK 31 Sandy soil, ridge-tops, flat NR, E
open areas
Ee Streambanks, on rock and NR
soil, wet areas
UMK 47 Shady wet areas, ridges,
often on big trees
UMK 50 Shady wet areas, ridges, NR
often on small trees
UMK00033 Open canopy area, steep
slopes, often on small trees
ILC Streambanks, on wet and NR
shady areas
UMK00059 On slopes, along the trail,
shady area
UMK00056 Streambanks, slopes NR, E
UMK00052 Stream gallery forests
UMKO00161 Open canopy areas, ridges, NR
often on small trees,
hemiepiphytic
ILC Open areas, along trail, NR
hemiepiphytic
Ke On slopes, ridges, on soil NR
and tree, surrounding with
leaf litter
UMK00159 Shady areas, flat areas along NR
trail, hemiepiphytic
IL{C Wet sites, hemiepiphytic NR
ILC Half open canopy area, NR
hemiepiphytic
Gardens’ Bulletin Singapore 63(1 & 2): 219-236. 2011 219
Diversity and assessment of plants
in Mt. Kitanglad Range Natural Park,
Bukidnon, Southern Philippines
Victor B. Amoroso’, Socorro H. Laraga and Bridget V. Calzada
Department of Biology, Central Mindanao University
University Town, Musuan, Bukidnon 8710, The Philippines
‘amorosovic@yahoo.com (corresponding author)
ABSTRACT. This research describes the vegetation types, determines the diversity and assesses
the conservation status of vascular plants in Mt. Kitanglad Range Natural Park, Bukidnon
Province. Twelve 20 m = 20 m nested plots were established per vegetation type. A transect
survey with 34 plots revealed three vegetation types, namely the agroecosystem, lower montane
forest and mossy forest, with 661 species, 264 genera, and 106 families enumerated. Plant
species richness and diversity decreases as the altitude increases, and the mossy forest had the
lowest species diversity. Lithocarpus sp. obtained the highest Species Importance Value (SIV)
for trees in both lower montane and mossy forests together, while Leptospermum sp. had the
highest SIV in the mossy forest. Tree profile analysis showed that the lower montane forest had
the highest mean number of species (7.9 spp.) and individuals (26.9 individuals), mean height
(11.12 m) and mean diameter at breast height (dbh, 39.40 cm). The upper mossy forest had the
lowest mean number of species (4.4 spp.), individuals (20.2 individuals), average height (7.03
m) and average dbh (16.60 cm). We assess 92 threatened and 82 rare species; 108 endemic
species, 50 economically important species, 56 species newly recorded in the locality and 20
species newly recorded for the Philippines. Policy recommendations are given for protecting
the remaining threatened, endemic and rare species of plants and their habitats.
Keywords. Conservation status, Mindanao, Mt. Kitanglad, Philippines, species diversity,
vascular plants, vegetation types
Introduction
Mt. Kitanglad Range Natural Park has been declared as a protected area on November
9, 2000 and an ASEAN Heritage Park on October 29, 2009. Encompassing 37,236
ha over the North Central portion of Bukidnon and with the highest elevation of
2938 m, the park is the headwaters of three major river systems and sees frequent
visitors (NORDECO 1998). Many faunal species were recorded in the park (Heaney
et al. 2006, NORDECO 1998) but meager information is available about the richness
and status of plants, especially at the northeastern part. It was therefore important
that an inventory and assessment of the floral resources be conducted to generate
knowledge on plant diversity and status for the conservation and protection of the
remaining biodiversity. The findings of this research will be used as the basis for policy
formulation by the Protected Area Management Board (PAMB) of Mt. Kitanglad,
i)
i)
S)
Gard. Bull. Singapore 63(1 & 2) 2011
Department of Environment and Natural Resources (DENR) and for conserving and
properly managing the threatened, endemic, rare and economically important species
of plants and their habitats.
Objectives
The project objectives were to obtain an inventory, and assess and conserve the
threatened, endemic, rare and economically important plants in the Mt. Kitanglad
Range Natural Park, Bukidnon, Southern Philippines. Specifically, it aimed to (1)
identify and describe the vegetation types; (2) determine the diversity of vascular
plants; (3) assess their conservation status; (4) record plant habitats and distribution;
and, (5) recommend policy measures regarding the protection and conservation of the
threatened/endemic species and their habitats.
Methods and materials
Prior Informed Consent (PIC) and selection of local researchers
To satisfy the legal requirements of EO 247 (Bioprospecting) and RA 9147 (Wildlife
Resources Conservation and Protection Act), prior informed consent from the
community was obtained by presenting the research proposal. Likewise, this research
proposal was presented to the members of the Protected Area Management Board
(PAMB) of Mt. Kitanglad Range Natural Park for their approval and eventual issuance
of the Gratuitous Permit from the Department of Environment and Natural Resources.
Selection of local researchers (Forest Guides) was made with the stakeholders
in Sitio Intavas, Barangay La Fortuna, Impasug-ong, Bukidnon based on their
sufficient indigenous knowledge of the floral resources in the study sites. Being co-
researchers and since the nature of the research is participatory, the Forest Guides were
compensated and involved during the entire duration of field work.
Identification and description of vegetation types
Field reconnaissance and a transect survey were conducted to identify and describe
the vegetation types by considering the species richness and dominance, canopy cover,
tree profile, altitude, location and other ecological parameters. A GPS was used to
determine the location of each vegetation type.
Survey, establishment of sampling sites, collection and processing of specimens
Several transects along the landscape were laid out to inventory and assess the plant
species observed. Likewise, a transect belt of 2 km x 10 m wide was established per
vegetation type. Within the transect belt, an inventory and assessment of plants were
conducted, and their local names, uses and altitude were recorded. Representative
specimens collected were pressed, poisoned and mounted as herbarium vouchers
using the wet method. Duplicates of the herbarium specimens were sent and deposited
Vegetation and plant diversity in Mt. Kitanglad Range Natural Park 2A
at the Philippine National Herbarium and Herbarium of the Taiwan Forestry Research
Institute.
Diversity Indices
A total of 34 sampling plots, each 20 m x 20 m, were established in all vegetation
types, each type with 12 sampling plots except for the upper mossy forest, with 10
plots. Tree enumeration was carried out for all individuals with a diameter at breast
height (dbh) of 10 cm and larger. Within these plots, a 5 m < 5 m subplot was laid out
to determine the species richness of pteridophytes, of which all occurring individuals
were assessed. We assessed species diversity by using the Shannon index of general
diversity (H’):
for trees: H’ = - © ndbh/Ndbh log ndbh/Ndbh
(where, ndbh = diameter at breast height of individual tree species
Ndbh = total diameter at breast height of all tree species);
and for pteridophytes: H’ = - = [n/N log [n,]/N
(where, n, = number of individuals in an area,
and N = total number of individuals in an area).
The Species Importance Value (SIV) was computed using the formula of Brower and
Banas) 7):
SIV or ni= RD + RF + Rdom
where, RD = relative density, RF = relative frequency, Rdom = relative
dominance.
Identification and assessment of conservation status
The collected plants were identified using taxonomic keys from floras and monographs
of Merrill (1923-1926): Linder (1987); Madulid (1995): Kalkman et al. (1995-1996);
Editorial Committee of the Flora of Taiwan (1996): Rojo (1999): Jebb & Cheek (2001):
Cootes (2001); Barcelona et al. (1996); Tan et al. (1996): Zamora & Co (1986): and
Amoroso et al. (1993, 1996, 1997).
The assessment of status for each species, whether threatened, endemic, rare or
economically important, was determined with the help of the national list of threatened
Philippine plants (Fernando et al. 2008), the IUCN (2007) and from published floristic
works and monographs.
Definitions of terms adopted from the International Union for the Conservation
of Nature (IUCN) (2007), Fernando ef al. (2008), Zamora (1986) and Department of
Environment and Natural Resources (DENR) Administrative Order No. 2007-01 (as
defined in its Section 5 of R.A. 9147) include:
a) Threatened Species - Actively threatened with extinction and its survival is unlikely
without protective measures. Threatened species fall under three categories as defined
by Fernando et al. (2008):
al) Critically Endangered — A taxon 1s critically endangered when it is facing
an extremely high risk of extinction in the wild in the immediate future.
i)
i)
i)
Gard. Bull. Singapore 63(1 & 2) 2011
a2) Endangered — A taxon is endangered when it is not critically endangered
but is facing a very high risk of extinction in the wild in the medium-term
future.
a3) Vulnerable — A taxon is vulnerable when it is not critically endangered
or endangered but is facing a high risk of extinction in the wild in the
medium term future.
a4) Other threatened species — A taxon belongs to other threatened species
when it is under threat from adverse factors, such as over collection,
throughout its range and is likely to move to the vulnerable category in the
near future.
a5) Other wildlife species —A taxon belongs to other wildlife species when it
has been evaluated but does not satisfy the criteria for any of the categories
Critically Endangered, Endangered, Vulnerable or Other Threatened
Species, but have the tendency to become threatened due to predation
and destruction of habitat or other similar causes as may be listed by the
Secretary upon the recommendation of the National Wildlife Management
Committee.
b) Rare Species — Not under immediate threat of extinction but occurring in such small
numbers or in such localised or specialised habitat that it could quickly disappear 1f the
environment worsens; needs monitoring.
c) Depleted Species — Although sufficiently abundant for survival, the species has been
heavily depleted and in decline as a result of natural causes of human activities.
d) Endemic Species — Confined to a certain geographical region or its parts.
e) Economically Important Species — Based on known usefulness whether medicinal,
ornamental, food, construction material, etc.
Results and discussion
Vegetation types and distribution
Transect survey and establishment of sampling plots was carried out along the trail in
the northeastern part of Mt. Kitanglad Range Natural Park to identify and describe the
vegetation types by considering the coordinates, species richness and dominance, tree
profile, altitude, and other ecological parameters (Fig. 1). Three vegetation types were
identified from 1200 m asl to the peak of the park as described below (Fig. 2).
The agro-ecosystem (08°10°17°N, 124°56°09”E) ranges from 1200 m to 1700
m asl. Potatoes (Solanum tuberosum L.), cabbage (Brassica oleracea L.), carrots
(Daucus carota L.) and tomatoes (Lycopersicum esculentum Mill.) dominated this
vegetation. The original vegetation was dipterocarp forest but this was logged and
later converted to agricultural land and planted with cash crops. Threatened plants
such as Cyathea contaminans (Wall.) Copel., (anonotong), Podocarpus macrocarpus
de Laub. and Dicranopteris linearis (Burm.) Underw. (agsam), which is an indicator
of a disturbed habitat, was observed at the edge of the vegetable plantation (Fig. 2A).
The agroecosystem in Mt. Malindang is also dominated by crops like vegetables,
cereals and agroforestry species (Amoroso et al. 2006).
Vegetation and plant diversity in Mt. Kitanglad Range Natural Park 223
Mindanao
VEGETATION
MT. KITANGLAD RANGE
NATURAL PARK
SCALE 1 : 224,000
~-S-ipmr@ MLN, OF MANOLO
oN se
LEGEND:
OTransect [Mf Lower
Walk Montane
Fig. 1. Location of Bukidnon Province on Mindanao (upper map), and layout of the transect
walk and 34 sampling plots in the Mt. Kitanglad Range Natural Park (lower map).
The /ower montane forest (08°09°54°N, 124°55°58”E) already begins from the
foot of Mt. Kitanglad and ranges from 1700 m to 2100 masl. This forest is characterised
by the presence of numerous species of mosses, lichens and other epiphytes. The
dominant tree species include Phyllocladus hypophyllus Hook.f. (mountain tungog),
Lithocarpus spp. (ulayan), Erythrina subumbrans (Hasskarl) Merr. (anii), while the
common shrubs observed included the endemic Hydrangea scandens Ser., Drimys
piperata Hook f. and several Medinilla spp. Emergent trees are 5—20 m tall, averaging
224 Gard. Bull. Singapore 63(1 & 2) 2011
Fig. 2. Panoramic view of Mt. Kitanglad Range Natural Park, Intavas, Impasug-ong, Bukidnon
showing the vegetation types. A. Agroecosystem. B. Montane Forest. C. Mossy Forest. Photo
by V. Amoroso, June, 2007.
12 m. The moss cover ranges from 50-75%. The edge of the montane forest is usually
inhabited by Trema orientalis (L.) Blume (andaluyong), Pteridium aquilinum (L.)
Kuhn (sigpang or bracken) and Cyathea spp. (Anonotong or tree ferns) (Fig. 2B). This
is similar to what was observed by Amoroso et al. (2006) in the montane forest on
Mt. Malindang, which has a high relative humidity and rainfall and has trees on the
average taller than in the mossy forest; they also noted that the moss layer was less
conspicuous than in the mossy forest.
The mossy forest begins from 2100 m asl up to the peak as described below.
The branches and trunks of trees and the forest floor were largely covered with mosses,
hence the name mossy forest (Amoroso et al. 2006). We consider there to be lower and
upper facies of these mossy forests from plant diversity considerations.
(a) The lower mossy forest (08°09°27”N, 124°55°49”E to 08°09°16”N, 124°55’30”E)
starts from 2100 m asl, reaching 2400 m asl. Moss cover ts thick in this vegetation type,
over the ground, roots, trunks and branches of trees. Lithocarpus sp., Phyllocladus
hypophyllus Hook.f. and Podocarpus spp. were the most abundant trees. Tree height
ranges 7-13 m, with an average of 10 m.
(b) The upper mossy forest (08°09°16”"N 124°55°30”E to the peak at 08°08°38"N
124°55°06” E) starts from 2401 m asl to 2900 m asl. Moss cover was very thick in
this vegetation type and largely covered the forest floor, roots, and the twisted trunks
and branches of trees. Leptospermum javanicum Blume was an abundant tree, while
Rhododendron spp., Dimorphanthera apoana (Merr.) Schltr, Vaccinium spp. and
Rubus spp. were abundant shrubs. Tree height ranges 6—12 m, with an average of 9 m.
The summit of the park has an abundance of the dwarf bamboo (Yushania
nitakavamensis (Hayata) Keng f.) and the wet ground is covered with Nertera diffusa
(Mutis ex L.f.) Druce, Sphagnum moss and lycopods.
i)
i)
Nn
Vegetation and plant diversity in Mt. Kitanglad Range Natural Park
Amoroso et al. (2004) observed that the mossy forest of Mt. Malindang was
characterised by the presence of small trees with prop roots and aerial roots developing
at 1-few meters from the base of irregularly shaped tree trunks. The presence of prop
roots appeared to be correlated to steep slopes, which were prevalent in this zone.
As in other mossy forests in the Philippines, the trees were dwarfed and their trunks
gnarled, especially those that grew near the mountain peak, possibly due to strong
wind pressure.
If Mt. Kitanglad has three vegetation types, Mt. Hamiguitan showed five
vegetation types (Amoroso et al. 2009), while Mt. Malindang exhibited six (Amoroso
et al. 2004).
Species richness and diversity
The sampling plots and transect survey enumerated a total of 661 species, 264 genera,
and 106 families of vascular plants (Table 1). There were 439 pteridophyte species, 11
gymnosperm species and 211 angiosperm species.
Table 1. Number of families, genera and species of plants in Mt. Kitanglad Range Natural
Park, based on the present study of sampling plots and the transect survey. Of 661 total taxa,
495 were identified to the species level.
Number of
Plant group
Families Genera Species / Taxa
Weeriocn —s- ery | }
Ferns 25 100 408
Lycopods 3 4 31
Gymnosperms 4 8 1]
Angiosperms 74 152 211
TOTAL 106 264 661
The Philippines has a total of 9,060 species of vascular plants, or perhaps
more (Madulid 1991). Of these, Mt. Kitanglad has 42.8% of the pteridophytes, 33.3%
of the gymnosperms and 2.6% of the angiosperms The species richness of vascular
plants in Mt. Kitanglad (7.3%) is lower than in Mt. Hamiguitan (9.6%) and Mt.
Malindang (12.8%). The lower species richness of angiosperms in Mt. Kitanglad is
due to the absence of dipterocarp forest, whereas this forest type was found on Mt.
Malindang and Mt. Hamiguitan. Mt. Kitanglad, however, has a higher species richness
of pteridophytes compared to the Mt. Hamiguitan and Mt. Malindang (Table 2).
Pteridophyte, as well as tree, diversity decreases from lower montane to mossy
forest (Table 3). The same pattern was observed for trees at Mt. Malindang and Mt.
Hamiguitan (Amoroso et al. 2006, 2009).
i)
i)
ON
Gard. Bull. Singapore 63(1 & 2) 2011
Table 2. Plant species richness of Mt. Kitanglad, Intavas, Bukidnon and two other protected
areas in Mindanao, compared with overall statistics for the Philippines.
Total number (and percentage) of species
Plant group Malindang Hamiguitan
- Philippines Mindanao’ (Amorosoet (Amorosoet Kitanglad
al., 2006) al., 2009)
632
Pteridophytes 1027 (61.5%) 280 (27.3%) 155(15.0%) 439 (42.8%)
é 0
Gymnosperms 33 No data 11(33.3%) 25 (75.8%) 11(33.3%)
Angiosperms 8000+ No data 873 (10.9%) 698 (8.7%) 211 (2.6%)
TOTAL 9060+ No data 1164 (12.8%) 878 (9.6%) 661 (7.3%)
Table 3. Plant species diversity (Shannon index of general diversity, H’) in different vegetation
types of the Mt. Kitanglad Range. Pteridophyte enumerations are based on 5 m x 5 m plots,
with 10 plots in Lower Montane Forest, 12 plots in Lower Mossy Forest, and 10 plots in Upper
Mossy Forest. Tree enumerations are based on 20 m =< 20 m plots, with 12 plots in Lower
Montane Forest and Lower Mossy Forest, and 10 plots in Upper Mossy Forest.
Mean number of
Vegetation type i Mean diversity value
cs YP Individuals Species
Pterido- Trees Pterido- _—‘ Trees Pterido- Trees
phytes phytes phytes
Lower Montane
;
(1700-2100 m) Pee 26.9 7.48 7.9 0.45 0.80
Lower Mossy
: ; : |
(2100-2400 m) 45.8 24.8 9.5 6.9 0.83 0.84
araieee 48.8 20.2 7.0 4.4 0.69 0.72
(2400-2800 m)
The pteridophyte diversity value was higher (H’=0.83) in the lower mossy
forest than in both the lower montane and upper mossy forests with H’= 0.45 and 0.69,
respectively. Among trees, however, the highest diversity value was obtained in the
lower mossy forest (H’=0.84), followed by the lower montane and upper mossy forests
with H’=0.80 and H’=0.72, respectively. This implies that the diversity is highest at
mid-elevations, although comparison with lowland vegetation types was not possible
in this study. As a comparison, mossy forests in Mt. Kitanglad and Mt. Hamiguitan had
lower diversity values. However, the upper mossy forest has a significantly reduced
diversity compared with the lower mossy forest. In terms of species richness, however,
Vegetation and plant diversity in Mt. Kitanglad Range Natural Park Dai}
it is clear that there are progressively fewer species from lower montane forest
upwards. These facts support the contention of various ecologists that the number of
species (i.e., richness) or diversity values at higher altitudes is lower as a response to
increasing environmental stresses like wind pressure, steep slopes, thin soil substrates,
etc. (Perez 2004).
Species Importance Values (SIV)
Species importance values determine the dominant species in an area and at the same
time provide an overall estimate of the influence of these species in the community.
The removal of these species from the community will greatly affect the physical and
biological structure of the community.
The species of pteridophytes which obtained the highest Species Importance
Value (SIV) in the lower montane forest (Table 4) were Mecodium reinwardtii (van
der Bosch) Copel., Trichomanes sp., Hymenophyllum sp., Nephrolepis cordifolia
(Linn.) Presl. and Grammitis sp. while for lower mossy forest, they were Plagiogvria
pycnophylla (Kunze) Mett., Phymatosorus sp., Plagiogyria christii Copel.,
Hymenophyllum sp. and Microsorum scolopendria (Burm.f.). Further, the upper mossy
forest species with the highest SIV were Hymenophyllum sp., Asplenium normale
Don, Dicranopteris, Humata repens (Linn.) Diels, and Plagiogyria glauca (Blume)
Mett. Hymenophyllum sp. and Plagiogyria christii were the top two pteridophytes
with the highest SIV values on Mt. Kitanglad. This observation is in consonance with
the results of Amoroso et al. (2009).
The tree species which obtained the five highest SIV (Species Importance
Values) in the lower montane forest includes: Lithocarpus sp., Fagraea blumei G.
Don, Melicope sp., Phyllocladus hypophyllus and Cinnamomum mercadoi Vidal. The
lower mossy forest had the following species with the five highest SIV: Lithocarpus
sp., Phyllocladus hypophyllus, Leptospermum sp., Syzygium sp. and Podocarpus
costalis C. Presl. For the upper mossy vegetation, Leptospermum sp., Dacrycarpus
cumingii (Parl.) de Laub., Fagraea blumei, Phyllociadus, and Podocarpus sp. were
the five species with the highest SIV. For the trees, Leptospermum sp. and Lithocarpus
sp. had the highest SIV value on Mt. Kitanglad. This finding is also supported by the
study of Amoroso et al. (2009). Lithocarpus sp. ranked first in both lower montane and
lower mossy forests of Mt. Kitanglad, with SIV of 115% and 94.32%, respectively. At
Mt. Hamiguitan, Agathis philippinensis Warb. ranked first in both lower montane and
mossy forest. It is noteworthy to mention that Hymenophyllum sp. and Plagiogyria
christii with high SIV were observed in the three vegetation types (Amoroso et al.
2009).
According to Krebs (1994), variation in important species may be caused by
differences in the response of various species to environmental conditions. He also
noted that elevation provides complex environmental gradients including temperature,
rainfall and relative humidity. It was noted that the species composition of the three
vegetation types here differed, suggesting that habitat differences catered differently
to the requirements of tree species.
228 Gard. Bull. Singapore 63(1 & 2) 2011
Table 4. Species importance values of pteridophytes and trees in the different vegetation types.
SIV Trees SIV
Vegetation Pteridophytes
(%) (%)
Mecodium reinwardtii 94.75 Lithocarpus sp. 115.0
Lower Montane oe : :
(1700-2100 m) Trichomanes sp. 80.67 Fagraea blumei 85.67
Hymenophyllum sp. 76.63 Melicope sp. 76.42
Nephrolepis cordifolia 72.27 ~~ Phyllocladus hypophyllus 66.95
Grammitis sp. 64.32 | Cinnamomum mercadoi 64.84
_ Plosion ap eneniiien 154.67 ; Lithocarpus sp. 94.32
Aone Phymatosorus sp. 137.00 Phyllocladus hypophyllus — 83.50
Plagiogyria christii 102.01 Leptospermum sp. 80.32
Hymenophyllum sp. 85.96 Syzygium sp. 70.16
Microsorum scolopendria 81.99 Podocarpus costalis 60.00
; a Fyereeine aa ; 209.66 enema i. : 168.91
Bae Asplenium normale 138.94 Dacrycarpus cumingii 78.99
Dicranopteris linearis 111.93. Fagraea blumei 72.91
Humata repens 97.46 ~Phyllocladus hypophyllus — 57.29
Plagiogyria glauca 95.11 | Podocarpus costalis 56.87
Tree profile
Table 5 shows the mean number of species and individuals, average height and average
diameter at breast height (dbh) of trees in the sampled plots of the different vegetation
types. The lower montane forest obtained the highest average number of individuals
in 20 m x 20 m plots with 26.9 individuals, while upper mossy forests had the lowest
average number of trees with 20.2 individuals. As altitude increases, the average
number of individuals (of 10 cm diameter or bigger) decreases. The highest average
height (11.12 m) and average dbh (39.30 cm) were recorded from trees of the lower
montane forest. These values decrease with elevation. The lowest value for average
height and average dbh were recorded in the upper mossy forest. Lithocarpus sp. was
the tallest tree recorded in the lower montane forest and lower mossy forest, attaining
25.0 m and 24 m, respectively. Dacrycarpus cumingii (20 m) were tallest in the upper
mossy forest.
Tree profile diagrams for the 34 plots were individually made to record the
species distribution and indicate relative heights of trees in the 20 m = 20 m plots.
From these, canopy cover in the sampling plots were calculated and ranged from 70—
95% in the lower montane forest compared 60-80% in the lower mossy forest, and
Vegetation and plant diversity in Mt. Kitanglad Range Natural Park 229
Table 5. Mean number of individuals, average height and diameter at breast height (dbh) of
trees in different vegetation types on Mt. Kitanglad Range, Intavas, Bukidnon.
Calculated based on enumerations of 12 plots each in Lower Montane Forest and Lower Mossy
Forest, and 10 plots in Upper Mossy Forest; each plot 20 m = 20 m.
Mean number of
Vegetation type py SIABE fverage
5 Individuals Species height (m) dbh (cm)
Lower Montane (1700-2100 m) 26.9 12 11.12 39.30
Lower Mossy (2100-2400 m) 24.8 6.9 9.60 21.69
Upper Mossy (2400-2800 m) 20.2 44 7.03 16.60
10—15% in the upper mossy forest. Since tree coordinates were taken in each of the 34
sampling plots, long-term monitoring of tree dynamics will be possible.
Assessment of conservation status
The habitat of the plants, whether epiphytic, terrestrial or petrophytic / lithophytic, was
noted. The distribution of threatened, endemic and economically important species of
plants was mapped based on elevation and vegetation types. Recording of altitudinal
distribution of the threatened, endemic, and economically important species in these
vegetation types will be an important basis in allocating priority to their protection and
conservation.
The conservation status of each species was noted and recorded. This was
carried out to establish a foundation for their protection, conservation and monitoring.
Of the total number of taxa, only 495 species have been identified up to the species
level. Of these species, about 92 species were recorded as threatened, 82 rare species,
108 endemic species, 50 economically important species, 56 species as new records
for the locality and 20 species as new records for the Philippines (Table 6).
Out of the 9060+ vascular plant species of the Philippines, 530 are threatened,
including 85 pteridophytes, 5 gymnosperms and 440 angiosperms (Fernando et
al. 2008). Of this number, 17.4% threatened species are located in Mt. Kitanglad
comprising 77, 7 and 8 species of pteridophytes, gymnosperms, and angiosperms,
respectively (Table 7). The percentage of threatened species in Mt. Kitanglad is 7%
and 11% higher than in Mt. Malindang and Mt. Hamiguitan, respectively.
The Philippines has a total of 3557 endemic species, including 351
pteridophytes, 6 gymnosperms, and 3200 angiosperms (Madulid 1991, Fernando et
al. 2008). Mt. Kitanglad has 21% of this endemism, which is higher than for Mt.
Malindang, but lower than for Hamiguitan. It has, however, a higher percentage
endemism of pteridophytes as compared to both mountain ranges (Table 8). The
high species endemism in Mt. Hamiguitan may be due to effects of the specialised
ultramafic geology there.
230 Gard. Bull. Singapore 63(1 & 2) 2011
Table 6. Number of threatened, endemic and economically important plants, and new records
for Mt. Kitanglad Range Natural Park, Intavas, Impasug-ong, Bukidnon. TS - Threatened
species; RS - Rare species; ECS - Endemic species; NRL - New record for locality; EIS -
Economically important species; NRP - New records for the Philippines.
Status
Plant groups
TS RS ECS EIS NRL NRP
Pteridophytes
Ferns UD 70 43 40 47 19
Lycopods 2 = 2 3 3 1
Gymnosperms 7 4 ] 0 0
Angiosperms 8 4 62 4 6 0
TOTAL TZ &2 108 50 56 20
Table 7. List of some threatened and endemic species in Mt. Kitanglad Range Natural Park.
Conservation status rankings: CR - Critically Endangered; EN - Endangered; OTS - Other
Threatened Species; OWS - Other Wild Species; VU - Vulnerable; ECS - Endemic. Vegetation
types: LMo - Lower Montane; LM - Lower Mossy; UM - Upper Mossy. A denotes the agro-
ecosystem lower down.
Vegetation
Family Species Status ane Altitude (m)
1 Aspleniaceae Asplenium nidus L. VU A La
2 Aspleniaceae Asplenium vittiforme Cav. VU LMo 1935, 2030
Blechnum fraseri (A.
7780-2
Cunningham) Luerss. ae ee pe
3. Blechnaceae
4 Cyatheaceae Cyathea elmeri Copel. VU LMo 1935-2050
Cyathea philippinensis Vie
5 C ; pi
5 Cyatheaceae Baker ECS LMo 030
GuaGyatneneese Cyathea contaminans ( Wall.) VU LMo 2030
Copel.
ree Dennstaedua williamsi EN LMo 3020
Copel.
8 Dryopteridaceae Polystichum elmeri Copel. OWS LM 2280
Om Ophiotiacsicear Botrychium daucifolium VU LM 7300
Wall.
Aglaomorpha cornucopia VU,
10 Pol eae : 179
olypodiaceae (Gapelnog: ECS Mo 797,
(ental podiaeene Aglaomorpha heraclea VU LM 2245
(Kuntze) Copel.
Vegetation and plant diversity in Mt. Kitanglad Range Natural Park
t2
21
22
Polypodiaceae
Polypodiaceae
Polypodiaceae
Psilotaceae
Araucariaceae
Podocarpaceae
Podocarpaceae
Ericaceae
Ericaceae
Lauraceae
Rosaceae
Microsorum punctatum (L.)
Copel.
Microsorum sarawakense
(Baker) Ching
Microsorum scolopendria
(Burm.f.)
Tmesipteris lanceolata
Dang.
Agathis philippinensis Warb.
Podocarpus costalis C.
Presl.
Podocarpus macrocarpus
de Laub.
Rhododendron kochii Stein
Rhododendron javanicum
(Blume)
Cinnamomum mercadoi
Vidal
Rubus heterosepalus Mert.
LMo
LMo
LMo, LM
LMo, LM,
UM
LMo
LMo
LMo, LM,
UM
LMo, LM
LMo, LM
LMo, LM
LMo
231
1711
1797
1797-2280
2050-2600
1945
2113
1985-2840
2245-2495
2245-2495
2050-2300
2010
Table 8. Endemism in the Philippines compared to that at Mt. Kitanglad Range, Intavas,
Bukidnon. Based on specimens identified to the species level. A dash refers to lack of data.
Malindang data from Amoroso et al. (2006); Hamiguitan data from Amoroso et al. (2009).
Plant group
Philippines Mindanao Malindang
Hamiguitan
Total number of species (Tspp) and number endemic (Espp)
Kitanglad
Tspp Espp Tspp Espp Tspp Espp Tspp Espp Tspp- Espp
Pteridophytes 1027 351 632 183 246 2s 99 Z 363 -
(11%) (9%) (12%)
Gymnosperms 33 6 as 3 11 s 13 1]
ymnosp (27%) (7%) (9%)
Angiosperms 8000+ 3200 — — 450 ee 365. 153" 121 a
25/0 (41%) (5 0)
138 163 108
TOTAL 9060+ 3557 — — 2
oo (16%) au (34%) (21%)
232 Gard. Bull. Singapore 63(1 & 2) 2011
New records
Nineteen species of fern and one species of fern ally are new records for the Philippines,
while 50 species of pteridophyte and six species of angiosperm are new records for
the locality (Table 6). A significant output of this research was the new record of
Athyrium erythropodum Hayata (Woodsiaceae) (Fig. 3) for the Philippine flora (Liu et
al. 2008). This species had previously been recorded as endemic to Taiwan and was
only subsequently discovered on Mt. Kitanglad. Another species newly recorded for
the Philippines is Huperzia monticola Underw. & F.E Loyd (Fig. 4). This was earlier
reported in Sumatra, Indonesia, and is now recorded for the first time on Mt. Kitanglad.
i 4 wt A er A TRS 5.
Fig. 3. Athyrium erythropodum Hayata, a new Philippine record. A. Frond. B. Sori with indusia.
C. Scales on stipe base. D. habit. Photographs from Yea-Chen Liu et al. 2009.
Vegetation and plant diversity in Mt. Kitanglad Range Natural Park 33;
Fig. 4. Huperzia monticola Underw. & F.E. Loyd. A. Habit. B. Aerial axis showing microphylls
(a) and sporangia (b). Photos by V. Amoroso.
Dissemination of Information, Education and Communication (IEC) Materials
The Information, Education and Communication (IEC) materials were prepared in the
form of flyers, with relevant information such as scientific name, local name, family
name and status of the floral species (Fig. 5). These help to enhance community
awareness for the conservation and protection of these species, and will be disseminated
to the community and local researchers and guides for their use.
Policy recommendations
The results of this project should be useful to Local Government Units (LGUs) of
[mpasug-ong and other municipalities and communities around the park; the Protected
Area Wildlife Division (PAWD) of the Department of Environment and Natural
Resources; and the Protected Area Management Board (PAMB) in the formulation of
policies and ordinances to protect and conserve the remaining botanical resources of
the Mt. Kitanglad Range Natural Park.
234 Gard. Bull. Singapore 63(1 & 2) 2011
Asplenium nidus Asplenium vittaeforme Blechnum fraserii
ASPLENIACEAE ASPLENIACEAE BLECHNACEAE
Botrychium daucifolium Ophioglossum pendulum Aglaomorpha cornucopia
OPHIOGLOSSACEAE OPHI POLYPODIACEAE
‘*& ” ches ” WP. cd ;
as = hare 2 3
Agathis philippinensis Podocarpus costalis Podocarpus macrocarpus
ARAUCARIACEAE PODOCARPACEAE PODOCARPACEAE
Rhododendron kochii Rhododendron javanicum Cinnamomum mercadoi
ERICACEAE ERICACEAE LAURACEAE
Fig. 5. Some threatened plants of Mt. Kitanglad, Bukidnon, Mindanao. Photos by V. Amoroso.
Recommended policies for the PAMB include: (1) The LGU’s of Intavas
and other LGUs around the park should officially organise their porters/guides for
long-term monitoring of threatened and endemic species. (2) Ex-situ and in-situ
conservation of species and habitats should be carried out to protect the remaining
endemic, threatened, rare, and economically important species of plants. For ex-
Vegetation and plant diversity in Mt. Kitanglad Range Natural Park 235
situ conservation, each municipality should have an economic garden / nursery to
propagate their threatened, endemic and economically important plants. (3) Collection
of threatened and endemic species should be regulated. (4) Denuded mountains should
be planted with indigenous tree species. (5) Threatened and endemic species found
in agricultural areas should be protected. (6) Mountaineers / hikers should be given
proper orientation before trekking and should follow forest guides in their trekking.
Temporary campsites at the middle zone (about 2000 m asl) should be discouraged
because of the presence of threatened vascular plants.
ACKNOWLEDGEMENTS. The authors gratefully acknowledge the financial support from
Central Mindanao University, Musuan, Bukidnon, Philippines. We appreciate our international
collaborators from the Taiwan Forestry Research Institute (TFRI) and the University of Zurich.
We are also thankful to previous research assistants during this study: A.J. Veloso, R.N.
Dopenio, and various graduate and undergraduate students of the first author.
References
Amoroso, V.B. (1993) Valuable Ferns in Mindanao. CMU J. Sci. 6(2): 23-36.
Amoroso, V.B. (1997) Ferns of the Philippines. Paper presented during a workshop
seminar. Quezon City: University of the Philippines, Diliman, Philippines.
Amoroso, V.B., Acma, F.M. & Pava, H.P. (1996) Diversity status and ecology of
pteridophytes in three forests in Mindanao. In: Canus, J.M. & Johns, R.J. (eds)
Pteridology in Perspective, pp. 53-60. Kew: Royal Botanical Gardens, Kew.
Amoroso, V.B., Arreza, G.R., Bornales, R.G., Cambel, T.L., Canencia, O.P., Estoista,
R.B., Lariosa, E.A., Mero, D.C. & Roxas, A.T. (2004) Participatory rural
appraisal in the lowland ecosystem of Mt. Malindang, Misamis Occidental,
Philippines. College Laguna: SEAMCO SEARCA and Quezon City: PDM Press,
Philippines.
Amoroso, V.B., Aspiras, R.A. & Polizon, J.J. (2009) Participatory inventory and
distribution of endangered, endemic and economically important plants in
Hamiguitan Range Wildlife Sanctuary, Davao Oriental. B/umea 54: 71—76.
Amoroso, V.B., Ruba, R.P., Demetillo, M.T., Stortelder, A., Lagunzad, D.A., Aranico,
E.C., Arances, J.B., Rufila, L.V., Kessler, P.J.A., Tan, B.C., Roscom, B.A., Gorne,
N.D., Opiso, G.S., Van Den Berg, A., Gruezo, W.Sm., Co L.L., & Alava, C.G.
(2006) Plant diversity in the northern landscape of Mt. Malindang Range and
environs, Misamis Occidental, Philippines. College Laguna: SEAMCO SEARCA
and Quezon City: PDM Press, Philippines.
Barcelona, J.F., Hernandez, B.F. & Price, M.G. (1996) Philippine Schizaea. Asia Life
Sci. 5: 27-34.
Brower, J.E. & Zar, J.H. (1977) Field and Laboratory Methods for General Ecology.
Dubuque, Iowa: C. Brown Company Publishers.
Cootes, J. (2001) The Orchids of the Philippines. Singapore: Times Editions.
236 Gard. Bull. Singapore 63(1 & 2) 2011
Department of Environment and Natural Resources (DENR) Administrative Order No.
2007-01. Establishing the National List of Threstened Plants and their Categories
and the List of other Wildlife Species. Diliman Quezon City: Department of
Environment and Natural Resources.
Editorial Committee of the Flora of Taiwan (1996) Flora of Taiwan Vol. 2,
Angiospermae, 2nd edn. Taipei: National Science Council of the Republic of
China.
Fernando, E.S., Co, L.L., Lagunzad, D.A., Gruezo, W.Sm., Barcelona, J.F., Madulid,
D.A., Lapis, A.B., Texon, G.I., Manila, A.C., & Zamora, P.M. (2008) Threatened
Plants of the Philippines: A Preliminary Assessment. Asia Life Sci. Suppl. 3: 1-52.
Heaney, L., Tabaranza, B.R., Rickart, E.A., Balete, D.S. & Ingle, N.R. (2006) The
mammals of Mt. Kitanglad Nature Park, Mindanao, Philippines. Fie/diana, n.s.
112: 1-63.
Jebb, M. & Cheek, M. (2001) Nepenthaceae. Flora Malesiana Ser. 1, 13: 145-168.
Kalkman, C. et al. (1995-1996). Flora Malesiana Series I, volume 12. Leiden:
Riyksherbarium / Hortus Botanicus, Leiden University.
Krebs, J.R. (1994) Introduction to Plant Biology. Dubuque, lowa: C. Brown Company
Publishers.
Linder, G.R. (1987) Monograph of the fern genus Goniophlebium (Polypodiaceae).
Blumea 34: 283-328.
Liu, Yea-Chen, Fraser-Jenkins, C.R., Amoroso, V.B. & Chiou, W.L. (2008) Athyrium
erythropodum (Woodsiaceae Pteridophyta), a new Philippine Record. Blumea 53:
447-457.
Madulid, D. (1991) The endemic genera of flowering plants in the Philippines. Acta
Manillana 39: 47-S8.
Madulid, D. (1995) A Pictorial Cyclopedia of Philippine Ornamental Plants.
Philippines: Bookmark, Inc.
Merrill, E.D. (1923-1926) An Enumeration of Philippine Flowering Plants. Volumes
I-IV. Manila: Bureau of Printing, Philippines.
NORDECO (1998) Nordic Agency for Development and Ecology (NORDECO) and
Department of Environment and Natural Resources (DENR) Technical Report.
Integrating conservation and development in protected area management in Mount
Kitanglad Range Range Natural Park, Philippines. Copenhagen: NORDECO and
Manila: DENR.
Perez, H. (2004) Diversity of trees along altitudinal gradient: Layawan River to
North Peak in Mt. Malindang National Park, Baranggay Lake Duminagat, Don
Victoriano, Misamis Occidental. Samu t-Sari 3(2).
Rojo, J.P. (1999) Revised Lexicon of Philippine Trees. College, Laguna: Forest Products
Research and Development Institute, Department of Science and Technology,
Philippines.
Tan, B.C., Fernando, E.S. & Rojo, J.P. (1996) An Updated list of Endangered Philippine
Plants. Yushania 3(2): 1—S.
Zamora, P.M. & Co, L.L. (1986) Guide to Philippine Flora and Fauna. Vol. IV.
Quezon City: Goodwill.
Gardens’ Bulletin Singapore 63(1 & 2): 237-243. 2011 237
Phylogenetic study of the Hottarum Group
(Araceae: Schismatoglottideae)
utilising the nuclear ITS region
S.L. Low'*, S.Y. Wong’, J. Jamliah' and P.C. Boyce*
‘Department of Plant Science & Environmental Ecology.
Faculty of Resource Science & Technology, Universiti Malaysia Sarawak
94300 Kota Samarahan, Sarawak, Malaysia
*shooklingS052(@hotmail.com (corresponding author)
*School of Biological Sciences, Universiti Sains Malaysia,
11800 Pulau Pinang, Malaysia
ABSTRACT. Recent phylogenetic analyses of the tribe Schismatoglottideae (Araceae)
elucidated a well-supported but internally unresolved crown group comprising Schismatoglottis
sarikeensis (Bogner & M.Hotta) A.-Hay & Bogner, previously placed in the genus Hottarum
Bogner & Nicolson, the genus Phymatarum M.Hotta, and a number of species either novel
or hitherto placed in Schismatoglottis Zoll. & Moritzi. The clade is particularly interesting in
that it is centred in northern central Sarawak (Malaysian Borneo), north of the Lupar Divide
and appears to represent an autochthonous radiation point for evolutionary activity isolated
from the major tribal radiations in south-western Sarawak. Former Hottarum species (with
the exclusion of H. truncatum) transferred to Piptospatha and Schismatoglottis are misplaced.
All except Bakoa lucens (Bogner) P.C.Boyce & S.Y.Wong belong to this supra-Lupar Divide
grouping. This study was undertaken to test the validity and phylogeny of the genus Hottarum
utilising the nuclear ITS region.
Keywords. Hottarum, ITS region, Lupar Divide, phylogeny, Phymatarum, Schismatoglottis
Introduction
The genus Hottarum Bogner & Nicolson previously comprised of four rheophytic
species, all endemic to Borneo (Mayo et al. 1997). These include the type species,
Hottarum truncatum (Bogner 1978), from Sg. (river) Kakus, Tatau, Bintulu, Sarawak,
and a further three species: H. /ucens (Bogner 1983), H. sarikeense (Bogner & Hotta
1983) and H. kinabaluense (Bogner 1984). Hottarum brevipedunculatum (Okada &
Mori 2000) was subsequently described.
In dismantling Hottarum, Bogner & Hay (2000) placed the constituent taxa
of Hottarum (i.e., including the type) into Piptospatha, and H. sarikeense was placed
into Schismatoglottis, based purely on morphological characters. Piptospatha sensu
Bogner & Hay (2000) differs from Schismatoglottis by its unconstricted spathe and
in having seeds with an extended micropylar appendage. Piptospatha sensu Wong &
Boyce (2010b) is further defined from Schismatoglottis by the peduncle erect at fruit
dispersal, with the persistent lower fruiting spathe forming a funnel-form splash cup,
238 Gard. Bull. Singapore 63(1 & 2) 2011
and pistils connate into a syncarpium (or rarely free but coherent).
Phylogenetic analyses of the tribe Schismatoglottideae (Wong et. al.
2010), with a well-supported molecular profile and compelling morphological
peculiarities, supported the removal of H. lucens to a new genus, Bakoa (= Bakoa
lucens; Wong & Boyce 2010a) and H. kinabaluense to the generically novel Ooia
(= Ooia kinabaluensis; Wong & Boyce 2010b). Recently, another former Hottarum
species, H. brevipedunculatum was shown to represent a second species of Bakoa
(B. brevipedunculata; Wong 2011). These transfers leave a core of species, including
the nomenclatural type (Hottarum truncatum), Hottarum sarikeense, Schismatoglottis
Josefii A.Hay, and three undescribed species in a Phymatarum + Hottarum clade.
Materials and methods
Sampling
Twenty taxa were selected. Nine taxa formed the ingroup, including the type of
Hottarum, H. truncatum; H. sarikeense (two accessions), Schismatoglottis josefii,
and its putative sister taxa, Phymatarum borneense (two accessions), and three
unplaced taxa (Schismatoglottis sp. A|AR-114], S. sp. B [AR-135] and S. ‘petradoxa’
[AR-920]). Eleven outgroup taxa were selected based on the results from Wong
et al. (2010), and comprise Schismatoglottis (2 species), Aridarum (7 species) and
Bucephalandra (2 accessions). Appendix A lists all the taxa and its respective localities,
together with voucher information and GenBank accession numbers. Vouchers are
deposited with the Herbarium of the Sarawak Forestry Department (SAR).
DNA extraction, PCR and sequencing
Total DNA was extracted using a modified version of the 2X CTAB protocol (Doyle
& Doyle 1987) with the addition of PVP (PolyVinylPyrrolidone) as described by
Gauthier et al. (2008). ITS1 (Internal Transcribed Spacer 1) and ITS 2 were amplified
using the primer pairs 1F/1R and 3F/4R, respectively (White et al. 1990). Polymerase
chain reactions (PCRs) were conducted in a total reaction volume of 20 ul comprising
1X buffer, 0.1mM dNTP mix, 0.2mM of each primer, 2.0mM MgCl, 2 units Taq DNA
polymerase and 2 ul of DNA extract). | ul of DMSO was added to each reaction for
improved amplification.
PCR conditions included an initial 2-min denaturation at 95°C, 40 cycles
of l-min at 95°C (denaturation), 1-min at 50°C-60'C (annealing), and 2-min at 72°C
(extension), followed by a final 10-min extension at 72°C. PCR products were
visualised on 1.5% or 2.0% agarose gels. Desired products were purified and sent for
sequencing.
Sequence alignment and phylogenetic analyses
All sequences obtained were manually checked, edited, assembled and aligned using
the BioEdit version 7.0.5 (Hall 1999). Gaps were treated as insertions or deletions of
nucleotides (indels).
The Hottarum Group analysed with the ITS region 239
Maximum parsimony (MP) analyses were performed using PAUP’ v.4.0b10
(Swofford 2002) according to the parameters described by Wong et al. (2010), except
that 100,000 trees were saved at the second round of tree bisection-reconnection (TBR)
branch swapping. Tree topologies were interpreted with bootstrap values generated
from RAxML (Randomized Axelerated Maximum Likelihood: Stamatakis et al. 2008)
for 100 replicates and repeated 10 times to generate 1000 replicates. Bootstrap support
values were taken as weak (50—74%), moderate (75-84%) or strong (85—100%) as
applied by Richardson et al. (2000).
Results and discussion
Analysis of the ITS region
Total aligned nucleotides for the ITS region for 20 taxa comprise 879 bp. The sequence
length varies from 770 bp (Aridarum crassum, AR-1605) to 803 bp (Phymatarum
borneense, AR-1931). A large deletion of indels (46 bp) was found in the type, Hottarum
truncatum (AR-3080) at the position 159 to 204. Indels were found less beyond the
position 500 bp for all 20 taxa. The ITS region was rich with the GC nucleotide.
All characters are of the type ‘unord’ and have equal weight. 771 characters from
the entire sequences are constant and 58 variables characters (6.6%) are parsimony-
uninformative. The remaining 50 characters (5.7%) were parsimony-informative,
resulting in 29 most parsimonious trees with a tree length of 136 steps, consistency
index (CI) of 0.85 and a retention index (RI) of 0.82. These trees also generated a
rescaled consistency index (RC) of 0.69 and homoplasy index (HI) of 0.15. CI and HI
with the exclusion of uninformative characters were 0.72 and 0.28, respectively.
The tree topology of the maximum parsimony (MP) 50% majority rule
(not shown) differed in the Bucephalandra + Hottarum truncatum (AR-3080) and
the Aridarum clades formed, as compared with the maximum likelihood (ML) tree
(Fig. 1) from RAxML. The bootstrap values (BS) for the MP tree were generated and
stopped at 411 replicates due to computing limitations. Both MP and ML trees strongly
support the previous phylogenetic study by Wong et al. (2010). Schismatoglottis josefii
together with H. sarikeense formed a monophyletic clade (Clade A in Fig. 1) with
parsimony bootstrap value, BS, and likelihood bootstrap value, BS,,, of 100%. Clade
A is weakly associated with Clade B (37% BS,,, ). Clade B itself is well supported
(85%) and comprises the three unplaced taxa (Schismatoglottis sp. A, Schismatoglottis
sp. B and Schismatoglottis ‘petradoxa’). The Schismatoglottis sp. B groups with
Schismatoglottis ‘petradoxa’ with weak BS,, and BS,,, of 60% and 72%, respectively.
Phymatarum borneense (Clade D) is recovered as a monophyletic genus (BS,,
BS,,,=100%). This genus is sister to the rest of the taxa including clusters containing
Hottarum truncatum (type species of Hottarum) + Aridarum (Clade C). Hottarum
truncatum (AR-3080) received low likelihood bootstrap support for its association
with the Aridarum species (BS,,, =55%) in Clade C. As this is considered insignificant,
the Hottarum truncatum + Aridarum clusters are taken as remaining unresolved.
240 Gard. Bull. Singapore 63(1 & 2) 2011
[ Schismatoglottis ‘macrocardia‘(AR-607)
Schismatogloitis adoceta (AR- 1408)
Phymatarum borneense (AR-1931) a
Phymatarum borneense (AR-1442)
Hottarum truncatum (AR-3080)
Anidarum crassum (AR-263) )
Andarum borneense (AR-2123) |
Anidarum nicolsonii (AR-480) |
Andarum pursegiovei! (AR-1887)
Andarum caulescens (AR-2311)
Aridarum purseglovei(AR-3096)
= Anidarum caulescens (AR-588) fe
Tae Schismatogliottis’ sp.A (AR-114) )
Schismatoglottis'sp.B (AR-135) | B
Schismatoglottis ‘petradoxa’ (AR- 920) ,
Schismatoglottis josefil (AR-1157)
_ Hottarum sankeense (AR-2394) | A
Hottarum sanikeense (AR-1605) J
| f—7— Bucephalandra motleyana(AR-252)
Bucephalandra motleyana (AR-2537)
Fig. 1. The maximum likelihood tree obtained from RAXxML with ITS sequences. Numbers
next to branches are values for BS,,/BS,,,. BS,, = bootstrap value for maximum parsimony.
BS,,, = bootstrap value for maximum likelihood.
Conclusion
Hottarum truncatum (=Piptospatha truncata, the type species for Hottarum), is shown
to be separated from H. sarikeense in this study. The rest of the species formerly
placed in Hottarum have been shown to be misplaced as well. The former Hottarum
species: H. lucens (= Piptospatha lucens) and H. brevipedunculatum (= Piptospatha
brevipedunculata) were transferred to a new genus, Bakoa P.C.Boyce & S.Y.Wong
(Boyce & Wong 2008, Wong 2011) and Hottarum kinabaluense (= P. kinabaluensis)
was transferred to a novel genus, Ooia (Wong & Boyce 2010b).
Although H. truncatum is sister to Aridarum, the weak likelihood bootstrap
values do not support the placement of Hottarum truncatum within Aridarum. A group
comprising Schismatoglottis josefii and H. sarikeense, and another group of 3 novel
species from central Sarawak, are well supported clades (with 100% and 85% support,
respectively). However, their generic assignment awaits the results of analyses with
additional gene regions.
ACKNOWLEDGEMENTS. This study is a part of a larger research project forming the
core of the first author’s M.Sc. research in UNIMAS. This study is partially funded by
the Ministry of Higher Education, Malaysia Fundamental Research Grant Scheme No.
FRGS/01(12)/709/2009(25) under Sarawak Forestry Department Research Permit No.
NPW.907.4.(1V)-133 and Park Permit No. 99/2009. The support from the Sarawak Forestry
Department is gratefully acknowledged. The first author is grateful to the Singapore Botanic
Gardens for the award of a bursary that enabled her attendance at the 8th Flora Malesiana
Symposium, where this paper was presented.
The Hottarum Group analysed with the ITS region 241
References
Bogner, J. (1983) A new Hottarum (Araceae) species from Borneo. Pi. Syst. Evol. 142:
49-55.
Bogner, J. (1984) Two new aroids from Borneo. P/. Syst. Evol. 145: 159-164.
Bogner, J. & Hay, A. (2000) Schismatoglottideae (Araceae) in Malesia 2: Aridarum,
Bucephalandra, Phymatarum and Piptospatha. Telopea 9(1): 179-222.
Bogner, J. & Hotta, M. (1983) A new Hottarum (Araceae) from Sarawak. Bull. Mus.
Nat. His. Nat. (Paris) 4(5): 27-31.
Boyce, P.C. & Wong, S.Y. (2008) Studies on Schismatoglottideae (Araceae) of Borneo
VII: Schottarum and Bakoa, two new genera from Sarawak, Malaysian Borneo.
Bot. Stud. 49: 393-404.
Doyle, J.J. & Doyle, J.L. (1987) A rapid DNA isolation procedure for small quantities
of fresh leaf tissue. Phytochem. Bull. 19: 11-15.
Gauthier, M-P.L., Barabé, D. & Bruneau, A. (2008) Molecular phylogeny of the genus
Philodendron (Araceae): Delimitation and infrageneric classification. Bot. J.
Linn. Soc. 156: 13-27.
Hall, T.A. (1999) BioEdit: a user-friendly biological sequence alignment editor and
analysis program for Windows 95/98/NT. Nucleic Acids Symp. Ser. 41: 95-98.
Hay, A. & Yuzammi (2000) Schismatoglottideae (Araceae) in Malesia I —
Schismatoglottis. Telopea 9(1): 1-177.
Mayo, S.J., Bogner, J. & Boyce, P.C. (1997) The Genera of Araceae, p. 370. Kew:
Royal Botanic Gardens, Kew.
Okada, H. & Mori, Y. (2000) Three new species of Schismatoglottideae, Araceae,
from Borneo. Acta Phytotax. Geobot. 51(1): 1-9.
Richardson, J.E., Fay, M.F., Cronk, Q.C.B., Bowman, D. & Chase, M.W. (2000)
A phylogenetic analysis of Rhamnaceae using rbcL and trnL-F plastid DNA
sequences. Amer. J. Bot. 87(9): 1309-1324.
Stamatakis, A., Hoover, P. & Rougemont, J. (2008) A rapid bootstrap algorithm for the
RAXxML Web-serves. Syst. Biol. 75(5): 758-771.
Swofford, D.L. (2002) PAUP*: Phylogenetic Analysis Using Parsimony (* And Other
Methods), version 4.0 beta 10. Massachusetts, Sunderland: Sinauer Associates.
White, T.J., Bruns, T., Lee, S. & Taylor, J. (1990) Amplification and direct sequencing
of fungal ribosomal RNA genes for phylogenetics. In: Innia, M., Gelfand, D.H..,
Sninsky, J.J. & White, T.J. (eds) PCR Protocols, pp 315-322. California, San
Diego: Academic Press.
Wong, S.Y. (2011) Studies on Schismatoglottideae (Araceae) of Borneo XV: A second
species of Bakoa from Indonesian Borneo. Acta Phytotax. Geobot. 61(3):127—
| Ped
Wong, S.Y. & Boyce, P.C. (2010a) Studies on Schismatoglottideae (Araceae) of Borneo
IX: A new genus, Hestia, and resurrection of Apoballis. Bot. Stud. 51: 249-255.
Wong, S.Y. & Boyce, P.C. (2010b) Studies on Schismatoglottideae (Araceae) of
Borneo XI: Ooia, a new genus and a new generic delimitation of Piptospatha.
Bot. Stud. 51: 543-552.
242 Gard. Bull. Singapore 63(1 & 2) 2011
Wong, S.Y., Boyce, P.C., Othman, A.S. & Leaw, C.P. (2010) Molecular phylogeny of
tribe Schimatoglottidae based on two plastid markers and recognition of a new
tribe, Philonotieae, from the neotropics. Jaxon 59(1): 117-124.
Appendix A. Taxa included in the study: GenBank accession numbers for sequences used,
voucher specimen numbers and collection localities. ‘Schismatoglottis’ = unplaced and
undescribed taxa; ‘macrocardia’ and ‘petradoxa’= nomenclatural taxa yet to be described; taxa
are arranged alphabetically following the taxonomic position of Mayo et al. (1997), Bogner &
Hay (2000), Hay & Yuzammi (2000) and Wong et al. (2010).
GenBank
Voucher : ;
Species accession Collection locality / coordinates
number
no.
Aridarum borneense JN544438 AR-2123 Sungai Bungen, Kubah National
Park, Matang, Kuching, Sarawak,
01°3630.9'N 110°11 35.0°E
Aridarum caulescens JN544440 AR-2311 Melinau Gorge, Mulu National Park,
Nanga Medamit, Limbang, Sarawak,
GPS position not available.
Aridarum caulescens JN544428 AR-588 Bukit Satiam, Bintulu, Sarawak,
02°58'47.6 N 112°56 37.5 E
Aridarum crassum JN544426 AR-263 Gunung Gaharu, Pantu, Sri Aman,
Sarawak, 01°01'19.5'N 110°52’'52.8°E
Aridarum nicolsonii JN544427 AR-480 Trail above Camp Permai, Santubong,
Kuching, Sarawak, 01°45 49.0'N
110°1907.4E
Aridarum purseglovei JN544435 AR-1887 Bukit Satiam, Bintulu, Sarawak,
02259 133) NUI2255) 51-5.
Aridarum purseglovei JN544444 AR-3096 Sg. Likau, GT Plantations, Tatau,
Bintulu, Sarawak, 02°43 53.6 'N
IL S27Sy US)sIL 1
Bucephalandra motleyana JN544425 AR-252 Gunung Gaharu, Pantu, Sri Aman,
Sarawak, 01°0239.5'N 110°53'18.3°E
Bucephalandra motleyana JN544442 AR-2537 Bukit Kelam, Sintang, Kalimantan
Barat, Indonesia, 00°05 30.1°'N
111°39'03.3'E
Hottarum sarikeense JN177489 AR-2394 — Sg. Pedali, Lubok Antu, Sri Aman,
Sarawak, 01°1158.9°N 112°03'27.0E
Hottarum sarikeense JN544434 AR-1605 Sg. Lepong, Sarikei, Sarawak,
01°5712.9'N 111°3034.9°E
Hottarum truncatum JN544443 =AR-3080 — Sg. Pandan Kecil, Trail behind Camp
C, GT Plantations, Tatau, Bintulu,
Sarawak, 02°42'40.1°N 113°2037.9"E
The Hottarum Group analysed with the ITS region
Phymatarum borneense
Phymatarum borneense
Schismatoglottis adoceta
Schismatoglottis josefii
Schismatoglottis ‘macrocardia’
Schismatoglottis ‘petradoxa’
‘Schimatoglottis sp. A
‘Schismatoglottis’ sp. B
JN544433
JN544436
JN544432
JN544431
JN544429
JN544430
JN544423
JN544424
AR-1442
AR-1931
AR-
1408
AR-1157
AR-607
AR-920
AR-114
AR-135
243
Belaga Road, Sebauh, Bintulu, Sarawak,
03°03°34.3°N 113°4216.4E
Trail to Deer Cave, Mulu National Park,
Mulu, Miri, Sarawak, 04°02 23.8°N
114°48'54.6E
Road junction, km 10 Bakun- Bintulu-
Miri, Kapit, Sarawak, 02°50'51.7'N
114°0157.6E
Sg. Pedali, Nanga Sumpa, Batang
Ai, Lubok Antu, Sri Aman, Sarawak,
01°1158.9'N 112°03 27.0°'E
Bukit Satiam, Bintulu, Sarawak,
02°59'26.1°N 112°55'54.4°E
Km 65 road to Camp Gahada, Rejang
Wood Concession, Nanga Gaat, Kapit,
Sarawak, 01°41°59.7'N 113°31 13.7E
Sg. Piat, Nanga Gaat, Kapit, Sarawak,
01°38'09.1'N 113°2409.9°E
Stream below Camp Gahada, Nanga
Gaat, Kapit, Sarawak, 01°41 49.4°N
113°2616.3°E
Gardens’ Bulletin Singapore 63(1 & 2): 243-257. 2011 245
Dendrobium (Orchidaceae): To split or not to split?
André Schuiteman
Herbarium, Library, Art and Archives.
Royal Botanic Gardens, Kew. Richmond. Surrey TW9 3AE. U_K.
A_Schuiteman@kew_ore
ABSTRACT. Dendrobium Sw. is one of the three largest orchid genera. with around 1580
species if certain currently accepted satellite genera are included. Until recently. no serious
attempts have been made to split up this imporiant genus into smaller genera. An infrageneric
classification at the sectional level, largely due to Schlechter, has been accepted by most
workers. Recent analyses based on DNA markers by Yukawa. Clements. and others have
provided new insights into the phylogeny of Dendrobium. Their work shows that Dendrobium
is not monophyletic when the satellite genera are excluded. This led to proposals to split up
Dendrobium into as many as fifty genera. largely along the lines of Schlechter’s sections.
However, the data do not suggest any single, evident way to do the splitting. Here it is argued
that a broad concept of the genus Dendrobium, which includes genera like Cadetia Gaudich..
Flickingeria A.D Hawkes and Epigeneium Gagnep.. among others, is to be preferred. The
comparable cases of other large orchid genera are briefly discussed and some observations are
made on character evolution in Dendrobium and the origin of the genus in light of DNA-based
phylogenies.
Keywords. Classification. Dendrobium, generic concepts, molecular phylogeny, Orchidaceae
Introduction
In its traditional, broad delimitation, Dendrobium Sw. is one of the three largest orchid
genera (Cribb & Govaerts 2005: 1197 spp.) as well as one of the most important in
commercial horticulture. Certain species of this genus are in high demand in traditional
Chinese medicine, which puts wild populations of these species under pressure from
(illegal) collecting in China and neighbouring countries. Habitat destruction threatens
the survival of many more species throughout the range of the genus. The taxonomy
of Dendrobium is therefore of interest not only to botanists, but also to orchid growers.
ecologists and conservationists.
Dendrobium belongs to subtribe Dendrobiinae in the tribe Dendrobieae of the
subfamily Epidendroideae, the largest of the five subfamilies within the Orchidaceae.
Most species of Dendrobium are epiphytes in primary forest, less often lithophytes:
only very few are obligate terrestrials. The range of the genus extends from Sri Lanka
and India throughout tropical Asia, north to Japan, east to Tahiti. and south to New
Zealand.
Prior to the advent of molecular methods, various classifications had been
proposed for Dendrobium and related genera, as summarised in Wood (2006). Dressler
246 Gard. Bull. Singapore 63(1 & 2) 2011
(1981, 1993) expressed a consensus view when he listed six genera as constituting
the subtribe Dendrobiinae: Cadetia Gaudich., Dendrobium, Diplocaulobium (Rchb.f.)
Kraenzl., Epigeneium Gagnep., Flickingeria A.D.Hawkes (Ephemerantha P.F.Hunt
& Summerh.), and Pseuderia Schltr. In this view, the Dendrobiinae consist of the
very large and complex genus Dendrobium next to a number of much smaller, more
homogenous genera, which had all been treated as sections of Dendrobium in the
past. Rudolf Schlechter (1911—1914) must be credited with presenting an infrageneric
classification of Dendrobium that has been adopted with few modifications by most
subsequent workers. This, however, applies only to the section level. Schlechter’s
system of subgenera is almost entirely artificial, being based on the application of
single defining character states, such as the presence or absence of a sheathing leaf
base. As we now know, this does not lead to phylogenetically meaningful groupings in
this subtribe.
The pioneering studies by Yukawa and co-workers (1993, 1996, 2000, 2001)
using DNA markers (matK and ITS) have provided a number of insights, which were
confirmed and extended by later studies (Clements 2003, 2006; Wongsawad et al.
2005; Burke et al. 2008; Sathapattayanon 2008):
1. Dendrobiinae consists of three main clades:
I. A predominantly continental Asian and West Malesian clade that includes
the type species of Dendrobium (D. moniliforme (L.) Sw.).
I]. A predominantly East Malesian - Australian - New Caledonian clade that
includes, e.g., D. bigibbum Lindl., as well as Cadetia, Diplocaulobium
and Flickingeria.
III. A much smaller clade that consists of the genus Epigeneium.
Following Clements (2003), these clades will here be referred to as the Asian,
Australasian and Epigeneium clade respectively.
2. Under the consensus classification, Dendrobium is polyphyletic.
Yukawa et al. (1993) found that the genus Pseuderia is not a member of the
Dendrobiinae, but appears to belong to the tribe Podochileae. In view of the
deviating vegetative and floral morphology of Pseuderia, both within Dendrobiinae
and Podochileae, its placement within the subfamily Epidendroideae needs further
study. In addition, Clements (2003) showed that a group of species traditionally
treated as section Oxystophyllum of Dendrobium properly belongs in the subtribe
Eriinae.
The taxonomic implications of these molecular studies found their expression
in two highly divergent views. Some authors presented arguments in favour of a very
large, monolithic and monophyletic genus Dendrobium, essentially comprising the
whole subtribe Dendrobiinae, except perhaps the genus Epigeneium (Yukawa 2001,
Burke et al. 2008). On the other hand, Clements & Jones (2002), in line with an earlier
suggestion by Yukawa et al. (1993), proposed that Dendrobium should be split up
in several smaller genera. Clements (2006) recognises as many as 50 genera in this
alliance. According to Clements (2003), the three main clades should be treated as
distinct subtribes (1: Dendrobiinae, II: Grastidiinae, and III: Epigenetinae). Wood
(2006), while provisionally following Dressler’s consensus view (except for Pseuderia
A broad generic concept of Dendrobium
and Oxystophyllum), expressed the hope that a middle ground between extreme
lumping and extreme splitting could be found.
Material and methods
The phylogram here shown (Fig. 1, 2) is based on sequences of the Internal
Transcribed Spacer | (partial), 5.8S ribosomal RNA gene (complete), and Internal
Transcribed Spacer 2 (partial) downloaded from GenBank (http://www.ncbi.nlm.nih.
gov). See Table | for taxa and accession numbers. Taxa were selected to make the
analysis comparable to earlier studies by Clements (2003, 2006) and Yukawa (2001).
Where possible, different but morphologically similar species were chosen. When a
taxon was represented by multiple accessions, initially all accessions were included
I
=a
52
Ila
71 D. cunninghamii I Winika
D. fractiflecum | Macrocladium
D. maidenianum I Cadetia
44 48 D. macrophyllum
Latouria
II 6 D. bifalce
D. canaliculatum
3 Spatulata
IIb a 76 D. nindii P
D. furcatopedicellatum
Grastidium
37 D. baileyi
29
29 15 D. ischnopetalum I Diplocaulobium
D. comatum I Flickingeria
aa D. monophyllum | Monophyllaea
D. racemosum | Rhizobium
44
D. kingianum var. pulcherrimum
85 24
D. adae
5 Dendrocoryne
D. speciosum
50 D. tetragonum
D. cymbidioides
Epigeneium
III 31 D. nakaharae
Bulbophyllum lobbii |
outgroup 1
100 Bulbophyllum nutans 9 P
Liparis kramer
outgroup 2
Eria ferruginea
100 Oxystophyllum sinuatum I outgroup 3
Fig. 1. One of 20 most parsimonious phylograms of selected Dendrobium species based on
ITS sequence data. Names on the right refer to the traditional sections and genera in which the
species would be included. Clade I is shown in Fig. 2.
248
35
57
21
33
36
52
44
Fig. 2. Phylogram of Clade I (from Fig. 1).
D. senile
D. ellipsophyllum
D. cariniferum
Gard. Bull. Singapore 63(1 & 2) 2011
Dendrobium ?
Distichophyllae
94 D. cuthbertsonii J Oxyglossum
93 D. bracteosum ff Pedilonum
17 D. lawesii J Calyptrochilus
36 D. smillieae — = Pedilonum
P2 34 D. cyanocentrum I Oxyglossum
93 D. lancifolium
D. victoriae-reginae Calcarifera
96 D. chameleon
36 79 D. terminale
38 D. aloifolium |
Aporum
D. spatella
ee D. quadrangulare I = Bolbodium
D. equitans
30 D. crumenatum Crumenata
D 40 D. goldfinchii
D. densifiorum I Densiflora
100 D. monticola
D. minutiflorum Ssaeyess
30 D. secundum
P1 D. amethystoglossum | ei
99 D. platygastrium I Platycaulon
35 D. mutabile [= Calcarifera
D. trigonopus | + Formosae ?
4
I
I
I
D. christyanum Fomgsae
D. lindleyi
100 D. jenkinsii Densiflora
D. chrysotoxum
D. fimbnatum
D. hancockii
D. devonianum Dendrobium
D. nobile
100 D. officinale
58 D. hercoglossum Breviflores
D. capillipes Dendrobium
in the alignment. In all cases these accessions gave identical results during phylogeny
inference, although there often were minor differences between the sequences, for
example in the three accessions of D. chrysotoxum Lindl. Afterwards one accession
was chosen arbitrarily to represent the taxon. Unfortunately, Yukawa’s and part of
Clements’s sequence data were not yet uploaded to GenBank at the time of this study.
A number of motif-based tests for the detection of pseudogenes were
performed (Harpke & Peterson 2008, Feliner & Rosselld 2007). Sequences were
aligned with Mega version 4 (Tamura et al. 2007), using the ClustalW algorithm, and
adjusted manually. The phylogeny of the 61 selected taxa was inferred with Mega4,
using maximum parsimony with the following options: close-neighbour interchange,
A broad generic concept of Dendrobium 249
Table 1. List of GenBank accession numbers.
Bulbophyllum lobbii Lindl. — EF195931; B. nutans Thouars — EF196038; Dendrobium adae
F.M.Bailey — EU430371; D. aloifolium (Blume) Rchb.f. — AY239951; D. amethystoglossum
Rchb.f. — AY239952; D. baileyi F.Muell. — AY240016; D. bifalce Lindl. — EU430373;
D. bracteosum Rchb.f. — AY239954; D. canaliculatum R.Br. — EU430375; D. capillipes
Rchb.f. — AF362035; D. cariniferum Rchb.f. — AF362027; D. chameleon Ames — AF521607;
D. christyanum Rchb.f. — EF629325; D. chrysotoxum Lindl. — EU477501; D. comatum
(Blume) Lindl. — AB289469; D. crumenatum Sw. — AF521608; D. cunninghamii Lindl. —
AY 240019; D. cuthbertsonii F.Muell. — AY239950; D. cyanocentrum Schltr. — AY239964;
D. cymbidioides (Blume) Lindl. — AY240011; D. densiflorum Wall. ex Lindl. — DQ058786;
D. devonianum Paxton — FJ384735; D. ellipsophyllum Tang & F.T.Wang — AF362033; D.
equitans Kraenzl. — AF521609; D. fimbriatum Hook. — EU003116; D. fractiflexum Finet —
AY 239949; D. furcatopedicellatum Hayata — AF521611; D. goldfinchii F.Muell. — AY239969;
D. hancockii Rolfe — EU003120; D. hercoglossum Rchb.f. — AF363685; D. ischnopetalum
Schltr. —- AY 240007; D. jenkinsii Wall. ex Lindl. - DQ058785; D. kingianum var. pulcherimum
Rupp — EU430385; D. lancifolium A.Rich. — AY239976; D. lawesii F.Muell. — AY239977;
D. lindleyi Steud. — DQ058784; D. macrophyllum A.Rich. — AY239979; D. maidenianum
Schltr. — AY239948; D. minutiflorum S.C.Chen & Z.H.Tsi (= D. sinominutoflorum S.C.Chen,
J.J.Wood & H.P.Wood) — DQ058800; D. monophyllum F.Muell. — EU430387; D. monticola
P.F.Hunt & Summerh. — DQ058798; D. mutabile (Blume) Lindl. — AY239984: D. nakaharae
Schltr. —AF521618; D. nindii W.Hill — AY239985; D. nobile Lindl. — EU477507; D. officinale
Kimura & Migo (= D. catenatum Lindl.) — EU592018; D. playgastrium Rchb.f. — AY239955;
D. quadrangulare Parish & Rchb.f. (= D. hymenanthum Rchb.f.) — EU840698; D. racemosum
(Nicholls) Clemesha & Dockrill — EU430389; D. secundum (Blume) Lindl. — AY239993; D.
senile Parish ex Rchb.f. - EU477509; D. smillieae F.Muell. — AY239996; D. spatella Rchb.f.
— AF362034; D. speciosum Sm. — EU430399; D. terminale Parish & Rchb.f. — DQ058801; D.
tetragonum A.Cunn. — EU430403; D. trigonopus Rchb.f. — DQ058793; D. victoriaereginae
Loher — EU840694; Eria ferruginea Lindl. — AF521071; Liparis krameri Franch. & Sav. —
AB289469; Oxystophyllum sinuatum (Lindl.) M.A.Clem. — AY 239995.
search level 3, with 50 random addition tree replications. Tree support was tested using
bootstrapping with 1000 replications. Gaps were treated as missing data (‘include all
sites’ option in Mega4). For various small subsamples of the taxa parsimony analyses
were conducted using the exhaustive max-mini branch-and-bound algorithm in Mega4.
These analyses produced tree topologies that were consistent with the ones found in
the complete analysis.
Results
A. Testing for pseudogenes
Harpke & Peterson’s test motif CGATGAAGAACGyAGC is not found in any
species included in this study; all have CGATGAAGAGCGCAGC instead (absence
of the test motif indicates potential pseudogene). On the other hand, the test motif
GAATTGCAGAAwyC is present in all species except in D. hancockii, which has
250 Gard. Bull. Singapore 63(1 & 2) 2011
AAATTGCAGAATCC. The motif GGCry-(4 to 7n)-GyGyCAAGGAA (Feliner &
Rossell6 2007) was found only in Eria ferruginea Lindl., D. mutabile Blume and
D. maidenianum Schltr. The motif GAATTGCAGAATTC, unlike the more general
GAATTGCAGAAwyC recommended by Harpke & Peterson, was not found in any
species, all had GAATTGCAGAATCC, except for D. hancockii Rolfe, which had
AAATTGCAGAATCC. A conserved EcoRV site, GATAC, was not present in D.
nobile Lindl., D. officinale Kimura & Migo (= D. catenatum Lindl.), D. hercoglossum
Rchb.f., D. victoriae-reginae Loher and D. chameleon Ames; these all had GATAT.
These results show that further testing for pseudogenes is indicated for all the species
included in this study, and that at least the sequence for D. hancockii here used is likely
to be a pseudogene. According to Burke et al. (2008) the inclusion of pseudogenes
did not have a significant influence on the phylogeny inferred in their study, except
that longer branch lengths were found as a result. They identified the sequence of
their accession of D. baileyi F.Muell., also used in the present study, as a potential
pseudogene. In theory, the use of paralogous sequences could influence the inferred
phylogeny considerably. However, in Dendrobium strongly supported results using
matK are usually replicated with strong support when using ITS (Wongsawad et al.
2005, Satthapattayanon 2008), and vice versa. This suggests that these results are not
much distorted by the inclusion of pseudogenes, or paralogous sequences in general.
The position of D. hancockii in Fig. 2 is in agreement with its membership of section
Dendrobium on morphological grounds. Likewise, D. baileyi nests with another
member of section Grastidium, as expected (Fig. 1). Nevertheless, the fact that quite
a few species of Dendrobium are unplaced, as discussed below, may indicate that the
role of pseudogenes and other genetic factors, such as ancient hybridization, need
further study.
B. Phylogeny
The aligned sequences had a length of 747 sites (including gaps), of which 395
were parsimony informative. One of the 20 most parsimonious trees (length 2409,
consistency index 0.404, retention index 0.619) is shown here (Fig. 1, 2). In its general
topology it agrees well with earlier studies by Yukawa and co-workers (Yukawa
2001, Yukawa & Uehara 1996, Yukawa et al. 1993, 1996, 2000), Clements (2003,
2006) and Burke et al. (2008). In line with these studies, three main clades (marked
I, II and III) can be distinguished in the ingroup. It is seen that Cadetia, Flickingeria
and Diplocaulobium are all nested within the Australasian clade (II), demonstrating
that recognition of these genera while maintaining Dendrobium in the broad sense
renders the latter paraphyletic, as first noted by Yukawa et al. (1993). The Asian clade
(1) generally shows longer branches than the Australasian clade. In the Asian clade
the average number of changes from the nearest node common to clade I and II to a
terminal node is 65.2; in the Australasian clade this is 28.9 changes.
The following discussion includes the results of the studies cited above. Some
of the sections and species mentioned below are not found in the phylogram shown
here because there were no sequences available from GenBank for these.
A broad generic concept of Dendrobium 251
In contrast to most other studies, bootstrap support for the Dendrobium clade as
a whole was here found to be low. No clear grounds for this anomaly could be detected.
The Epigeneium clade (II) is well supported; it is basal to the two other clades, of which
the Asian clade (1) has very strong bootstrap support here as well as in all studies cited.
The Australasian clade (II) is less strongly supported, as is also seen in other studies.
The basal dichotomy in the Australasian clade is again a result common to all studies.
It represents a split between a clade (Ila) consisting of species from New Zealand and
New Caledonia, traditionally included in section Macrocladium on the one hand, and
a clade (IIb) consisting of numerous Australian/Asian species in many other sections
on the other. Clements (2006) included a wider sample of New Caledonian species,
showing that the sections Macrocladium, Kinetochilus, Dendrobates, Inobulbum,
Finetianthe and Tetrodon are closely related, but the topology of the subtree is not
well supported. The very limited sequence divergence at the higher nodes in this New
Caledonian group hardly validates the recognition of so many sections. In Clements’s
analysis, the New Guinean D. herpetophytum Schltr. (section Herpethophytum) is
nested deeply within this clade. In Yukawa’s (2001) phylogram, on the other hand,
an unidentified species of sect. Herpethophytum resides in a clade with sections
Grastidium, Pleianthe and Biloba (syn. sect. Monanthos), which agrees much better
with morphology and biogeography. As is evident from Clements (2006) and Yukawa
(2001), most of the subclades within clade IIb correspond well with morphologically
recognised sections, such as Cadetia, Diplocaulobium, Brevisaccata (syn. sect.
Trachyrhizum), Crinifera (genus Flickingeria), and Latouria. The last, however, 1s not
strongly supported, even when the anomalous position of D. spectabile (Blume) Miq.
in Clements (2006) is disregarded. In the present study support for clade IIb is much
stronger than it is in Clements (2006).
Sections Spatulata, Phalaenanthe and Eleutheroglossum together form
a well supported clade that is consistent with morphology. It appears that at least
sect. Phalaenanthe is nested within sect. Spatulata and may not warrant continued
recognition.
Another strongly supported, morphologically and biogeographically plausible
clade consists of sections Grastidium (including Eriopexis and Dichopus), Biloba,
Herpethophytum and Pleianthe. Not enough species have been sampled to assess the
robustness of the sections within this Grastidium clade, but the four mentioned here
are easily distinguished morphologically.
Two species of section Fugacia (syn. sect. Euphlebium) were sampled by
Clements (2006). Although this section is well-characterised by vegetative and floral
morphology, the two species appear in two separate clades within clade Ib, but without
strong support one way or the other.
Sections Dendrocoryne, Monophyllaea (syn. Australorchis) and Rhizobium
were analysed in detail by Burke et al. (2008) and Adams et al. (2006). Their work
suggests that, although morphologically distinct, sections Rhizobium and Monophyllaea
are nested within a broadly defined sect. Dendrocoryne. Sect. Rhizobium seems to
represent a recently evolved clade with xeromorphic adaptations (Yukawa et al. 2000).
252 Gard. Bull. Singapore 63(1 & 2) 2011
Finally, still within clade Ib, Yukawa (2001) finds D. bulbophylloides Schltr.
of sect. Microphytanthe to be sister to Flickingeria, but the clade combining these
two has low bootstrap support. In Clements’s (2006) phylogram D. bulbophylloides is
sister to D. toressae (F.M.Bailey) Dockrill (sect. Lichenastrum), with the pair in turn
sister to Flickingeria, but again with low support.
In summary, in the Australasian clade the sectional classification based on
morphology is largely supported by the molecular phylogenies. The relationships
between the clades are largely unresolved, however. The position of Cadetia, for
example, is still unclear.
Turning to the Asian clade (I), the picture is much more confusing. There are
several well supported clades, but also quite a few species that do not nest inside any
of those clades, e.g., the closely related D. /indleyi Steud. and D. jenkinsii Wall. ex
Lindl., D. senile Parish ex Rchb.f., D. trigonopus Rchb.f., and D. capillipes Rchb.f.,
which must all be considered unplaced at present. A broad sampling of the florally
very similar sections Dendrobium and Densiflora (syn. sect. Callista) by Wongsawad
et al. (2005) confirmed that sect. Densiflora is polyphyletic, as found by earlier studies
(e.g., Yukawa 2001), with D. chrysotoxum deeply nested in a well-supported clade
that contains species of sect. Dendrobium and sect. Breviflores. Two other clades
of sect. Densiflora are each strongly supported, one consisting of the species pair
D. lindleyi and D. jenkinsii, the other consisting of the relatives of D. densiflorum
Wall. ex Lindl. (clade D in Fig. 2). However, the position of these clades within the
Asian clade is still undetermined. It is hard to explain how the continental Asian D.
densiflorum alliance could be sister to D. sect. Amblyanthus, a morphologically quite
different section with a predominantly East Malesian distribution, and not to one of
the much more similar clades from continental Asia. Yet, this is suggested by the
analyses by Yukawa (2001), Wongsawad et al. (2005) and the ITS-based analysis of
Sathapattayanon (2008), although not by her matK-based analysis. It should be noted
that branch lengths inferred from ITS sequences in the D. densiflorum clade and in the
Amblvyanthus clade are much shorter (by about a factor 0.5) than in the neighbouring
clades, for instance the second Pedilonum clade (clade P2 in Fig. 2) mentioned below.
This could be a factor in explaining the rather counterintuitive position of clade D.
Section Dendrobium is largely supported by Wongsawad et al.’s study, except for a
few species (Wongsawad et al.’s clade 7, here represented by D. capillipes in Fig. 2)
that form a small, rather inexplicably unplaced group outside the main Dendrobium
clade. Section Breviflores, still according to Wongsawad et al., is polyphyletic, and
should probably be included in sect. Dendrobium.
Section Formosae has recently been analysed in detail by Sathapattayanon
(2008), who found that the section as traditionally circumscribed falls apart into two
distinct sections next to two morphologically aberrant species that have to remain
unplaced for the time being (D. trigonopus and D. jerdonianum Wight). The core of
sect. Formosae forms a well-defined monophyletic group of some 40 species. The
sections Distichophyllae and Conostalix are, according to ITS data, sister to the as-yet-
unnamed clade split off from sect. Formosae, but not according to matK data.
(Nn
Lo
A broad generic concept of Dendrobium 2
Section Stachyobium is a strongly supported clade with high sequence
divergence. Its position within the Asian clade varies more than any other subclade
in the various studies here cited, and this morphologically distinctive section must be
considered unplaced.
Perhaps the greatest problem in the Asian clade is presented by the species of
the sections Pedilonum, Calcarifera, Oxyglossum, Calvptrochilus, Platycaulon and
Dolichophyllum. \t was generally believed that they formed a single, monophyletic
alliance within Dendrobium. The work of Clements (2003, 2006) suggests, however,
that this alliance consists, at the highest level, of two strongly supported clades that
are not sister clades. To some extent, these two clades are supported by geography,
with one clade, P! in Fig. 2, being predominantly West Malesian (this clade includes
the type species for sect. Pedilonum, D. secundum (Blume) Lindl.), and the other, P2
in Fig. 2, being predominantly East Malesian. However, section Platycaulon, which
falls in the West Malesian clade, is well represented in East Malesia. There are few,
if any, morphological characters that can serve to distinguish P| and P2. Species of
the sections Pedilonum and Calcarifera, as traditionally understood, are found both in
clade P| and in P2.
The sister clade to P2 is a strongly supported alliance consisting of the sections
Crumenata, Aporum, Strongyle and Bolbodium. This clade is probably best considered
as a single section (to be called Aporum), because species with virtually identical,
ephemeral flowers occur in each of the four sections. There is some support, however,
for a subdivision into two subgroups, one having stems with a few swollen internodes,
the other having stems without any swollen internodes. Section Bo/bodium is nested
within sect. Crumenata, and appears to represent a case of vegetative reduction or
neoteny. The Aporum clade and clade P2 together form a strongly supported clade in
all studies.
In summary, the Asian clade contains several well-supported subclades, but
these are much less congruent with morphology than in the Australasian clade, and the
relations between the subclades are, in many cases, still unclear.
Discussion
Do these results, and the studies cited here, support the splitting up of Dendrobium into
smaller genera? To some extent they do, in that the three basal clades in the Dendrobiinae
would at first sight be good candidates for recognition at genus level. Not only are
they well supported, there is also a distinct geographical signal present. Unfortunately,
there appear to be no consistent morphological characters to distinguish between
the Asian clade and the Australasian clade. Both groups, if recognised as separate
genera, would be highly heterogenous and an identification key for these two entities
would be too complex to be useful. Therefore, if splitting is considered desirable,
then this has to occur at lower levels. But, as I have pointed out repeatedly above, the
delimitation of several clades is still problematic. There are unplaced species, there are
fragmented morphology-based sections, and there are strongly supported clades that,
254 Gard. Bull. Singapore 63(1 & 2) 2011
if recognised as genera, would be very hard to distinguish from other genera. There is,
beyond the level of the three basal clades, no obvious way to split up Dendrobium into
monophyletic genera.
Apart from these practical difficulties, there are several arguments that can be
invoked in favour of, or against splitting. The main arguments in favour of splitting,
mostly paraphrased from Clements (2003, 2006), are:
1. A very large genus of more than 1500 species is impractical; it makes species-
identification more difficult and membership of such a large genus conveys little
information about the characters of any particular species.
2. Many clades are very distinctive, and some have already been recognised at genus
level for a long time, such as Cadetia and Flickingeria.
3. It is inconsistent to have numerous small genera in one large subtribe, such as the
Aeridinae, and only one big genus in another, even larger one.
4. Splitting is unavoidable, because the group under consideration is polyphyletic.
These are some possible counter arguments:
1. A large genus may be impractical, but so is a large number of genera that are
difficult to distinguish.
2. Many clades are not distinctive at all.
3. There are no rules for the size of genera. The size of a genus depends on a specific,
historical pattern of speciation and extinction, therefore a wide range of sizes among
different genera is to be expected. Moreover, the number of recognised genera 1n the
Aeridinae is probably far too large.
4. After the removal of Pseuderia and Oxystophyllum, the polyphyly argument no
longer applies to Dendrobium.
The main arguments against splitting are as follows:
1. There is no obvious ‘best’ way to split Dendrobium into smaller genera.
2. Splitting is at any rate premature because the presently available phylogenies show
too little resolution, and the sampling has been inadequate for most sections.
3. While many of the segregate genera are easily recognised (Cadetia, Diplocaulobium),
others would be hard to distinguish, even by specialists.
4. Splitting will result in a huge number of name changes, rendering older publications
obsolete and herbarium management more difficult. Reassigning Cadetia and other
segregates back to Dendrobium will do this to a far lesser extent.
5. The horticulturally extremely important D. bigibbum and its hybrids will no longer
belong to the genus Dendrobium.
6. The popular image of taxonomists as people who endlessly tamper with well-
established names will be confirmed.
All arguments considered, and recognising that one of the few objective
criteria that may guide us in the delimitation of genera is the criterion of monophyly,
it seems to me preferable to take a broad view of Dendrobium, and to include in it
all the former segregate genera (except Pseuderia and Oxystophyllum). In principle,
the genus Epigeneium could be kept as a separate genus, as it is sister to all the other
species in the Dendrobiinae. However, morphologically Epigeneium is not particularly
distinctive when compared with such species as Dendrobium carrii Rupp & C.T.White.
bo
Nn
Nn
A broad generic concept of Dendrobium
Inclusion of Epigeneium would not add any new character states, or even any obvious
combinations of character states to Dendrobium. Therefore, apart from its status as
sister genus, there are hardly any solid arguments to maintain this relatively small
genus of some 40 species next to the huge conglomerate of Dendrobium. There is
also no particularly large sequence divergence between Epigeneium and the rest of
Dendrobium.
The case of Dendrobium is not unique: similar problems are posed by other
very large orchid genera including Bulbophyllum, Epidendrum, Eria, Habenaria, and
Pleurothallis, among others. In the case of Eria and Pleurothallis, the arguments in
favour of splitting have prevailed (Pridgeon et al. 2005), and in my opinion rightly so,
since the formerly broadly defined genera were clearly highly polyphyletic. Generic
delimitation is still debatable in some cases. Splitting of Epidendrum (Pridgeon et al.
2005) and Bulbophyllum (Vermeulen, pers. comm.) has so far not been favoured by
modern experts in these genera. At least in the case of Bulbophyllum, the arguments
are very much like those listed above for Dendrobium, except that Bulbophyllum has
far less significance in horticulture.
Character evolution and the origin of Dendrobium
Roughly speaking, there are two main habit types in Dendrobium: plants with
pseudobulbs and plants with cane-like stems. This is an oversimplification, as there are
many intermediate cases, but it does raise the question what the earliest dendrobiums
looked like. I believe the molecular phylogenies enable us to answer this question. We
now know that the Epigeneium clade is sister to all the other Dendrobium species, and
we also know that Bulbophyllum is sister to Dendrobium s.1. All Bulbophyllum species
have heteroblastic pseudobulbs with one or two, non-sheathing apical leaves. The
same is true for Epigeneium. Moreover, Epigeneium has distinct, creeping rhizomes,
a character-state also found in numerous Bulbophyllum species. These facts suggest
that the earliest dendrobiums were plants with heteroblastic pseudobulbs with one
or two terminal, non-sheathing leaves and creeping rhizomes. Species, indeed whole
sections, with such a habit are still found in the Australasian clade, for example in
sections Diplocaulobium, Rhizobium, Microphytanthe, and others. However, they are
not basal in this clade, and are absent from the Asian clade. This suggests that the
most recent common ancestor to the Asian and Australasian clade was a plant with a
cane-like habit, and that cases like Diplocaulobium represent reversals to the earliest
ancestral type.
How early were those earliest ancestors? There are various theories about the
origin of Dendrobium. The clear split between an Asian and an Australasian clade (or
what Wood (2006) refers to as the Northern and Southern clade) has led some authors
(Brieger 1981, Wood 2006) to propose a very early origin for Dendrobium, as they
invoke the splitting up of Gondwanaland (around 60—70 million years ago) to explain
the existence of the two clades. However, if that were the case, then one would expect
far greater sequence as well as morphological divergence between the two clades than
is actually observed. A more recent origin in continental Asia, or possibly Africa. with
subsequent dispersal to Australia, New Caledonia and New Zealand, as advocated
256 Gard. Bull. Singapore 63(1 & 2) 2011
by Lavarack et al. (2000), seems more plausible. Dispersal in the opposite direction
would explain why certain members of the Australasian clade are in fact better
represented in the western (Asian) part of the range of Dendrobium, in particular sect.
Crinifera (the former genus Flickingeria). The radiation of clade P2 in New Guinea
may be the result of a relatively recent secondary west-to-east dispersal. The complex
geology of Wallacea, with land rapidly appearing and disappearing within a few
million years around 10 million years ago (Hall 2009), may explain these secondary
dispersals. As for the primary dispersal from Asia to Australia, New Caledonia, and
New Zealand, this must have occurred more than 23 million years ago, assuming that
a recently discovered, c. 23 million year old fossil from New Zealand described as D.
winikaphyllum Conran, Bannister & Lee (Conran et al. 2009), is indeed a Dendrobium
species. Gustaffson et al. (2010), using this fossil to provide a calibration point, dated
Dendrobium to about 32 mya, with a 95% confidence interval of 25 to 40 my.
References
Adams, P.B., Burke, J.M. & Lawson, S.D. (2006) Systematic analysis of Dendrobium
Swartz section Dendrocoryne in the Australian region. P/. Syst. Evol. 260: 65-80.
Brieger, F.G. (1981) Subtribus Dendrobiinae. In: Brieger, F.G., Maatsch, R. &
Senghas, K. (eds) Die Orchideen by Schlechter, R., 3rd ed, (11—12): 636-752.
Berlin: Verlag Paul Parey.
Burke, J.M., Bayly, M.J., Adams, P.B. & Ladiges, P.Y. (2008) Molecular phylogenetic
analysis of Dendrobium (Orchidaceae), with emphasis on the Australian section
Dendrocoryne, and implications for generic classification. Austral. Syst. Bot. 21:
1-14.
Clements, M.A. (2003) Molecular phylogenetic systematics in the Dendrobiinae
(Orchidaceae), with emphasis on Dendrobium section Pedilonum. Telopea 10:
247-298.
Clements, M.A. (2006) Molecular phylogenetic systematics in Dendrobieae
(Orchidaceae). Aliso 22: 465-480.
Clements, M.A. & Jones, D.L. (2002) Nomenclatural changes in the Dendrobieae
(Orchidaceae). 1: The Australasian region. Orchadian 13: 485—497.
Conran, J.G., Bannister, JM. & Lee, D.E. (2009). Earliest orchid macrofossils: early
Miocene Dendrobium and Earina (Orchidaceae: Epidendroideae) from New
Zealand. Amer. J. Bot. 96: 466-474.
Cribb, P. & Govaerts, R. (2005) Just how many orchids are there? In: Raynal-Roques,
A., Roguenant, A. & Prat, D. (eds) Proceedings of the 18th World Orchid
Conference. Turriers: Naturalia Publications.
Dressler, R.L. (1981) The Orchids: Natural History and Classification. Massachusetts,
Cambridge: Harvard University Press.
Dressler, R.L. (1993) Phylogeny and Classification of the Orchid Family. Oregon,
Portland: Dioscorides Press.
A broad generic concept of Dendrobium 257
Feliner, GIN. & Rossell6, A.N. (2007) Better the devil you know? Guidelines for
insightful utilization of nrDNA ITS in species-level evolutionary studies in plants.
Molec. Phylogen. Evol. 44: 911-919.
Gustafsson, A.L.S., Verola, C.F. & Antonelli, A. (2010). Reassessing the temporal
evolution of orchids with new fossils and a Bayesian relaxed clock, with
implications for the diversification of the rare South American genus
Hoffmannseggella (Orchidaceae: Epidendroideae). BMC Evol. Biol. 10: 177
(http://www.biomedcentral.com/1471-2148/10/177).
Hall, R. (2009) Southeast Asia’s changing palaeogeography. B/umea 54: 148-161.
Harpke, D. & Peterson, A. (2008) 5.8S motifs for the identification of pseudogenic ITS
regions. Botany 86: 300-305.
Lavarack, P.S., Harris, W. & Stocker, G. (2000) Dendrobium and its Relatives. Oregon,
Portland: Timber Press.
Pridgeon, A.M., Cribb, P.J.. Chase, M.W. & Rasmussen, F.N. (eds) (2005) Genera
Orchidacearum, Epidendroideae, vol. 4(1). Oxford: Oxford University Press.
Sathapattayanon, A. (2008) Taxonomic revision of orchids in the genus Dendrobium
Sw. section Formosae (Benth. & Hook.f.) Hook-f. in Thailand and adjacent areas.
Thesis, Chulalongkorn University, Bangkok.
Schlechter, R. (1911-1914) Die Orchidaceen von Deutsch-Neu-Guinea. Repert. Spec.
Nov. Regni Veg. Beih. 1: 1-1079.
Tamura, K., Dudley, J.. Nei, M. & Kumar, S. (2007) MEGA4: Molecular Evolutionary
Genetics Analysis (MEGA) software version 4.0. Molec. Biol. Evol. 24: 1596-1599.
Wongsawad, P., Handa. T. & Yukawa, T. (2005) Molecular phylogeny of Dendrobium
Callista - Dendrobium complex. In: Nair, H. & Arditti, J. (eds) Proceedings of the
17th World Orchid Conference, Shah Alam 2001. Kota Kinabalu: Natural History
Publications (Borneo).
Wood, H.P. (2006) The Dendrobiums. Ruggell: A.R.G. Gantner Verlag.
Yukawa, T. (2001) Molecular phylogeny of Dendrobium. In: Nagata, H. & Ichihashi,
S. (eds) Proceedings of the seventh Asia Pacific orchid conference (APOC7), pp.
69-71. Nagoya, Japan.
Yukawa, T. & Uehara, K. (1996) Vegetative diversification and radiation in subtribe
Dendrobiinae (Orchidaceae): evidence from chloroplast DNA phylogeny and
anatomical characters. P/. Syst. Evol. 201: 1-14.
Yukawa, T., Kita, K. & Handa, T. (2000) DNA phylogeny and morphological
diversification of Australian Dendrobium (Orchidaceae). In: Wilson, K.L. &
Mossison, D.A. (eds) Monocots: Systematics and Evolution, pp. 465-471.
Melbourne: CSIRO Publishing.
Yukawa, T., Kurita, S., Nishida, M. & Hasebe, M. (1993) Phylogenetic implications of
chloroplast DNA restriction site variation in subtribe Dendrobiinae (Orchidacaeae).
Lindleyana 8: 211-221.
Yukawa, T., Ohba, H., Cameron, K.M. & Chase, M.W. (1996) Choloroplast DNA
phylogeny of subtribe Dendrobiinae (Orchidaceae): insights from a combined
analysis based on rbcL sequences and restriction site variation. J. P/. Res. 109:
169-176.
Gardens’ Bulletin Singapore 63(1 & 2): 259-275. 2011 259
Implications from molecular phylogenetic data
for systematics, biogeography and growth form evolution
of Thottea (Aristolochiaceae)
Birgit Oelschlagel'’, Sarah Wagner’, Karsten Salomo’,
Nediyaparambu Sukumaran Pradeep*, Tze Leong Yao’,
Sandrine Isnard'’, Nick Rowe’,
Christoph Neinhuis' and Stefan Wanke’
‘Institute for Botany, University of Technology Dresden, D-01062 Dresden, Germany
*birgit.oelschlaegel@tu-dresden.de (corresponding author)
* Molecular Phylogenetics Lab, Tropical Botanic Garden & Research Institute,
Palode, Thiruvananthapuram — 695 562, Kerala, India
* Forest Biodiversity Division, Forest Research Institute Malaysia,
52109 Kepong, Selangor, Malaysia
> Univ. Montpellier2, UMR AMAP, Montpellier, F-34000, France;
CNRS, UMR AMAP, Montpellier, F-34000, France
ABSTRACT. The genus 7hottea comprises about 35 species distributed from India throughout
Southeast Asia. However, most of the species have a narrow distribution. A first molecular
phylogeny based on the chloroplast #K intron, matK gene and t7nK-psbA spacer is presented
and confirms the monophyly of the genus according to Hou in 1981. Earlier subdivisions into
the sections or genera Apama and Thottea could not be substantiated since both proved to be
paraphyletic with respect to each other. The taxonomic and systematic history of Thottea is
discussed with respect to molecular and morphological data. Thottea piperiformis is sister to all
other species, which gives limited recognition to Asiphonia piperiformis as proposed by Huber
(1985). Thottea tomentosa, one of the smallest and most widespread species is subsequently
sister to all remaining species. Thottea diversified in two biogeographic regions: the Western
Ghats in India and the Indo-Malayan region. A high degree of endemism is observed resulting
from the presence of very few species shared between islands, which might be the result of a
single colonisation and subsequent radiation. Within Piperales, Thottea holds a key position
between the herbaceous Asaroideae and the woody Aristolochioideae.
Keywords. Aristolochiaceae, biogeography, growth form evolution, molecular phylogeny.
Thottea
Introduction
During the last two centuries, the number of accepted genera within Aristolochiaceae
has differed according to the dataset investigated (e.g., seven by Gonzalez & Stevenson
2002; five by Neinhuis et al. 2005 and Ohi-Thoma et al. 2006). Currently, the family
Aristolochiaceae is divided into two subfamilies Aristolochioideae and Asaroideae
and four genera are consistently recognised (Saruma Oliv., Asarum L., Thottea Rottb.
260 Gard. Bull. Singapore 63(1 & 2) 2011
and Aristolochia L.) (Wanke et al. 2007a). However, Aristolochiaceae has turned out
to be paraphyletic with respect to Lactoris fernandeziana Phil. and probably also
Hydnoraceae (Nickrent et al. 2002; Wanke et al. 2007a), which will not be addressed
here further. The subfamily Asaroideae contains small-sized herbaceous plants with
flowers characterised by an actinomorphic perianth. It consists of two genera: the
monotypic Saruma (S. henryi Oliv.), endemic to central China (Zhou et al. 2010)
and Asarum with about 86 species from temperate areas of North America, Europe
and Asia (Kelly 1998, Kelly & Gonzalez 2003, Wanke et al. 2006a). In contrast, the
Aristolochioideae are distributed from tropical to temperate climate zones (Neinhuis
et al. 2005). Thottea includes about 35 shrubby species with an actinomorphic perianth
restricted to tropical Asia, while Aristolochia is the most species-rich genus with about
400 species representing geophytes, perennial herbs, climbers and shrubs (Wanke et
al. 2006a).
Most likely due to the lack of Thottea in ex situ collections, only a few species
have ever been included in molecular-based phylogenetic studies (e.g., Neinhuis et
al. 2005, Ohi-Thoma et al. 2006, Wanke et al. 2006a, 2007a). However, all studies
have assumed the monophyly of the genus based on morphological characters. In
addition, traditional taxonomic concepts and infrageneric relationships have not yet
been addressed using molecular data.
Since the genus 7hottea was described by Rottb6ll (1783) seven further genera
have been published and used by later authors at tribal or sectional levels:
Thottea Rottb. (type: T. grandiflora Rottb.), Nye Dansk. Vidensk. Selsk. Skrift. 11.
(1783) 529s 2:
Apama Lam. (type: A. siliquosa Lam.), Encycl. (Lamarck) 1(1). (1783) 91;
Bragantia Lour. (type: B. racemosa Lour.), Fl. Cochinch. 2. (1790) 528;
Ceramium Blume (type: C. tomentosum Blume), Bijdr. Fl. Ned. Ind. 17. (1826-27)
1134, nom. illeg.: renamed as Munnickia Rchb., Consp. Regn. Veg. 85 (1828),
Vanhallia Schult.f., Syst. Veg. 7 (1829) xviii &166, and Cyclodiscus Klotzsch,
Monatsb. Akad. Berl. (1859) 591.
Trimeriza Lindl. (type: T. piperina Lindl.), Edwards’s Bot. Reg. 18. (1832) sub t.
1543;
Asiphonia Griff. (type: A. piperiformis Griff.), Trans. Linn. Soc. London 19. (1845)
Ese) lies) ie
Lobbia Planch. (type: L. dependens Planch.), London J. Bot. 6. (1847) 144, t. 3;
Strakaea C.Presl. (type: S. melastomaefolia C.Presl.), Epimel. Bot. (1851) 221;
Different taxonomic concepts are shown in Table |. Klotzsch (1859) accepted
three genera in two tribes, while Duchartre (1864) recognised only two: Thottea and
Bragantia, and put all other genera in synonymy. He distinguished both genera by
the number and arrangement of stamens. Whereas Thottea possesses 16—36 stamens
arranged in two whorls (e.g., TZ. abrahamii M.Dan, P.J.Mathew, Unnithan & Pushp.,
Fig. 1A), Bragantia exhibits one whorl with only 6 to 10 stamens (e.g., 7. barberi
(Gamble) Ding Hou, Fig. 1B). Hooker adopted this classification but noted that both
genera might “well be united” (Hooker 1890) because of their morphological similarity.
Solereder (1894) also accepted the division into two genera but renamed Bragantia
261
ylogeny of Thottea
Molecular ph
‘JID viuoydisy
VINVAY “JOOS
VALLOH [JOOS
‘qnoy Payoy]
(S861) toqny
[PUI] PZMaUA]
Qnory DIANOY |
S[SOq’) VAPYDAS
“Your]d ViqgoT
‘YOSZIOTS
SNISIPSOJIA)
SOUIN] | WWUD.10,y
SINO'T VIUDSDAG
JQ piuoydisy
Suey] piupdy “jou
‘qnoy vaioy]
(1861) no
Y9SZ}O[
SNISIPOJIA) SOU] |
WNIUDAIY “UI
SNOSIGOTIAD
‘TIT (9s
[PUL] VZMOULAT
“yuey] puipdy “out
VZRIAINET, {[] “190s
‘[S81d"D
DODYDAIG NOT]
DIUDBDAG “JID
piuoydisp “{out
VIINVOVUg :] ‘1998
‘wey vupdpy
‘qnoy vanoy]
Your] viqqo7] ‘jour
‘qnoy Payoy |
(P68 1) 19po19]0g
[Pury]
DZIAIUAT, S|SO.1g' dD
DIDYDAJS YISZIO| S|
SNISIPOJIA) SOU] |
WNIUDAIY SANO7T
DIUDBDAG °° UeT
puwdy yao
piuoydisp *{oul
‘INO'T VIJUDSBPAG
‘qQnoy vayoy |
“Youe]d VIGgGOT "oul
“qQnoy Paoy |
(0681) 9400]
YOSZ}O]¥
SNISIPOJIA,) SOUIN]
WNIWUDAD) {OUT
SNOSIGOTIAD
‘TTT “1095
[PUL] VZMOULAT
Syuey puipdy “out
VZRIAWRI], i] “JOOS
‘[Sed'D
DIDYOMS INO7
DIUDBDAG 2 IJLAD)
piuoydisp “out
VILNVOVUEA 2] JOOS
INO] VIULSDAG
your} PIggoT [out
VIO] ‘[] 1998
‘qUuoy Payoyy you
VALLOH I], ‘] ‘1998
YOsZO,y VaoYy |
(P98 |) omseyonq
YOsZJO] 4
SNISIPOJIAD SUN] |
WNMWUDAI) “JOUI
YOSZO[Y SNISIPOJIAD
SRIUTOSIPO]DAD SQ.
[Pury
DZIMIWIAT SNOT
DIUDBPAG SILO)
piuoydisp “ou
INO’7T Pyuvspig
‘qnoy Payoy ]
Your] PIggoOT ‘jour
‘qnoy Payoy
ovonuvseig SNQUy,
(6S81) YOSZIO] y
‘|S PAOYT SNUIS dy} JO S1d99UOd SILUOUOXe} JUSLING puK [LUOTIPeAL, *[ 9qQuy,
262 Gard. Bull. Singapore 63(1 & 2) 2011
Fig. 1. Diversity of flowers and vegetative organisation within 7hottea s.1. (Hou 1981). A.
Flower of 7. abrahamii M.Dan, P.J.Mathew, Unnithan & Pushp. (sect. Thottea) with stamens
arranged in two whorls. B. Flower of 7. barberi (Gamble) Ding Hou (sect. Apama), possessing
9 stamens in one whorl. C. Growth habit of 7. grandiflora Rottb. D. T. tomentosa (Blume)
Ding Hou, reaching only 40 cm in height. The flowers of this species appear at ground level
(indicated by an arrow). E. 7: piperiformis (Griff.) Mabb., growing up to 2 m or more with
flowers borne in the axils of leaves. In addition to seed anatomy (Huber 1985), 7. piperiformis
also differs in growth form from all other species of the genus by its acrotonic branching.
as Apama, since this is the older name and has priority. About a century later, Hou
published the most comprehensive study on the genus Thottea so far (Hou 1981). His
study revealed that the differentiating characters of both genera were highly variable
within species and even within one specimen. Based on these results he merged Thottea
and Apama into one large genus Thottea. Furthermore, he explicitly pinpointed that
no infrageneric subdivision is needed. A few years later, Huber (1985) again excluded
one species, Asiphonia piperiformis Griff., due to the conspicuous seed anatomy. More
recently six new species were described from India (Swarupanandan 1983; Sivarajan,
1985; Sivarajan & Babu 1985; Pandurangan & Nair 1993; Dan et al. 1996; Kumar et
Molecular phylogeny of Thottea 263
al. 2000). Although the flower and inflorescence morphology of the new Indian species
were investigated from a systematic perspective (Shajiu & Omanakumari 2009, 2010),
no further comprehensive study on the whole genus 7hottea has been performed since
then.
The aims of this study are to |) test the monophyly of the genus Thoffea sensu
lato (s.1.) (Hou 1981) based on about half of the described species; 2) compare the most
recent taxonomic concepts of Hou (1981) and Huber (1985), namely the treatment of
the species as one single genus Thottea, and the separation of Asiphonia with the
subdivision of the genus into the sections Apama and Thottea (Duchartre 1864), by
means of molecular phylogenetic approaches; 3) compare results of a molecular
phylogeny of the Indian species with the recently published results on flower and
inflorescence morphology; and 4) provide a first molecular phylogenetic hypothesis
as a Starting point for more detailed studies addressing biogeographical questions,
character evolution with respect to growth forms in Thottea and Aristolochiaceae and
a revision of the genus reflecting natural relationships.
Methods
For this study, full sequences of three chloroplast regions (#7nK intron, matK gene
and trnK-psbA spacer) were generated for 15 Thottea species, as well as 21 species
representing the other lineages of Aristolochiaceae and 3 outgroup genera of
Saururaceae. For the latter, sequences of the ¢rnK intron and the matK gene were derived
from earlier studies (Wanke et al. 2006a, b, 2007a), whereas the t1nK-psbA spacer was
sequenced for this study. In Appendix A the origin of the material, voucher information
and botanical garden accession numbers as well as GenBank accessions are provided.
Total genomic DNA was isolated from herbarium specimens or leaves collected from
the field or botanical gardens and dried in silica gel. A double-extraction approach with
CTAB was used according to Borsch et al. (2003). After precipitation in ethanol and
resuspension of the pellets in TE, DNA was cleaned by using the NucleoSpin® Extract
II-Kit (Macherey-Nagel).
The amplification of the entire gene cluster was performed in one part for
silica-dried material or in three parts with several 100 bp overlap for material from
herbarium specimens. Primer sequences for amplification and sequencing are listed in
Table 2. PCR products were obtained using a 25 ul reaction containing | ul template,
15.3 pl ddH,0, 2.5 yl 10x Taq buffer (15 mM MgCl), 1 ul of 25 mM MgCl, 0.5
ul of each primer (10 pmol/l), 4 wl dNTP mix (1.25 mM each), 0.2 ul Taq DNA
polymerase (Promega). Amplification conditions were: one cycle of 1.5 min at 96°C,
1 min at 50°C, 2 min at 68°C, 34 cycles of 0.5 min at 95°C, 1 min at 48°C, 2 min at
68°C, and a final extension of 20 min at 68°C in a T3 Thermocycler (Biometra). PCR
products were purified and extracted from a 1.2% agarose gel, using the NucleoSpin®
Extract II-Kit. Sequences were run with an in-house Beckman Coulter 8000 capillary
sequencer or sent to Macrogens’ sequencing service (Macrogen Inc., Korea).
264 Gard. Bull. Singapore 63(1 & 2) 2011
Table 2. Amplification and sequencing primers used.
Primer name Direction Sequence (5’-3’) Design Primer used for
trnK-F forward GGG TTG CTA ACT Wicke & all taxa
CAA TGG TAG AG Quandt (2009)
psbA-R reverse CGC GIC TCheiTA Steele & all taxa
AAATTGCAGTCAT Vilgalys (1994)
AR-matK-2400R reverse ATT TTC TAG CAT Wanke et al. Aristolochia
ITGACIT EE (2007a)
AR-matK-2660F forward CTT ATG ATG AAG this study Aristolochia
AAA TGG AAA TA
AR-psbA-3720R _ reverse CCC ATT TGY TAT this study Aristolochia
TIC GGAT
AR-trnK-3480F — forward ATT CTG AAA TGT this study Aristolochia
TTA CRC AGT AGT
Th-matK-I510R reverse TAA ACT CCT GAA this study Thottea
AGA GAA GTG G
Th-matK-2000F — forward TIA TGG GCT ATC this study Thottea
TTT CAA GTC G
Th-matK-2190R reverse TAT CAG AAT CAG this study Thottea
ACG AAT CGG C
Th-matkK-910F forward GAC TGT ATC GCA this study Thottea
CTA TGT ATC G
Sequences were manually edited and aligned using the Phylogenetic Data Editor
PhyDE* v0.995 (www.phyde.de) following alignment rules proposed by Kelchner
(2000) and Borsch et al. (2003). Sixteen hotspots were excluded from the original
dataset prior to the phylogenetic analyses due to ambiguous homology assessments.
The dataset contained two inversions, one in the genus 7hoftea s.l1. and one in Asarum
and Saruma. To use both for phylogenetic reconstruction, the information on presence/
absence of the inversion as well as the mutational events within, the inversions were
coded in two additional columns at the end of the alignment and reversed in the
alignment. Subsequently, indels were automatically coded using the simple indel coding
approach according to Simmons & Ochoterena (2000) as implemented in SeqState
(Miiller 2005a), a PhyDE® plugin. The alignment and the indel matrix are available
from TreeBASE (www.treebase.org). For phylogenetic reconstruction Maximum
Parsimony and Bayesian Inference methods were employed. The most parsimonious
trees where obtained by using the parsimony ratchet (Nixon 1999), as implemented in
PRAP2 (Miiller 2005b). Ratchet settings were set at 20 random addition cycles of 500
ratchet replicates up weighting randomly 25% of the characters during each iteration.
A strict consensus tree was calculated and nodes were evaluated by bootstrapping
Molecular phylogeny of Thottea 265
(BS) in PAUP* v.4.0 (Swofford 2002) using 1000 replicates. MrBayes v3.1 (Ronquist
& Huelsenbeck 2003) was utilised for Bayesian Inference analysis. The GTR + I
+ | model was applied for sequence data, and the restriction site model (“F81”) for
the indel matrix after testing the best fitting model using jModeltest 0.1.1 (Posada
2008). Ten independent runs with 1,000,000 generations and 4 chains each were run
simultaneously. Every 100th generation of each run was collected. The burnin was
evaluated using Tracer v1.5 (Drummond & Rambaut 2007). A consensus tree and the
posterior probabilities (PP) were calculated after discarding the first 50,000 sampled
generations of each run as burnin.
Results
Characterisation of the molecular dataset
The total length of the alignment comprises 4440 bp, the mean sequence length 2887
bp (min = 2730 bp, max = 3272 bp), while the mean sequences in Thottea had a length
of 2806 bp (min = 2763 bp, max = 2821 bp). Two inversions were detected. One
in Thottea (position 226 to 253 in the ¢rnK intron) forming a hairpin with a poly-T
microsatellite as terminal loop (Wanke et al. 2007a) and one in Asarum and Saruma
(position 4178 to 4185 in the trnK-psbA spacer). The dataset exhibited a large number
of length mutations (indels), 222 of which were identified by SeqState. The combined
data matrix (excluding hotspots) comprised a total number of 3802 characters, 1361 of
which were variable and 906 parsimony-informative.
Phylogenetic reconstruction
The phylogram obtained by Bayesian Inference is shown in Fig. 2. Maximum
Parsimony (MP) analyses resulted in 18 most-parsimonious trees of 2248 steps (CI =
0.736, RI = 0.901). The topology of the Bayesian and the MP strict consensus tree are
virtually identical among early nodes and therefore only one tree is shown. However,
within Thottea s.|. differences in branching patterns are observed for nodes lacking
statistical support.
All Thottea species sampled form a single clade, which has maximum
statistical support in both MP and Bayesian analyses (Fig. 2), as well as the sister group
relationship of Thottea to Aristolochia. Thottea piperiformis, which Huber (1985)
separated from Thottea as a monotypic genus (Asiphonia piperiformis), appears as
sister to all other Thottea species. Thottea tomentosa, the most widespread species,
is subsequent sister to the remaining species (PP 0.99, BS 97). The remaining species
are found in two main clades. One 1s statistically highly supported (PP 1.00, BS 100),
containing all Southeast Asian species, whereas the other is statistically unsupported
and contains all Indian species. Within the clade of Indian species 7. abrahamii and T.
dinghoui are branching first, followed by T. barberi. Thottea ponmudiana is subsequent
sister to the remaining species. Relationships among the rest are statistically supported
for T. dalzellii being sister to T. sivarajanii, but their sistergroup relationship to T.
siliquosa, T. idukkiana, T. duchartrei and a yet unidentified accession is unsupported.
Gard. Bull. Singapore 63(1 & 2) 2011
266
‘ASO;OYdIOW IBY)
uo paseq sdnosd dary} dS9y} 0} PAUdIsse O19M PIPUT WOIJ saisads paqiiosap Apuases sYL (S861) JoqnH pue (P98) aeyonq Jo s}daouod o119u08
ayy squasaidas (161 NOH) ‘TS PavoyZ ULYM WOBOYISse]S dLouasqns sy} ‘PoArosqo ayes UONNISGNs JeNjoe IY} JO qjuaosod (yg Ajuo juosoidas
//, Sq paidnuayur sayourig ‘pousisse jou oe ‘Ajaatoodsas ‘Auowlssieg UINUUIXe|) pue uvIsokeg JO YS PUB CY UR} Sse] SonyeA Ajiqeqold
‘soyourig ay} MOojaq UMOYS ose sIsXjeuR AUOLISIed WINUITXe|Y Woy pouleygo (Sq) sonjea densjoog pur aroge uadts ore (dd) samtyiqeqoid
JOLIN|SOg *SUOTIEINU YASUI] papoo Surpnyoul ‘uoiso1 pqsd-yjou- yu. yse|doroyyo ay} UO paseq ‘siskjeue uvIsadeg Wo poAaliop weiso]Ayd “7 “SIA
[ ByOsNIe eIydO|O\SUY
eu06i06 eiysojojsuy
ByeuUIWNOR eIYIO|O}SHY
eye}09}9eIg BIYD0|0}sL\y ——
BpuNjO) BIydo|O\SHLYy
| shijewiayo elysojo)suy ‘\n =
e1yoo|o\sid eIyDo|O\SUY a 50
PYJUBIOIL BIYDO|O\SUY 001 nal
uaupul| ae ke; Ca =r | — oe |
BIYDO/OISHY eyeige| e1490/0)suy —————> 9
si|A\sojoy e1y90|0}Si\y ——————
eueiselo) e1y90|0\Sl\y¥ ——— ——-
BUND} BIYO|O}Sy ————— 001
essiwoid eiydojo\si\y —— F ea:
sisuaweued eIyoo|o}suy a:
| sisuasnysueW eI4yoo|o}suiy aT oot
| PYE|NIIJad EIYDO|O\SUY j ae
| 004 |
Seapiolly 0j0}sUy/ | euejuadias elYysojo\suiy - t |
| eyeqojjiuad eayjoyul | 89
| | ]
eisy-aS ea}yoyL
sIsuaauJog Bayou, t
BsoyiAsed eayoul | ee
laseyonp ovo |
OL i}
euelyyNp! Bayou L
| |
gads eayjouL
|
esonbijis BayouL Ta
ewedy| t
| | | | iuefeseAis eayjoyuL iz oc |
J's BONOUL | e/pul ljjazjep eayouy — °° ‘ | | t
i | | be 00
| | eueip od eayoyL J T [46 —— a = =
| | uaqieg eayouL yt) |B |
| inoyBuip eayVoyL, — oni : | |
| eaNouL Ea | = = Oa = =|
| lueyeuge BAYOU) I | i
| BISW-3S + BIpul |
BSO}JUBWO) BAYOU] |
| | eIsy-3S |
siwJojedid eay;oy, ——_@@ —_A——
asuaebueyobuiys wmesy
aeapioesy |
P' MA iAsuay eBuimes
aesoesnines
Molecular phylogeny of Thottea 267
The Southeast Asian clade contains four species (7. parviflora, T. borneensis, T.
grandiflora and T. penitilobata).
Discussion
Monophyly of Thottea and taxonomic concepts
Phylogenetic analyses, based on half of the currently accepted species, demonstrate
that Thottea s.|. is monophyletic. The study includes a small but biogeographically
representative selection of the Southeast Asian species, as well as a complete taxon
sampling of the Indian species, which were newly described since the last revision of
Thottea by Hou in 1981. Furthermore, the specimens studied represent all traditional
taxonomic subdivisions and all the morphological diversity of the genus. Therefore
the systematic concept of Hou (1981) based on morphology is substantiated as well as
the assumption of Hooker (1890), that both genera, Apama and Thottea, *...may well
be united...”
We assigned species published after the treatment of Hou (1981) and Huber
(1985) to the respective sections based on morphological characters that have
traditionally been used to delimit the sections. Both sections are paraphyletic with
respect to each other. Although the relationships between the different Apama and
Thottea representatives are not highly supported in all nodes, a broader sampling
or better resolution will not achieve monophyly. Consequently, the morphological
characters used by Duchartre (1864), i.e. the number and arrangement of the stamens,
are a result of parallel evolution. It is well known that the number of floral organs in
basal angiosperms is not strictly determined (Soltis et al. 2009; Chanderbali et al., in
press) and might therefore be of less systematic value at the species level. Thottea
shows a comparatively high variability of flowers and especially a high plasticity of
the androecium (Hou 1981, Leins et al. 1988, Shajiu & Omanakumari 2010). Whereas
in earlier studies only two stamen whorls have been proposed, the detailed study of
Hou (1981) revealed up to four whorls. Hou (1981) found 6 to 46 stamens and 2 to
20 styles per flower. Both, the number of styles and stamens per flower have been
shown to vary between different individuals of the same species and even within one
single individual (Hou 1981, Shajiu & Omanakumari 2010). However, species with
low stamen numbers tend to have a lower stamen variability (e.g., 7. duchartrei, 8-10
stamens) or the number is even constant (e.g., 7. tomentosa 6 stamens), whereas in
species with higher stamen numbers greater variability is observed (e.g. T. dinghoui,
15-30 stamens) (Hou 1981, Shajiu & Omanakumari 2010).
The sister relationship of 7. piperiformis (Fig. 1E) to all other Thottea species
(Fig. 2) may appear to lend support to the treatment by Huber (1985), who segregated
this species from Thottea s.1. under its former name Asiphonia piperiformis. However,
after comparing the sequences, the number of substitutions is not higher than in other
Thottea species (Fig. 2). We therefore follow Hou (1981) in accepting only one single
genus Thottea including Asiphonia piperiformis.
268 Gard. Bull. Singapore 63(1 & 2) 2011
Say
Morphological characters of the flowers and inflorescences investigated by
Shajiu & Omanakumari (2009, 2010) substantiate the relationships within the Indian
species complex. The sister group relationship of Thottea abrahamii and T. dinghoui to
the remaining species is characterised by a racemose inflorescence and bi-lobed floral
bracts (Shajtu & Omanakumari 2009) as well as a high number of (15-39) dorsifixed
stamens that are arranged in two whorls (Shajiu & Omanakumari 2010). In contrast,
all other Indian species show cymose inflorescence patterns (Shajiu & Omanakumari
2009) as well as a lower number of (mostly 9) ventrifixed stamens arranged in one
whorl (Shajiu & Omanakumari 2010). The segregation of the next clade, 7. barberi,
is supported by the equal distribution of the stamens around the styles, whereas a
pattern of 3+3+3 stamens substantiates the relationship of 7: duchartrei, T. idukkiana,
T. ponmudiana, T. siliquosa and T. sivarajanii. The close relationship of the latter five
is also confirmed by the presence of a gynostemium that, in contrast, is absent in 7:
abrahamii, T: dinghoui and T. barberi. Furthermore, the affinity of 7) idukkiana and
T. duchartrei 1s retrieved in terms of the following morphological characters of their
flowers and inflorescences: the very small prophyll in comparison to the floral bracts,
fused sepals, the presence of sterile appendages on the gynostemium that are assumed
to be staminodes and the co-occurrence of entire as well as bifid stylar lobes (Shajiu
& Omanakumari 2009, 2010). A morphological investigation of the undetermined
species, which is cultivated in the Botanical Garden, Dresden, and resolved in the
phylogeny together with 7) idukkiana and T. duchartrei, revealed—besides a few
differences—a high affinity to T. idukkiana. However, it has been used erroneously in
our former studies as 7: si/iqguosa (Neinhuis et al. 2005; Wanke et al. 2006a; Wanke et
al. 2007a, b).
Outlook on biogeography and growth form evolution
From a biogeographic point of view, Thottea possibly represents an interesting case
to study Southeastern Asian biogeography west of the Wallace line (Wallace 1859,
1863), as well as floristic affinities of this region to the Western Ghats in India, and
island biogeography in general (Fig. 3). At first sight, the distribution of the genus
seems rather constrained: from India to the Philippines and to the Greater Sunda
Islands including one species crossing the Wallace line to Sulawesi (7: celebica). It 1s
clear that Thottea diversified in two biogeographic regions: the Western Ghats in India
and the Malesian region. In addition, a comparison of the biodiversity, distribution
and similarity of species across the Islands, indicates that in most cases, only one
species is shared between them, resulting in a high degree of endemism. Exceptions
to this include only Sumatra, the Malay Peninsula, and Kra Isthmus, which share four
species. However, floristic similarities of Sumatra and Malay Peninsula (Welzen et al.
2005) as well as Kra Isthmus and Malay Peninsula (Woodruff 2003) are well known.
Thottea tomentosa (Fig. 1D) presents the link between the Indian species and the
Southeast Asian species (excluding 7 piperiformis). It is the smallest shrub within
the genus that normally bears only 2 or 3 leaves per stem and is found throughout
the western distribution area of the genus. Recently, Sumathi et al. (2004) reported
the occurrence of T paucifida from the Andaman Islands (not sampled in this study).
|
Molecular phylogeny of Thottea 269
Fig. 3. Distribution and biodiversity similarity diagram for Thottea s.1. (Hou 1981) showing the
number of species known for the respective biogeographic regions (in circles) and the number
of species shared between them. Two diversity hotspots are observed: one in India (southern
Western Ghats) and one in the Malesian region.
This species has only been reported from Borneo previously. This finding requires
confirmation because 7. paucifida and T. tomentosa can be confused due to superficial
similarities. In addition, 7. paucifida from the Andaman Islands is known only with
fruits, whereas 7. paucifida from Borneo was known only with flowers — which could
exacerbate a comparison.
Thottea holds a potential key position with respect to growth form and
woodiness evolution in the Piperales, being a potential link between the herbaceous
Asaroideae and the woody Aristolochioideae. The genus Aristolochia is dominated by
vines or lianas, but rarer shrub-like species are known. Close relationships between
species having wide-ranging growth forms pose a number of questions concerning
the processes by which highly different growth forms have evolved. Analysis of the
developmental shifts in both primary and secondary development of the stem provides
an implicit framework for identifying which structural and anatomical traits are
adapted for life as herbs, shrub or lianas (Speck et al. 2003, Rowe & Speck 2005,
Isnard et al. 2011, Wagner et al. in prep.). Ongoing studies are investigating to what
extent heterochrony and relatively basic changes in developmental rate can radically
influence the growth form and how specialisation and/or canalisation of developmental
traits, play a role in modifying the overall size and growth form of species within the
Aristolochiaceae and Piperales.
270 Gard. Bull. Singapore 63(1 & 2) 2011
ACKNOWLEDGEMENTS. Financial support for this study came from SYNTHESYS, a
European Union-funded Integrated Activities grant, as well as from the German Research
Foundation (DFG)-funded project NE 681/11-1. Many Herbaria provided specimens on loan
for morphological study and/or molecular work, which is thankfully acknowledged (A, AAU,
BKF, BONN,: BRI, C, COL, DR, G, Ko KLAKLU, Ly MO; NYSPH. PSUS ORG. ST SING
TBGT, USJ, WU). We especially would like to thank the Nationaal Herbarium Nederland (L)
for kindly being the host of the first author during the Synthesys fellowship. We thank Michael
Stech, Marcel Eurlings, Ding Hou (NHN) and René Glas (University Leiden) for hosting and
support in Leiden, and last but not least Anna-Magdalena Barniske (TU Dresden) for many
valuable discussions.
References
Borsch, T., Hilu, K., Quandt, D., Wilde, V., Neinhuis, C. & Barthlott, W. (2003)
Noncoding plastid trn7-trnF sequences reveal a well resolved phylogeny of basal
angiosperms. J. Evol. Biol. 16: 558-576.
Chanderbali, A., Yoo, M.-J., Zahn, L., Brockington, S., Wall, K., Gitzendanner, M.,
Albert, V., Leebens-Mack, J., Altman, N., Ma, H., dePamphilis, C., Soltis, D.
& Soltis, P. (in press) Origin and evolution of the flower: conservation and
canalization of gene expression during Angiosperm diversification. PNAS.
Dan, M., Mathew, P.J., Unnithan, C.M. & Pushpangadan, P. (1996) Thottea abrahamii
a new species of Aristolochiaceae from Peninsular India. Kew Bull. 51: 179-182.
Drummond, A.J. & Rambaut, A. (2007) BEAST: Bayesian evolutionary analysis by
sampling trees. BMC Evol. Biol. 7: 214.
Duchartre, P.E.S. (1864) Aristolochiaceae. In: Candolle, A.P. de (ed), Prodromus
Systematis Naturalis Regni Vegetabilis 15: 421-498. Paris: Sumptibus Sociorum
Treuttel & Wiirtz.
Gonzalez, F. & Stevenson, D. (2002) A phylogenetic analysis of the subfamily
Aristolochioideae (Aristolochiaceae). Revista Acad. Colomb. Ci. Exact., Fisic.
Nat. 26: 25-60.
Hooker, J.D. (1890) The Flora of British India 5: 72-74. London: L. Reeve & Co.
Hou, D. (1981) Florae Malesianae Praecursores LXII On the genus Thottea
(Aristolochiaceae). Blumea 27: 301-332.
Huber, H. (1985) Samenmerkmale und Gliederung der Aristolochiaceen. Bot. Jahrb.
107: 277-320.
Isnard, S., Wagner, S., Frenzke, L., Prosperi, J., Wanke, S., Neinhuis, C. & Rowe, N.
(2011) Success and failure of growth forms diversification in basal angiosperms.
XVIII Intern. Botanical Congress Melbourne, Australia.
Kelchner, S.A. (2000) The evolution of non-coding chloroplast DNA and its application
in plant systematics. Ann. Missouri Bot. Gard. 87: 482-498.
Kelly, L. (1998) Phylogenetic relationships in Asarum (Aristolochiaceae) based on
morphology and ITS sequences. Amer. J. Bot. 85: 1454-1467.
Molecular phylogeny of Thottea 271
Kelly, L. & Gonzalez, F. (2003) Phylogenetic Relationships in Aristolochiaceae. Syst.
Bot. 28: 236-249.
Klotzsch, F. (1859) Die Aristolochiaceae des Berliner Herbariums, pp. 571—626, t.1.
Monatsberichte der Koniglichen Preusischen Akademie der Wissenschaften zu
Berlin.
Kumar, E., Khan, A. & Binu, S. (2000) A new species of Thottea Rottb.
(Aristolochiaceae) from Kerala, South India. Rheedea 10: 117-120.
Leins, P., Erbar, C. & Heel, W. van (1988) Note on the floral development of Thottea
(Aristolochiaceae). Blumea 33: 357-370.
Miiller, K. (2005a) SeqState — primer design and sequence statistics for phylogenetic
DNA data sets. Appl. Bioinformatics 4: 65-69.
Miller, K. (2005b) The efficiency of different search strategies in estimating parsimony
jackknife, bootstrap, and Bremer support. BMC Evol. Biol. 5: 58. (PRAP2)
Neinhuis, C., Wanke, S., Hilu, K., Miiller, K. & Borsch, T. (2005) Phylogeny of
Aristolochiaceae based on parsimony, likelihood, and Bayesian analyses of trnL-
trnF sequences. PI. Syst. Evol. 250: 7-26.
Nickrent, D., Blarer, A., Qiu, Y., Soltis, D., Soltis, P., & Zanis, M. (2002) Molecular
data place Hydnoraceae with Aristolochiaceae. Amer. J. Bot. 89: 1809-1817.
Nixon, K.C. (1999) The parsimony ratched: a rapid means for analyzing large data
sets. Cladistics 15: 407-414.
Ohi-Toma, T., Sugawara, T., Murata, H., Wanke, S., Neinhuis, C. & Murata, J. (2006)
Molecular phylogeny of Aristolochia sensu lato (Aristolochiaceae) based on
sequences of rhcL, matK, and phyA genes, with special reference to differentiation
of chromosome numbers. Syst. Bot. 31: 481-492.
Pandurangan, A. & Nair, V. (1993) Thottea idukkiana - anew species of Aristolochiaceae
from Kerala, India. J. Econ. Tax. Bot. 17: 465-467.
Posada, D. (2008) jModelTest: phylogenetic model averaging. Molec. Biol. Evol. 25:
1253-1256.
Ronquist, F. & Huelsenbeck, J.P. (2003) MrBayes 3: Bayesian phylogenetic inference
under mixed models. Bioinformatics 19: 1572-1574.
Rottbéll (1783) Thottea grandiflora. Det Kongelige Danske Videnskabers Selskabs
Skrifter 2: 529-531, t. 2.
Rowe, N. & Speck, T. (2005) Plant growth forms: an ecological and evolutionary
perspective. New Phytol. 166: 61—72.
Shaiju, P. & Omanakumari, N. (2009) Inflorescence morphology and systematics of
the genus Thottea Rottb. (Aristolochiaceae) from the Western Ghats, India. Syst.
Biodivers. 7: 445-451.
Shaiju, P. & Omanakumari, N. (2010) Floral morphology and systematics of the genus
Thottea Rottb. (Aristolochiaceae) from the Western Ghats, India. P/. Syst. Evol.
288: 213-225.
Simmons, M.P. & Ochoterena, H. (2000) Gaps as characters in sequence-based
phylogenetic analyses. Syst. Biol. 49: 369-381.
Sivarajan, V. (1985) A new species of Thottea (Aristolochiacheae) from India. P/. Syst.
Evol. 150: 201-204.
D2: Gard. Bull. Singapore 63(1 & 2) 2011
Sivarajan, V. & Babu, A. (1985) A new species of Thottea Rottb. (Aristolochiaceae)
with notes on the identity of Thottea siliquosa (Lam.) Ding Hou. Indian J. For.
8: 265-268.
Solereder H. (1894) Aristolochiaceae. In: Engler, A. & Prantl, K. (eds) Die natiirlichen
Pflanzenfamilien 3: 264-273.
Soltis P.S., Brockington, S.F., Yoo, M.-J., Piedrahita, A., Latvis, M., Moore, M.J.,
Chanderbali, A.S. & Soltis, D.E. (2009). Floral variation and floral genetics in
basal angiosperms. Amer. J. Bot. 96: 110-128.
Speck T., Rowe, N.P., Civeyrel, L., Classen-Bockhoff, R., Neinhuis, C. & Spatz, H.-C.
(2003) The potential of plant biomechanics in functional biology and systematics.
In: Stuessy, T.F., Mayer, V. & H6érandl, E. (eds) Deep Morphology: Toward a
Renaissance of Morphology in Plant Systematics, pp. 241—271.
Steele, K.P. & Vilgalys, R. (1994) Phylogenetic analyses of the Polemoniaceae using
nucleotide sequences of the plastid gene matK. Syst. Bot. 19: 126-142.
Sumathi, R., Maina, V. & Lakra, G. (2004) Thottea paucifida Ding Hou, family
Aristolochiaceae, a new record for India. J. Bombay Nat. Hist. Soc. 101: 195-196.
Swarupanandan, K. (1983) A new species of Thottea Rottb. (Aristolochiaceae) from
India. Blumea 28: 407-411.
Swofford D.L. (2002) PAUP*. Phylogenetic Analysis Using Parsimony (*and Other
Methods), version 4. Sunderland, Massachusetts: Sinauer Associates.
Wallace, A.R. (1859) On the zoological geography of the Malay Archipelago. In:
Alfred Russel Wallace, Classic Writings 2009, paper 4.
Wallace, A.R. (1863) On the physical geography of the Malay Archipelago. J. Roy.
Geogr. Soc. London 33: 217-234.
Wanke, S., Gonzalez, F. & Neinhuis, C. (2006a) Systematics of Pipevines: Combining
morphological and fast-evolving molecular characters to investigate the
relationships within subfamily Aristolochioideae (Aristolochiaceae). /ntern. J.
Pl. Seis lose 12151227.
Wanke, S., Samain, M.-S., Vanderschaeve, L., Mathieu, G., Goetghebeur, P. &
Neinhuis, C. (2006b) Phylogeny of the genus Peperomia (Piperaceae) inferred
from the trnK/matK Region (cpDNA). P/. Biol. 8: 93-102.
Wanke, S., Jaramillo, M., Borsch, T., Samain, M.-S., Quandt, D. & Neinhuis, C.
(2007a) Evolution of Piperales—matK gene and ¢rnK intron sequence data reveal
lineage specific resolution contrast. Molec. Phylogen. Evol. 42: 477-497.
Wanke, S., Vanderschaeve, L., Mathieu, G., Neinhuis, C., Goetghebeur, P. & Samain,
M.-S. (2007b) From forgotten taxon to a missing link? The genus Verhuellia
(Piperaceae) revealed by molecules. Ann. Bot. 99: 1231-1238.
Welzen, P.C. van, Slik, J.W.F. & Alahuhta, J. (2005) Plant distribution patterns and
plate tectonics in Malesia. Biol. Skr. 55: 199-217.
Wicke, S. & Quandt, D. (2009) Universal primers for the amplification of the plastid
trnK/matK region in land plants. Ann. Jard. Bot. Madrid 66: 285-288.
Woodruff, D.S. (2003) The location of the Indochinese-Sundaic biogeographic
transition in plants and birds. Nat. Hist. Bull. Siam Soc. 51: 97-108.
Molecular phylogeny of Thottea
Zhou, T.-H., Li, S., Qian, Z.-Q., Su, H.-L., Huang, Z.-H., Guo, Z.-G., Dai, P.-F., Liu,
Z.-L. & Zhao, G.-F. (2010) Strong phylogeographic pattern of cpDNA variation
reveals multiple glacial refugia for Sarwma henryi Oliv. (Aristolochiaceae), an
endangered herb endemic to China. Molec. Phylogen. Evol. 57: 176-188.
Permanent URL of the alignment and analyses data on TreeBASE:
http://purl.org/phylo/treebase/phylows/study/TB2:S11815
Appendix A. Field origin or Botanical Garden (BG) accession numbers, voucher information
and GenBank accessions used in the present study. * For these specimens sequences of the
trnK intron and matK gene were derived from earlier studies. The 7K 3° exon and #K-psbA
spacer were newly generated and submitted to genbank as an update of the original sequences.
Taxon Field ongin Voucher (herbarium) Genebank accession no.
Botanical Garden ;
tee ee tnK intron, Source
matK gene,
trnK-psbA
spacer
Saruma Oliv.
S. henrvi Oliv. BG Bonn, 02618 —_ Borsch 3456 DQ532033 Wanke et al. 2007a
(BONN) & this study*
Asarum L.
A. chingchengense BG Bonn, 02680 =Neinhuis 90 (DR) DQ882196 Wanke et al. 2007a
C.Y.Cheng & & this study*
C.S.Yang
Aristolochia L.
A. acuminata Lam. BG Bonn, 17417 Wanke & Neinhuis DQ532063 Wanke etal. 2007a
146 (DR) & this study*
A. acutifolia Duch. Colombia, Meta Gonzalez- DQ532048 Wanke et al. 2006a
4176 (COL) & this study*
A. bracteolata Lam. BG Bonn, 16714 _— Neinhuis 94 (DR) DQ532059 Wanke et al. 2007a
& this study*
A. clematitis L. Croatia. Is Hovik/ Starmiiller (KL) DQ532060 Wanke et al. 2006a
Asinello & this study*
A. gorgona Heredia: Puerto Blanco 1752 (USJ) DQ532051 Wanke etal. 2007a
M.A.Blanco Viejo de & this study*
Sarapiqui, Costa
Rica
A. holostylis BG Bonn, 02193 =Neinhuis 116(DR) DQ532057 Wanke etal. 2007a
F.Gonzalez & this study*
A. labiata Willd. BG Bonn, 09867 ~=Neinhuis 96 (DR) DQ532055 Wanke et al. 2007a
& this study*
274
Gard. Bull. Singapore 63(1 & 2) 2011
A. lindneri A.Berger Bolivia, San Jose Ibisch s.n. (DR) DQ532047 Wanke et al. 2006a
de Chiquitos & this study*
A. manshuriensis BG Bonn, 13085 Neinhuis 104(DR) DQ532040 Wanke et al. 2007a
Kom. & this study*
A. micrantha Duch. priv. coll. Neinhuis 103 (DR) DQ532046 Wanke et al. 2007a
B. Westlund & this study*
Texas, USA
A. panamensis Panama, Panama Gonzalez-4018B DQ532043 ~=Wanke et al. 2006a
Standl. (COB) & this study*
A. pistolochia L. France, Cassis, leg. Kreft, Wanke 37 DQ532062 Wanke et al. 2007a
Calenque d’En (DR 25372) & this study*
Veau
A. promissa Mast. BG Bonn, 13014 —Neinhuis 118(DR) =DQ532065 = Wanke et al. 2007a
& this study*
A. reticulata Nutt. priv. coll. Neinhuis 108 (DR) DQ532037 + Wanke et al. 2007a
B.Westlund & this study*
Texas, USA
A. rojasiana BG Miinchen s.n., Wanke s.n. (DR) DQ861635 Wanke et al. 2006a
(Chodat & Hassl.) Brazil, Mato & this study*
F.Gonzalez Grosso
A. rotunda L. France, Corsica Wanke 015 (DR) DQ532061 Wanke et al. 2006a
& this study*
A. serpentaria L. priv. coll. Neinhuis 112(DR) DQS532038 Wanke et al. 2007a
B. Westlund & this study*
Texas, USA
A. triactina Hook.f. BG Bonn, 12767 Neinhuis 119(DR) DQ532066 Wanke et al. 2007a
& this study*
Thottea Rottb.
T. abrahamii M.Dan, India, Kerala Pradeep s.n.(TBGT) JN415669 this study
P.J.Mathew, Trop. Bot. Garden
Unnithan & Pushp.
T. barberi (Gamble) India, Kerala Pradeep s.n.(TBGT) JN415675 this study
Ding Hou Trop. Bot. Garden
T. borneensis Valeton Hort.Bogor van Steenis 24294 JN415668 this study
XI1.B.XII1.134, (L 240977)
origin: Borneo
T. dalzellii (Hook.f.) | India, Kerala Pradeep s.n.(TBGT) JN415677 this study
Karthik. & Moorthy Trop. Bot. Garden
T. dinghoui India, Kerala Pradeep s.n.(TBGT) JN415670 this study
Swarupan. Trop. Bot. Garden
T. duchartrei Sivar., India, Kerala Pradeep s.n.(TBGT) JN415678 this study
A.Babu & Balach. Trop. Bot. Garden
Molecular phylogeny of Thottea
T. grandiflora Rottb.
T. idukkiana Pandur.
& V.J.Nair
T. parviflora Ridl.
T. penitilobata Ding
Hou
T. piperiformis
(Griff.) Mabb.
T. ponmudiana Sivar.
T. siliquosa (Lamkey)
Ding Hou
T. sivarajanii
E.S.S.Kumar,
A.E.S.Khan & Binu
Thottea sp.
T. tomentosa (Blume)
Ding Hou
Saururaceae Rich.
Anemopsis
californica (Nutt.)
Hook. & Arn.
Gymnotheca
chinensis Decne.
Saururus chinensis
Hort. ex Loudon
Peninsular
Malaysia,
Selangor, Genting
Sempah
India, Kerala
Trop. Bot. Garden
Thailand,
Songkhla
Province, Hat Yai
Borneo, Brunei,
Temburong,
Batu Apoi Forest
Reserve
Malaysia
India, Kerala
Trop. Bot. Garden
India, Kerala
Trop. Bot. Garden
India, Kerala
Trop. Bot. Garden
BG Bonn, 09037,
origin: India,
Kerala, Thrissur
District (Bogner
86-3421)
Thailand,
Phatthalung, Tha
Mot
BG Bonn, 06422
BG Bonn, 17072
BG Bonn, 00312
B.C. Stone 6112
(PH 0961499)
Pradeep s.n. (TBGT)
S.Chantanaorrapint
1265 (PSU)
Poulsen, A.D. |
(AAU)
Weber &
Anthonysamy
870519-1/1 (WU)
Pradeep s.n. (TBGT)
Pradeep s.n. (TBGT)
Pradeep s.n. (TBGT)
Neinhuis 121 (DR)
Larsen et al. 43958
(AAU)
Wanke 002 (DR)
Wanke 004 (DR)
Wanke 001 (DR)
JN415671
JN415680
JN415672
JN415673
DQ532036
JN415676
JN415679
JN415681
DQ532035
JN415674
DQ882198
DQ882199
DQ212713
Ds
this study
this study
this study
this study
Wanke et al. 2007a
& this study*
this study
this study
this study
Wanke et al. 2007a
& this study*
this study
Wanke et al. 2007a
& this study*
Wanke et al. 2007a
& this study*
Wanke et al. 2006b
& this study*
Gardens’ Bulletin Singapore 63(1 & 2): 277-286. 2011 PEAT)
Asian Begonia: out of Africa via the Himalayas?
S. Rajbhandary', M. Hughes’, T. Phutthai’,
D.C. Thomas? and Krishna K. Shrestha!
' Tribhuvan University, Institute of Science and Technology,
Central Department of Botany, Kirtipur, Kathmandu, Nepal
"Royal Botanic Garden Edinburgh, 20a Inverleith Row, Edinburgh, EH3 SLR, U.K.
m.hughes(@rbge.ac.uk (corresponding author)
*Princess Maha Chakri Sirindhorn Natural History Museum
& Centre for Biodiversity of Peninsular Thailand,
Department of Biology, Faculty of Science,
Prince of Songkla University, Hat Yai, Songkhla, Thailand 90112
*University of Hong Kong, School of Biological Sciences,
Pok Fu Lam Road, Hong Kong, PR China
ABSTRACT. The large genus Begonia began to diverge in Africa during the Oligocene. The
current hotspot of diversity for the genus in China and Southeast Asia must therefore be the
result of an eastward dispersal or migration across the Asian continent. To investigate the
role of the Himalayas as a mesic corridor facilitating this migration, we constructed a time-
calibrated molecular phylogeny using ITS sequence data. Himalayan species of Begonia were
found to fall into two groups. The first is an unresolved grade of tuberous, deciduous species
of unknown geographic origin, with evidence of endemic radiations in the Himalayan region
beginning c. 7.4 Ma, coinciding with the onset of the Asian monsoon. The second is a group
of evergreen rhizomatous species with a probable origin in China, which immigrated to the
Himalayan region c. 5.1 Ma, coinciding with an intensification of the monsoon. The hypothesis
of the Himalayas being a mesic migration route during the colonisation of Asia is not refuted,
but further data is needed.
Keywords. Begonia, biogeography, China, Himalayas, molecular phylogeny, southeast Asia
Introduction
The large genus Begonia has around 750 species in Asia, with the bulk occurring
in Southeast Asia, and the Malesian region having c. 440 of these (Hughes 2008).
Reconstructions of the phylogenetic history of Begonia show an early divergent
African clade, with Asian and American species nested within an African grade (Plana
et al. 2004, Forrest et al. 2005, Goodall-Copestake et al. 2010) and Asian (including
Socotran) species supported as monophyletic (Goodall-Copestake et al. 2010, Thomas
2010). Given the initial diversification of Begonia in Africa in the Oligocene or late
Eocene (Goodall-Copestake et al. 2009) and the monophyly of Asian species, there
must have been a migration or dispersal of an ancestor eastwards from Africa which
has speciated and led to the various diversity hotspots across Asia.
278 Gard. Bull. Singapore 63(1 & 2) 2011
Molecular divergence age estimates indicate that the origins of Asian
(including Socotran) Begonia date to 18-15 Ma (Thomas 2010) coinciding with the
mid-Miocene climatic optimum, a warm phase which led to the expansion of tropical
vegetation in Asia as far north as southern Japan and as far east as the northwest of the
Indian subcontinent (Zachos et al. 2001, Morley 2007). However it would seem that a
straightforward migration of Begonia across the Arabian peninsula to Asia 1s unlikely,
given the dry conditions thought to have existed across much of southwest Asia and
the Arabian peninsula at the time (Morley 2007). Hence a long distance dispersal event
does not seem an unlikely hypothesis for the entry of an African ancestor into Asia,
although this was possibly facilitated by the greater expanse of tropical climate during
the mid-Miocene. Begonia from Sri Lanka, the Western Ghats and Socotra are at the
western limit of the Asian Begonia clade (Fig. 1), and have been found to be early
divergent lineages within a phylogeny of Asian species (Thomas 2010).
The Himalayas began to form during the initial collision of the Asian and
Indian plates during the early Eocene c. 35 Ma (Ali & Aitchison 2008) and by the time
of the arrival of Begonia in Asia c. 15 Ma, the High Himalayas were present (Amano
& Taira 1992). Hence, depending on climate, there was potentially a significant area
of montane habitat at the north of the Indian continent which could have provided
a mesic link as a migration route eastwards towards the current Begonia diversity
hotspots of Indo-China and Malesia (Fig. 1).
To investigate the role of the Himalayas as an easterly migration route during
the colonisation of Asia, Himalayan species of Begonia were placed in a phylogeny
to allow their interpretation in a geographical and temporal context. This paper builds
on previous studies by considerably increasing the sampling of Himalayan species
and coding the region as a distinct geographic unit in a biogeographic analysis. In
particular, this paper addresses the following questions:
—Is there evidence for the Himalayas behaving as a link in an eastward migration of
Begonia?
—How has paleo-climatic change and the Himalayan orogeny influenced the evolution
of Nepalese Begonia?
Materials and methods
The molecular phylogenetic analysis was based on nuclear ribosomal internal transcribed
spacer (ITS) sequences, obtained from Genbank and also de novo for this study
(Appendix A) following the methods in Thomas (2010). Phylogenetic reconstruction,
divergence times and ancestral area reconstruction were performed simultaneously
using Bayesian inference as implemented in BEAST v1.6.1 (Drummond & Rambaut
2007), using 4 x 10’ generations and sampling every 1000th generation. The sequence
data was divided into three partitions, namely the two internal transcribed spacers
(ITS1 and ITS2) and the 5.8s ribosomal gene. Models of sequence evolution for each
nucleotide sequence partition were determined using jModelTest (Posada 2008). The
divergence time for the split between African and Asian species was set as 17.8 Ma
Onigin and divergence of Asian Begonia 279
Fig. 1. The distribution of Begonia in the Old World, highlighting the disjunction between the
African and Asian floras, and the westward linear distribution along the Himalayas. Data from
the GBIF data portal and Hughes & Pullan (2007).
(Thomas 2010), implemented as a normally distributed prior with a standard deviation
of 3.4. Bayesian ancestral area reconstructions were performed using the continuous-
time Markov chain (CTMC) model for discretized diffusion specified by Lemey et al.
(2009). considering diffusion amongst 10 geographic areas defined as Africa, China,
Himalaya, India, Indo-China, Philippines, Socotra, Sulawesi, Sunda Shelf and Taiwan.
Maximum clade credibility consensus trees were made using TreeAnnotator v1.6.1
(Drummond & Rambaut 2007) with a burn-in of 25%, and visualised using FigTree
v1.3.1 (Rambaut 2007).
The Bioclim parameter precipitation seasonality (Busby 1991) was extracted
from the WorldClim database (Hijmans et al. 2005a) using DIVA-GIS (Hijmans et al.
2005b) and georeferenced herbarium specimen data from Hughes & Pullan (2007). It
is calculated as the standard deviation of weekly precipitation estimates expressed as
a percentage of the annual mean of those estimates. Scores for species with multiple
localities were averaged.
Perennation organ type was scored from literature records and herbarium
specimens available from Hughes & Pullan (2007).
280 Gard. Bull. Singapore 63(1 & 2) 2011
Results
Sequence characterisation
The aligned sequences gave a dataset of 873 characters in length (ITS1 330 bases,
5.8s 148 bases, ITS2 395 bases). The bases at positions 492-499 and 505-525 were
excluded from the analysis due to difficulty in reliably assigning positional homology.
The remaining 845 bases were included 1n the analysis, of which 561 were variable and
433 (51%) were potentially informative (ITS1, 237 variable and 190 (58%) potentially
informative; 5.8s, 31 variable and 11 (7%) potentially informative; ITS2, 294 variable
and 232 (63%) potentially informative). All three regions were analysed under a
GTR+G+I model (general time reversible model, gamma distributed rate variation,
plus a proportion of invariable sites).
Phylogenetic analysis—geography and dating
The Asian + Socotra ingroup was supported as monophyletic with a posterior
probability of | (Fig. 2). Relationships at the base of the tree were generally poorly
supported, and hence inferences regarding timing and geography of cladogenesis on
this part of the tree are not possible with confidence.
Clade A contains species from China, Indo China and Malesia, with several
well supported sub-clades. One of these contains the Thai species B. alicida and
B. smithiae, together with three species of Begonia sect. Parvibegonia. Two other
supported subclades are highlighted (Fig. 2) concerning the large Begonia sect.
Petermannia with a probable origin on the Sunda Shelf in the late Miocene followed
by later diversification in Wallacea.
Himalayan species are present only in Clade B (Fig. 2), in both the ‘Diploclintum
grade’ and the clade consisting of members of Begonia sect. Platycentrum and
Sphenanthera, marked PLA-SPH in Fig. 2. The ‘Diploclinitum grade’ has a largely
unsupported topology at the base, but began diversifying sometime between 14.7
(+7.6) and 12.3 (+6.7) Ma in the mid-Miocene. Species from the Himalayan region
in this grade are intermixed with species from China, Indo-China and Taiwan. Due
to the unresolved backbone, the area of origin for this tuberous, seasonally adapted
grade cannot be reconstructed. Supported subclades within the grade have areas
of origin reconstructed as the Himalaya, Indo-China or China. One consists of the
Taiwan endemic B. ravenii sister to two accessions of the Himalayan B. dioica; the
geographic origin of this clade is equivocal, with the Himalayas having the highest
posterior probability, though only of 0.37; it dates from 7.4 (+5.7) Ma. Another clade
consists of four species from Thailand, probably representing speciation originating
in that country (PP 0.84), dating from 8.1 (+5.2) Ma. A clade of species from Nepal,
with B. picta appearing as sister to the remaining species, likely has a Himalayan
origin (PP 0.90) dating 7.4 (+4.6) Ma. A clade with Begonia puttii (Thailand) and B.
labordei (China) along with an unidentified species from China diverged 6.4 (+4.8)
Ma in China (PP 0.82). The most highly nested subclade in the Diploclinium grade,
sister to the PLA-SPH clade, contains three accessions of Begonia rubella from Nepal
Origin and divergence of Asian Begonia 281
AUG B canoes SAfrica
PAR B.tenuifolia Bali
PAR B.vanabilis Vietnam
PAR B.variabilis Thailand
RE! B.muricata Java
PEL B.samhaensis Socotra
PEL B.socotrana Socotra
DIP B.chioroneura Phillipines
DIP B.hernandioides Phillipines
DIP B suborbiculata Palawan
DIP B.cleopatrae Palawan
PET B.chlorosticta Borneo
PET B.laruei Sumatra
PET B.burbidgei Borneo
PET B.amphioxus Borneo
PET B.malachosticta Borneo
PET B.macintyreana Sulawesi
PET PoaEer cae Sulawesi
PET B.incisa Philippines
PET B.ramosii Philippines
IGN B.demissa Thailand
IGN B-rabillii Thailand
DIP Bdiscreta Thailand
DIP B.aceroides Thailand
B.picta2 Nepal
DIP B bicta Nepal
DIP dees wise
i]
Uv
@
3
Ss
Q
g
o
z
2
B
apes6 wniuijoojdiq
PLA B.alpina Pen Malaysia
PLA B.pavonina Pen jaysia
SPH B robusta Java
SPH B.multangula Java
SPH B.balansana Vietnam
PLA B.versicolor China
SPH B.cerat China
DIP B.setifolia China
DIP Bwenshanensis China
DIP B.ruboides China
PLA B.longicarpa China
PLA B.cathayana China
SPH B:silletensis China
SEE ean cme
tera Sulawesi
Attica China SPH B acetosella Thailand
F SPH B.Jongifolia Taiwan
Taiwan Himalaya SPH B.longifolia Sumatra
MON B.nepalensis Nepal
SPH B.halconensis_ Phillipines
Sulawesi Indo-China DIP Biaiwaniana Taiwan
SPH B.roxburghii Nepal
Philippines Sunda Shelf
PLA B.xanthina India
PLA B.rex India
PLA B.barbata India
PLA B.sikkimensis Soh
PLA B.teysmanniana_ Sumatra
PLA B.hemsleyana China
PLA B.circumlobata_ China
PLA B.formosana Taiwan
B.hatacoa2 Nepal
PLA B.taligeral Nepal
PLA B.taligera2 Nepal
PLA B.baviensis Indo China
PLA B.panchtharensis Nepal
PLA B.annulata2 Nepal
PLA B.annulata Nepal
PLA B.nuwakotensis Nepal
Perenniation organs
Tuber
Rhizome
None
ORSeReee:
15.0 12.5 10.0 7.5 5.0 25 0.0 Procpitetion soaponaily
Fig. 2. Maximum clade credibility chronogram derived from an analysis of nuclear ribosomal
ITS sequences using BEAST v1.6.1 (Drummond & Rambaut 2007). Asterisks denote
clades with a posterior probability of less than 0.8; pie charts show Bayesian ancestral
area reconstructions on supported nodes; lower scale denotes time in millions of years; the
monochrome and coloured boxes at the branch tips show the perenniation organ type and
precipitation seasonality respectively; missing data coded as white. Three-letter codes indicate
the sections of Begonia (ALI, sect. Alicida; AUG, sect. Augustia; COE, sect. Coelocentrum:
DIP, sect. Diploclinium; HAA, sect. Haagea; IGN, ignota (section unknown); MON, sect.
Monopteron; PAR, sect. Parvibegonia; PET, sect. Petermannia; PLA, sect. Platyvcentrum; REI,
sect. Reichenheimea; RID, sect. Ridleyella, SPH, sect. Sphenanthera).
282 Gard. Bull. Singapore 63(1 & 2) 2011
and one of B. tribenensis, also from Nepal. This subclade probably has an origin in the
Himalayas (PP 0.95) dated 6.4 (+4.3) Ma.
The origin of the PLA-SPH clade dates from 6.5 (+4) Ma and hence is
considerably younger than the ‘Diploclinium grade’. Himalayan species in the PLA-
SPH clade are intermixed with species from China, Indo-China, the Sunda shelf,
Taiwan, the Philippines and Sulawesi. The geographic area of origin for this clade is
most probably China (PP 0.73). Three supported subclades within the PLA-SPH clade
also have an area of origin optimised as China, with posterior probabilities of 0.65,
0.70 and 0.94. Only one subclade has a probable area of origin in the Himalaya (PP
0.99), dating from 3.0 (+2.3) Ma and with one dispersal into Indo-China. The overall
picture for the PLA-SPH clade is of a fairly rapid diversification with its origins in
China following loss of the tuberous habit, with dispersal to the Himalayan region and
throughout Southeast Asia followed by localised speciation.
Phylogenetic analysis—perennation organs, precipitation and seasonality
All of the species in the ‘Diploclinium grade’ are tuberous and the vast majority are
found in climates with a marked seasonality in rainfall (Fig. 2). They tend to die back
and lose their leaves completely in the dry season. Species within the PLA-SPH clade
either have a rhizome or no specialised perennation organs. The rainfall seasonality
of the distributions of most of the species is much less marked than for those in the
‘Diploclinium grade’, and they do not lose their leaves during the dry season. Only
one clade lacks perennation organs entirely, consisting of Begonia sect. Petermannia
species distributed in the largely everwet Malesian region.
Discussion
Although support for the backbone of the tree was weak, the taxon composition
of the two major clades, A and B, matches that of strongly supported clades found
using chloroplast sequence data by Thomas (2010), with much greater sampling of
Himalayan species in this study. In addition, the overall topology of clade B matches
the results of Thomas (2010) with respect to a clade of Begonia sects. Platycentrum
and Sphenanthera (PLA-SPH in Fig. 2) being nested within a grade consisting mainly
of Begonia sect. Diploclinium species from the continent (labelled ‘Diploclinium
grade’ in Fig. 2); also congruent is the Chinese species B. grandis being sister to the
remaining taxa. The basal relationships in Clade A are unsupported, but the supported
subclades within Clade A are similar to those found by Thomas (2010), and generally
correspond to species of a single section or geographic region. This congruence gives
some confidence in the tree topology, despite the disappointing levels of support.
The Himalayas as a link in an eastward Begonia migration
If the Himalayas acted as a migration corridor in the early evolution of Asian Begonia,
we would expect to see other species in Asia nested within a Himalayan grade in the
phylogeny. Due to the unsupported nodes at the base of the clade containing all the
Origin and divergence of Asian Begonia 283
Himalayan species, evidence for this is equivocal. Of the two major sections in the
Himalayan Begonia flora (Begonia sects. Diploclinium and Platycentrum), Begonia
sect. Diploclinium has the oldest lineages which date to between 14.7 (=7.6) and 12.3
(+6.7) Ma and accounts for nearly 60% of the species in Nepal. Some subclades in
this group do show a likely Himalayan origin, but the supported nodes are dated in the
late Miocene at c. 7.5 Ma and so post-date the early divergence of the section. Further
work on other gene regions is needed to resolve the relationships in the *‘Diploclintum
grade’. The hypothesis of basally branching Himalayan lineages in Begonia sect.
Diploclinium is not refuted by our results. However, the fact that the Chinese species
B. grandis is possibly sister to the rest of the ‘Diploclinium grade’ is tantalising and
raises the possibility of China being an ancestral area for this group.
The other major section, Begonia sect. Platycentrum, accounts for 35% of
Nepalese species. The species are all included in the PLA-SPH clade with an origin
in the late Miocene—early Pliocene. There is no evidence for the Himalayan species
in Begonia sect. Platycentrum being relicts from an eastward migration early in the
evolution of Asian Begonia, in fact the emerging picture is the exact opposite. This
clade has a probable Chinese origin, and represents an entry of the genus into the
Himalayan region separate from those in the ‘Diploclinium grade’. The Himalayan
species are largely highly nested within this clade and there is evidence for more
than one migration into the Himalayan region, including a very recent one in the late
Pliocene—early Pleistocene (bottom of Fig 2; clade from B. hatacoa to B. nuwakotensis).
Other major clades containing the bulk of species diversity in Southeast
Asia (Begonia sect. Petermannia from Malesia; Begonia sect. Diploclinium from the
Philippines: Begonia sect. Reichenheimia from the Sunda shelf) remain unsupported
in Clade A. Our results do not show any affinity of these clades with Himalayan taxa,
and the origin of these large radiations remains enigmatic.
Paleo-climatic change, Himalayan orogeny and evolution of Nepalese Begonia
It is possible to highlight two main events in the evolution of the Himalayan Begonia
flora—the mid- to late Miocene diversification of the tuberous, seasonally adapted
*Diploclinium grade’ and the late Miocene—Pliocene immigration and diversification
of the rhizomatous and evergreen PLA-SPH clade.
The early diversification of tuberous clades with a likely origin in the
Himalayas can be dated to between 6.4 (+4.8) and 7.4 (+4.6) Ma. This coincides
with the development of the Asian monsoon 7.4 Ma (Copeland 1997), due to the
Tibetan plateau having reached sufficient altitudes to affect a major re-organisation
of atmospheric circulation over the Asian continent (Zheng et al. 2004). The tuberous
species would be well adapted to this seasonal monsoon climate, with concentrated
periods of intense rainfall interspersed with significant dry seasons. These species are
deciduous, and their flowering periods are strongly constrained to the three months
after the onset of the monsoon (Rajbhandary et al. 2010). The subsequently formed
dry dehiscent capsular fruits are then able to disperse their seeds either passively or
through wind assistance during the following dry season.
284 Gard. Bull. Singapore 63(1 & 2) 2011
The immigration of members of the PLA-SPH clade into the Himalayan region
began 5.1 (+3.2) Ma, with an endemic radiation starting 3.0 (+2.1) Ma and continuing
throughout the Pleistocene. The immigration coincides with a further intensification of
the Asian monsoon around 5 to 3.6 Ma, possibly linked to a further surge in the uplift
of the Himalayas and the Tibetan plateau (Zheng et al. 2004, Zhisheng et al. 2001)
and changes in ocean currents in the Indo-Pacific region (Srinivasan & Sinha 2000).
The evolution of an evergreen, rhizomatous habit in this clade suggests adaptations to
wetter conditions, but the details of how the Pliocene monsoon intensification affected
total and seasonal rainfall in the Chinese (presumably SW China) region where the
PLA-SPH clade originated are unknown. The Himalayan species in this clade currently
occur in areas with similar rainfall seasonality to the tuberous species, but occupy
different micro habitats where water is more constantly available; their flowering is
not strongly constrained by the monsoon (Rajbhandary et al. 2010). They have fruits
which are adapted to rain-splash seed dispersal, and hence depend on either rainfall
or drips and splashes from waterfalls to disperse. Our results are congruent with those
of Tebbitt et al. (2006), who suggested that members of Begonia sect. Platyvcentrum,
which have fruit morphologies indicative of rain dispersal, evolved from wind-
dispersed Asian taxa following the colonisation of wetter habitats.
Of the two groups of Begonia in the Himalayas, seasonally deciduous
and evergreen, only the former could possibly be relicts of an eastward migration.
However, due to the unresolved relationships of clades within this group, their
origins remain unknown. The evergreen species represent a re-entry to the Himalayan
region, most likely from China, and have speciated in response to further tectonic
uplift and Pliocene—Pleistocene climatic cycles and changes in the monsoon intensity.
These factors are also likely to have strongly influenced the recent diversification of
other elements of the Himalayan flora. Further phylogenetic studies of Himalayan
plants will reveal the relative contribution of relict clades and recent speciation to the
considerable plant diversity in the region.
ACKNOWLEDGEMENTS. We are grateful to the M.L. MacIntyre Trust and the University
Grants Commission (UGC) Nepal for funding, to Ching-I Peng for providing DNA sequence
data, and to Michelle Hollingsworth, Alex Clarke and Ruth McGregor for assistance in the lab.
This research was supported by the Scottish Government’s Rural and Environment Research
and Analysis Directorate.
References
Ali, J.R. & Aitchison, J.C. (2008) Gondwana to Asia: Plate tectonics, paleogeography
and the biological connectivity of the Indian sub-continent from the Middle
Jurassic through latest Eocene (166-35 Ma). Earth Sci. Rev. 88: 145—166.
Amano, K. & Taira, A. (1992) Two-phase uplift of Higher Himalayas since 17 Ma.
Geology 20: 391-394.
Origin and divergence of Asian Begonia 285
Busby, J.R. (1991) BIOCLIM - A Bioclimatic Analysis and Prediction System. In:
Margules, C.R. & Austin, M.P. (eds) Nature Conservation: Cost Effective
Biological Surveys and Data Analysis, pp. 64—68. Canberra: CSIRO.
Copeland, P. (1997) The when and where of the growth of the Himalaya and the
Tibetan Plateau. In: Ruddiman, W.F. (ed) Tectonic Uplift and Climate Change.
New York: Plenum Press.
Drummond, A.J. & Rambaut, A. (2007) BEAST: Bayesian evolutionary analysis by
sampling trees. BMC Evol. Biol. 7: 214.
Forrest, L.L., Hughes, M. & Hollingsworth, P.M. (2005) A phylogeny of Begonia
using nuclear ribosomal sequence data and morphological characters. Syst. Bot.
30: 671-682.
Goodall-Copestake, W.P., Harris, D.J. & Hollingsworth, P.M. (2009) The origin of
a mega-diverse genus: dating Begonia (Begoniaceae) using alternative datasets,
calibrations and relaxed clock methods. Bot. J. Linn. Soc. 159: 363-380.
Goodall-Copestake, W.P., Pérez-Espona, S., Harris, D. & Hollingsworth, P.M. (2010)
The early evolution of the mega-diverse genus Begonia (Begoniaceae) inferred
from organelle DNA phylogenies. Biol. J. Linn. Soc. 101: 243-250.
Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G. & Jarvis, A. (2005a) Very high
resolution interpolated climate surfaces for global land areas. /ntern. J. Climatol.
25: 1965-1978.
Hiymans R.J., Guarino, L., Mathur, P. & Jarvis, A. (2005b) DIVA-GIS Version 5.0.
http://www.diva-gis.org/
Hughes, M. (2008) An Annotated Checklist of Southeast Asian Begonia. Edinburgh:
Royal Botanic Garden Edinburgh.
Hughes, M. & Pullan, M. (2007) Southeast Asian Begonia Database. Electronic
publication accessible via www.rbge.org.uk.
Lemey, P., Rambaut, A., Drummond, A.J., & Suchard, M.A. (2009) Bayesian
phylogeography finds its roots. PLoS Computat. Biol. 5.
Morley, R.J. (2007) Cretaceous and Tertiary climate change and the past distribution of
megathermal rainforests. In: Flenley, J.R., & Bush, M.B. (eds) Tropical Rainforest
Responses to Climatic Changes, pp. 1-31. Berlin: Springer.
Plana, V., Gascoine, A., Forrest, L.L., Harris, D. & Pennington, R.T. (2004) Pleistocene
and pre-Pleistocene Begonia speciation in Africa. Molec. Phylogen. Evol. 31:
449-461.
Posada, D. (2008) jModelTest: Phylogenetic model averaging. Molec. Biol. Evol. 25:
1253-1256.
Rajbhandary, S., Shrestha, K.K. & Hughes, M. (2010) Distribution patterns of Begonia
species in the Nepal Himalaya. Botanica Orientalis 7: 73-78.
Rambaut, A. (2007) FigTree, software package downloaded from http://tree.bio.ed.ac.
uk/software/figtree/
Srinivasan, M.S. & Sinha, D.K. (2000) Ocean circulation in the tropical Indo-Pacific
during early Pliocene (5.64.2 Ma): Paleobiogeographic and isotopic evidence.
Proc. Indian Acad. Sci., Earth Planet. Sci. 109: 315-328.
286 Gard. Bull. Singapore 63(1 & 2) 2011
Tebbitt, M.C., Forrest, L.L., Santoriello, A., Clement, W. & Swensen, S.M. (2006)
Phylogenetic relationships of Asian Begonia, with an emphasis on the evolution
of rain-ballist and animal dispersal mechanisms in sections Platycentrum,
Sphenanthera and Leprosae. Syst. Bot. 31:327-336.
Thomas, D.C. (2010) Phylogenetics and Historical Biogeography of Southeast Asian
Begonia (Begoniaceae). U.K.: University of Glasgow.
Zachos, J., Pagani, M., Sloan, L., Thomas, E. & Billups, K. (2001) Trends, rhythms
and aberrations in global climate 65 Ma to present. Science 292: 686-693.
Zheng, H., Powell, C. McA., Rea, D.K., Wang, J. & Wang, P. (2004) Late Miocene and
mid-Pliocene enhancement of the East Asian monsoon as viewed from the land
and sea. Global Planet Change 41: 147-255.
Zhisheng, A., Kutzbach, J.E., Prell, W.L. & Porter, S.C. (2001) Evolution of Asian
monsoons and phased uplift of the Himalaya-Tibetan plateau since late Miocene
times. Nature 411: 62-66.
Appendix A. List of the Genbank accession numbers of the DNA sequences used in the
phylogenetic analysis.
B. alpina, AY 753717; B. alveolata, AY048977; B. amphioxus, AF485150; B. annulata, HQ729060;
B. annulata2, HQ729059; B. aptera, AF48510; B. atricha, HQ729047; B. balansana, AF485091;
B. barbata, AF485095; B. baviensis, AF485087; B. brevipes, HQ729048; B. brvophila, HQ729030;
B. burbidgei, HQ729049; B. cathayana, AF280106; B. ceratocarpa, AY 048978; B. chiasmogyna,
HQ729050: B. chloroneura, AF485134 ; B. chlorostricta, AF485153; B. circumlobata, AY753721;
B. cleopatrae, AF485133; B. crispipila, HQ729051; B. demissa, HQ729026; B. dioical, HQ729039;
B. dioica2, HQ729038; B. dipetala, AF485142; B. discreta, HQ729024; B. dregei, AY429336;
B. flagellaris, HQ729031; B.formosana, AF485119; B. geranioides, AF469120; B. goegoensis,
AF485138; B. grandis, AF485089; B. grandis2, AF485088; B. halconensis, AF485106; B. handelii,
AY 048982; B. hatacoa2, AF485111; B. hemsleyana, AF485099; B.hernandioides, AF485135; B.
hyatae, AJ287262; B. incisa, AF485148; B. josephii2, HQ729037; B. kingiana, AF485139; B.
koordersii, HQ729052; B. labordei, AF485122; B. laruei, HQ729058; B. leprosa, AY 753722; B.
leptoptera, HQ729036; B. leptoptera2, HQ729033; B. leptoptera3, HQ729034; B. leptopterad,
HQ729035; B. longicarpa, AF485109; B. longifolia, AF485105; B. macintvreana, HQ729054;
B. malabarica, AF468141; B. malachosticta, AF485156; B. masoniana, AF485123; B. morsei,
AF485130; B. multangula, AY753724; B. muricata, AY753725; B. negrosensis, HQ729055;
B. nepalensis, AY753726; B. nuwakotensis, HQ729061; B. ovatifolia, HQ729032; B. palmata,
AF485113.1; B. palmata, AF485115.1; B. panchtharensis, HQ729062; B. parvula, GU176066; B.
pasamanensis, HQ729070; B. pavonina, AY753727; B. picta, HQ729042; B. picta2, HQ729041;
B. pseudolateralis, HQ729053; B. putii, HQ729025; B. rabillii, HQ729027; B. rajah, AF485136;
B. ramosii, HQ729057; B. ravenii, HQ729040; B. rex, AY753728; B. robusta, AY753729; B.
roxburghii, AF485092; B. rubella, AF485112; B. rubella2, HQ729043; B. rubella5, HQ729044;
B. ruboides, AY 048987; B. samhaensis, AF469122; B. scottii, HQ729063; B. setifolia, AY048990;
B. sikkimensis, HQ729064; B. silletensis, AF485094; B. sinwaensis, HQ729029; B. smithiae,
HQ729065; B. socotrana, AF469121; B. suborbiculata, HQ729069; B. taiwaniana, AY753731;
B. taligeral, HQ729066; B. taligera2, HQ729067; B. tenuifolia, HQ873478; B. teysmanniana,
HQ729068; B. tribenensis, HQ729045; B. variabilis, AY753732; B. variabilis, HQ729046; B.
varipeltata, HQ729056; B. versicolor, AF485090; B. wenshanensis, AY048974; B. xanthina,
AY 753733; Begonia sp. (China), GU176063; Begonia sp. (Nepal), HQ729028.
Gardens’ Bulletin Singapore 63(1 & 2): 287-298. 2011 287
Distribution patterns in Malesian Callicarpa (Lamiaceae)
Gemma L.C. Bramley
Herbarium, Library, Art and Archives,
Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3AE, U.K.
g.bramley@kew.org
ABSTRACT. A revision of the 55 species of Callicarpa L. (Lamiaceae) in Malesia is almost
complete. There appear to be two major centres of diversity, in terms of species numbers:
Borneo has 23 (44%) of the species (Bramley 2009), with 19 (83%) endemic; the Philippines
has 26 (50%) of the species of which 16 (61%) are endemic (Bramley, in press, a). Callicarpa
species have an extensive variation in distribution patterns; this paper focuses on the Pan-
Malesian species, and the species of Borneo and the Philippines, the two islands / island groups
that are the centre of Callicarpa species diversity. Fifteen of the 19 Callicarpa species endemic
to Borneo belong to the “Geunsia’ group, an informal group used here to recognise Callicarpa
pentandra and its relatives. The Geunsia group appears to be restricted to Malesia, and is
only represented by C. pentandra outside of Borneo, the Philippines and Sulawesi. The 16
Callicarpa species endemic to the islands of the Philippines represent a number of different
informal morphology-based groups containing species from other areas of Malesia, China or
Indo-China, or, they do not appear to belong to any particular group.
Keywords. Borneo, Callicarpa, centres of diversity, distribution, endemism, Malesia,
Philippines
Introduction
Arevision of Callicarpa L. (Lamiaceae) in Malesia is almost complete. It will form part
of an addition and update (Bramley et al., in prep.) to Keng’s Flora Malesiana account
of the Labiatae (1968). Callicarpa and a number of other genera in the Verbenaceae,
including Vitex L., Premna L. and Clerodendrum L., have been transferred to the
Lamiaceae (Cantino et al. 1992, Harley et al. 2004) based on morphological characters,
especially cymose versus racemose inflorescence, and a tubular and bilabiate versus
salverform corolla; these characters are supported by embryological and pollen
characters, and corroborated by analyses of cpDNA sequences (e.g., Olmstead et al.
2001).
There are about 140 species of Callicarpa, occurring in both temperate and
tropical regions, although this number may be inflated due to currently unrecognised
synonyms. In the New World there are 33 species recognised, particularly in the
Caribbean Islands (24 species currently recognised on Cuba). The genus is more
species rich in the Old World, with one species in Madagascar, c. 48 species in
Temperate Asia, particularly China, 55 in Malesia, seven in Australia, and three in
the Pacific. In general, the common and well-known species are shrubs or small trees
288 Gard. Bull. Singapore 63(1 & 2) 2011
found in disturbed areas such as secondary forest or roadsides, but a number of lesser
known species inhabit primary forest only. Due to their use in horticulture, Ca/licarpa
americana L., C. japonica Thunb., C. dichotoma (Lour.) K.Koch and C. bodinieri var.
giraldii (Hesse ex Rehder) Rehder are widespread.
The 55 Malesian species of Callicarpa are unevenly distributed across the
region (Fig. 1). Only two species have a pan-Malesian distribution, being present on
all islands and groups of islands: Callicarpa longifolia Lam. and Callicarpa pentandra
Roxb. There appear to be two major centres of diversity, in terms of species numbers:
Borneo has 23 (44%) of the species (Bramley 2009), with 19 (83%) endemic; the
Philippines has 26 (50%) of the species of which 16 (61%) are endemic (Bramley, in
press, a). Less species rich 1s the Thai-Malay Peninsula with a total of eight species, of
which one is endemic. Sumatra and Java share seven of these species, neither island
has any endemic species. Sulawesi has six species currently recognised, one of which
is endemic, and three further endemic species are being described as new to science
(Bramley, in press, b). The Lesser Sunda Islands have only three species, and the
Moluccas four species, none of which are endemic; New Guinea 1s home to seven
species.
¢ |Total number of Callicarpa taxa in Malesia = 55
Shown in boxes: Endemic species/Total species
Kilometers
0 250500 1,000 1,500 2,000
Fig. 1. Map showing the numbers of Callicarpa species in each island / island group across
the Malesian region. In each box the number to the left represents the number of Ca/licarpa
species endemic to that area; the number to the right is the total number of Ca/licarpa species
found in that area.
Malesian Callicarpa 289
Understanding why some species are widespread and others endemic to
particular areas is becoming especially important in this era of habitat destruction
and climate change. If we are to meet the targets set out by the Global Strategy for
Plant Conservation, there must be an increase in the rate of production of species
level conservation assessments (Nic Lughadha et al. 2005). Now that the taxonomy of
Malesian Callicarpa has been studied, and species delimitation clarified, it is possible
to map the distribution of each species based on data from Herbarium specimens.
Not only does this allow preliminary conservation assessments to be undertaken,
it encourages thought on what morphological characters might facilitate species to
inhabit a particular environment, or enable a species to be widespread and apparently
successful in more varied environments (Graham et al. 2004). Thoughts on the latter
may be useful in any attempts to restore vegetation in disturbed areas.
Malesia has a rich geological history and many studies have focused on
distributions of plant families and genera within it, with less of a focus on species.
Divisions between areas have been postulated (e.g., Wallace’s line), and frontiers
marked between different floristic regions (e.g. van Steenis 1950; van Welzen &
Slik 2009). By examining the distributions of Malesian Callicarpa species, | wish to
determine whether there are any common and easy to define patterns that represent
groups linked by morphological characters and geography. In this way I might be able
to suggest key characters that enable survival or speciation in particular environments.
I will make brief comparisons between patterns within Ca/licarpa and other Lamiaceae
genera in Malesia. There will be no formal analysis: following the rationale of Baker et
al. (1998), this paper is a descriptive precursor to any future analytical biogeographical
project.
Materials and methods
Label information from specimens of Malesian Callicarpa from BM, BO, C, E, GH,
K, KEP, L, NY, SING, SNP, US (abbreviations following Index Herbariorum, Thiers
[continuously updated]) was captured in a Microsoft Access database; collecting
localities were georeferenced if latitudes and longitudes were not already provided on
the specimen label. The online gazetteer GEOnet Names Server (http://earth-info.nga.
mil/gns/html/index.html) was used as a source for place names as well as the Google
search engine (www.google.com) and printed maps.
Specimen data was exported from Access into ArcView 3.3, and each species
distribution plotted using the Conservation Assessment Tools (CATS) extension
developed at RBG Kew (Moat 2007).
Distribution patterns, species groups and phytogeographical relationships
Callicarpa species show extensive variation in distribution patterns and each pattern
is not described in detail here. Instead I focus on three groups of particular interest:
290 Gard. Bull. Singapore 63(1 & 2) 2011
the Pan-Malesian species, because only this distribution is surprisingly rare, and the
Bornean and Philippine species, because these two islands / island groups are the
centres of diversity. For a quick reference to the distribution of individual taxa, see
Table 1.
Table 1. Malesian Callicarpa species and the areas in which they occur (excluding three new
species from Sulawesi, Bramley, in press, b). (Left to right) THA = Thailand; MLY = Peninsular
Malaysia; SUM = Sumatra; BOR= Borneo; PHI = Philippines; JAW = Java; SUL = Sulawesi;
LSI = Lesser Sunda Islands; MOL = Moluccas; NWG = New Guinea; AUS = Australia. Other
Areas: C = China, SC = South China, PAC = Pacific, T = Taiwan, V = Vietnam.
CallneeiDe THA MLY SUM BOR PHI JAW SUL LSI MOL NwG Aus Oe
species areas
albidotomentella x
angusta X ?V
angustifolia xX x x Vv
anomala XG
apoensis
arborea x Xe x oy di x X
argentil
badipilosa
barbata
basilanensis %
bicolor = x x x PAC
candicans X x x x x x x x C
caudata x x x x
cinnamoea x
clemensorum xX
coriacea x
denticulata x
dolichophylla % IEXC
endertii Xx
erioclona X Xx X Xx Xx X x Vv
fasciculiflora Sry nce i. Xx
flavida XK
fulvohirsuta X
furfuracea x Xx X X
glabrifolia
havilandii
hispida
involucrata
kinabaluensis
~ Ke KM mR |
longifolia x x x
Malesian Callicarpa 291
longipetiolata x
magnifolia X
maingayil xX XxX
micrantha X
pachyclada
paloensis Xx
pedunculata SC,
pentandra X X X X
platyphylla
plumosa
Pata a Ee a
ramiflora
rubella x x Xx Vac
saccata
scandens
Stapfii
subaequalis
subalbida X
subglandulosa
s~ | OM
subintegra
superposita x
surigaensis x
teneriflora X
Pan-Malesian species
Only two species have a distribution that includes all areas within Malesia: Callicarpa
pentandra Roxb., and C. longifolia Lam. (Fig. 2).
Callicarpa pentandra is a small tree commonly found along roadsides and in
secondary forest. Its distribution extends to Thailand in the North and to New Britain
in the East, it is possibly present in the Solomon Islands but the material from the
area has not yet been examined closely. Callicarpa pentandra is distinct from typical
members of the genus because it has pentamerous rather than 4-merous flowers, larger
oblong rather than elliptical anthers and a fruit with five locules each with two ovules
giving a total of ten seeds, rather than the typical four (Fig. 3). As discussed in Bramley
(2009) it formed part of the genus Gewnsia Blume (now included in Callicarpa), on
the basis of the characters listed above, and others. Further reference to this “Geunsia’
group will be made in a later section on the Borneo Endemics. Callicarpa pentandra
has been noted as a pioneer species (Tsai 1991); its bright red fruit is likely to be bird-
dispersed, and given that it has more than double the typical number of seeds than
other member of the genus, it is likely to have an advantage in numbers of seed in the
soil seed bank.
292 Gard. Bull. Singapore 63(1 & 2) 2011
Fig. 2. Distribution of Callicarpa pentandra (grey squares) and C. longifolia (black circles).
Callicarpa longifolia extends further south than C. pentandra, to Australia. Its
morphology is more typical for the genus; it has small 4-merous flowers, it is a shrub
rather than a tree, and has only 4 locules per ovary, with | ovule per locule. The fruit
however is white, rather than the more typical purple, and very fleshy—although this
is difficult to see in dry material (Fig. 3). In this case I assume its success in secondary
habitats is due to the attractiveness of the juicy fruit which is readily dispersed by birds
or rodents such as treeshrews (Fletcher 1938, Snow 1981).
Borneo endemics
There are 19 Callicarpa species endemic to Borneo. Four of these endemic species
have the 4-merous flowers and four ovules that are typical for the genus. The remaining
fifteen species belong to the ‘Geunsia’ group, an informal group that I am using to
recognise Callicarpa pentandra and its relatives. Along with C. pentandra, they have
features once recognised under the genus Geunsia Blume (see Bramley 2009: 417).
Unlike C. pentandra, they typically have 4-merous flowers but still have two rather
than one ovule per locule than typical for the genus. Unusually, C. hispida (Moldenke)
Bramley can have up to 7-merous flowers, with a total of 14 ovules (2 in each of
its seven locules) (Fig. 4). Some of these species are widespread across Borneo (C.
havilandii (King & Gamble) H.J.Lam), others are only known from much smaller areas
(e.g., C. argentii Bramley; C. anomala (Ridl.) B.L.Burtt; C. swhaequalis Bramley). All
of these species have a fruit that ripens red which ts larger than typical for Callicarpa,
presumably because of the larger number of developing seeds (Fig. 3). In addition, a
number of the species have a dense ferruginous indumentum of various different hair
Malesian Callicarpa 293
Fig. 3. Callicarpa longifolia: A. Flowering branch. B. Fruits. Callicarpa pentandra: C.
Flowering and fruiting branch. D. Fruiting branch. E. Habit.
types (see Bramley 2009). All tend to be found in areas of primary forest, often with
some degree of disturbance.
Of additional interest are C. kinabaluensis Bakh. & Heine and C. clemensorum
Moldenke, found at high altitude (1600-2500 m) on the peaks surrounding Mount
Kinabalu: they are both extremely hairy and have inflorescences made up of almost
globose cymes, perhaps an adaptation to their environment.
Philippines endemics
The 16 Callicarpa species endemic to the islands of the Philippines represent a
number of different informal morphology-based groups containing species from other
294 Gard. Bull. Singapore 63(1 & 2) 2011
Fig. 4. Examples of Geunsia group species from Borneo. A. Callicarpa hispida (Moldenke)
Bramley: Al, 7-merous flowers; A2, fruits cut open to reveal 14 ovules. B. Callicarpa havilandii
(King & Gamble) H.J.Lam. C. Callicarpa involucrata Mert.
3 s
3
Wa ?
im - bs
if j
s
pe e
f x
= )
{J a Cc
4 ij
tw xe
(A <<
$ oa
nS [Pers |» as
1A. o { oft
SA e Het \
i =. + - \
ar a0 S
- 2 oe
\ ene * BI
; 4 ‘Sa - G
_~
Fig. 5. Distributions of species endemic to the Philippines. Most species have narrow
distributions (each species is signified by a different symbol), the exception being C. micrantha
(grey squares). The Geunsia group of species are restricted to the southern islands (circular
symbols).
Malesian Callicarpa 295
areas of Malesian, China or Indo-China, to be detailed in the forthcoming revision of
Philippine Callicarpa (Bramley, in press, a), or, they do not appear to belong to any
particular group.
Most of the endemic species appear to have narrow ranges (Fig. 5). Only
C. micrantha Vidal occurs throughout the Philippine islands. It is morphologically
similar, through its delicate inflorescence, to the widespread C. japonica, native to
China and Japan. There is a general divide between the Luzon and the Mindanao
island groups. Most species present in the Luzon area are not present further south than
the northern tip of Samar, with the exception of C. micrantha. The species endemic to
the Mindanao islands are members of the Geunsia group (Basilan and Mindanao—C.
basilanensis Merr., C. flavida Elmer, C. ramiflora Merr., C. surigaensis Merr.;
Samar—C. ramiflora), with the exception of C. apoensis Elmer (endemic to Mount
Apo). The furthest north the Geunsia group extends is Morong, in Rizal Province,
Luzon (14°31°N 121°14’E; Vidal 3430), represented by the widespread C. pentandra.
The Geunsia group appears to be restricted to Malesia, and is only represented by C.
pentandra outside of Borneo, the Philippines and Sulawesi.
The species that occur in the Philippines as well as other areas appear to have
eastern Malesian distributions, sometimes extending to Indo-China, but they do not
occur in Sundaland (Fig. 6). For example, C. pedunculata R.Br. is a widespread species
that does not occur further west than the Philippines. Perhaps this can be explained by
a habitat requirement for seasonality: as described by van Steenis (1979), the western
side of the Philippines, as well as parts of Sulawesi and New Guinea, are considered
Fig. 6. Examples of Callicarpa species with Eastern Malesian distributions: C. pedunculata
R.Br. (triangles), C. erioclona Schauer (grey circles), C. caudata Maxim. (black squares), C.
dolichophylla Mert. (grey squares).
296 Gard. Bull. Singapore 63(1 & 2) 2011
seasonal or monsoon areas. Likewise, C. erioclona Schauer is found from Vietnam,
through the Philippines, on Java, in Sulawesi and New Guinea and has been recorded
once from each of Kudat and Banggi Islands off Sabah. Lack of spread to the west is
again perhaps likely to be a result of a preference for a degree of seasonality.
Callicarpa in comparison to other Lamiaceae genera
Callicarpa can be described as a Sunda Shelf and Wallacea oriented genus, with little
presence in the Sahul Shelf (regions as defined by van Welzen & Slik (2009)). Premna
(14 spp.; de Kok, in press), Vitex (16 spp., de Kok (2008)), Teijsmanniodendron (23
spp., de Kok et al. (2009)) are other Lamiaceae genera with similar distributions
but are much smaller than Ca/licarpa in terms of species numbers in Malesia. Only
Clerodendrum (65—70 spp., J. Wearn, pers. comm.) surpasses Callicarpa in terms of
species numbers. Premna and Vitex have been revised by Rogier de Kok (2007, 2008;
submitted), and Clerodendrum is under revision (Wearn & Mabberley, in prep.).
In terms of the numbers of species endemic to islands or island groups,
Callicarpa is unusual. For example, there are only seven out of 22 species of
Clerodendrum endemic to the Philippines (J. Wearn, pers. comm.). On Borneo, the
only comparable genus may be Teijsmanniodendron: all of its 23 species occur on the
island, and 11 of these are endemic (de Kok et al. 2009).
A team of volunteers have begun to database the Malesian Lamiaceae
collections at K: once all specimens have been georeferenced and distribution maps
completed for each genus, the baseline data for a comparative study of generic
distributions will be provided. A particularly interesting question 1s why some genera
appear to have radiated on the island of Borneo, the Geunsia group of Callicarpa
especially. In addition, van Welzen & Slik (2010) recently reported that Lamiaceae
sensu Lindley, treated for the Flora Malesiana by Keng (1968), is an example of a
family with relatively high numbers of species (more than expected) in Wallacea.
This makes for an interesting comparison with the woody genera transferred from the
Verbenaceae discussed here.
Questions to be addressed in future research
This descriptive paper aims to act as a precursor to any future phylogenetic or
biogeographic analyses. To be able to test whether the morphological groups
discussed reflect evolutionary relationships between species, we need to elucidate the
relationships between species (to date I only have a limited phylogeny). Furthermore,
to understand the distribution of Callicarpa more generally, and to make comparisons
with other Lamiaceae genera, I need to understand the position of Ca/licarpa within the
Lamiaceae, a project also requiring phylogenetic work. In some preliminary analyses it
has appeared close to the Australian subfamily Prostantheroideae Luerssen (Bramley,
unpubl.). Indeed Callicarpa does share, at least superficially, some characters with the
tribe Chlorantheae Benth. & Hook.: actinomorphic flowers, branched hairs, variable
number of stamens. There is currently no solid evidence to support this position but it
would be interesting to pursue this line of investigation, and also the developmental
basis of the actinomorphic flower structure, rare within the Lamiaceae as a whole.
Malesian Callicarpa 297
References
Baker, W.J., Coode, M.J.E., Dransfield, J., Dransfield, S., Harley, M.M., Hoffmann, P.
& Johns, R.J. (1998) Patterns of distribution of Malesian vascular plants, in Hall,
R. & Holloway, J.D. (eds) Biogeography and Geological Evolution of South East
Asia pp. 243-258.
Bramley, G.L.C. (2009) The genus Callicarpa (Lamiaceae) on Borneo. Bot. J. Linn.
Soc. 159: 416-455.
Bramley, G.L.C. (in press, a) The genus Callicarpa (Lamiaceae) in the Philippines.
Kew Bull.
Bramley, G.L.C. (in press, b) Three new species of Callicarpa (Lamiaceae) from
Sulawesi: Callicarpa anisodonta, C. mendumiae and C. pseudoverticillata. Kew
Bull.
Bramley, G.L.C., de Kok, R.P.J., Wearn, J., Walsingham, L., Davies, N.J. & Mabberley,
D.J. (in prep.) Lamiaceae II for the Flora Malesiana.
Fletcher, H.R. (1938) The Siamese Verbenaceae. Bull. Misc. Inform. Kew 1938: 401—
445.
Graham, C.H., Ferrier, S., Huettman, F., Moritz, C. & Townsend Paterson, A. (2004)
New developments in museum-based informatics and applications in biodiversity
analysis. Trends Ecol. Evol. 19: 497-503.
Harley, R.M., Atkins, S., Budantsev, A.L., Cantino, P.D., Conn, B.J., Grayer, R.,
Harley, M.M., de Kok, R.P.J., Krestovskaja, T., Morales, R., Paton, A.J., Ryding,
O., & Upson, T. (2004) Labiatae. In: Kubitzki, K. (ed) The Families and Genera
of Vascular Plants VM. Pp 167-275.
Hoffmann, P. (2005) Antidesma in Malesia and Thailand. Kew: Royal Botanic
Gardens, Kew.
Keng, H. (1968) Labiatae. Flora Malesiana 8. Alphen aan den Rijn : Sijthoff &
Noordhoff.
Kok, R.P.J. de (2007) The genus Vitex L. in New Guinea and the South Pacific Islands.
Kew Bull. 587-603.
Kok, R.P.J. de (2008) The genus Vitex (Labiatae) in the Flora Malesiana region,
excluding New Guinea. Kew Bull. 63: 17-40.
Kok, R.P.J. de (in press) The genus Premna (Lamiaceae) in the Flora Malesiana region.
Kew Bull.
Kok, R.P.J. de, Rusea, G. & Latiff, A. (2009) The genus 7eijsmanniodenron Koord.
(Lamiaceae) Kew Bull. 64: 587-625.
Moat, J. (2007) Conservation assessment tools extension for ArcView 3.3, version 1.2.
Kew: Royal Botanic Gardens, Kew. (http://www.rbgkew.org.uk/gis/cats)
Nic Lughadha, E., Baillie, J., Barthlott, W., Brummitt, N.A., Cheek, M.R., Farjon, A.,
Govaerts, R., Hardwick, K.A., Hilton-Taylor, C., Meagher, T.R., Moat, J., Mutke,
J., Paton, A.J., Pleasants, L.J., Savolainen, V., Schatz, G.E., Smith, P., Turner,
I., Wyse-Jackson, P. & Crane, P.R. (2005) Measuring the fate of plant diversity:
towards a foundation for future monitoring and opportunities for urgent action.
Philos. Trans., Ser. B 360: 359-372.
298 Gard. Bull. Singapore 63(1 & 2) 2011
Snow, D.W. (1981) Tropical frugivorous birds and their food plants: A world survey.
Biotropica 13: \—14.
Steenis, C.G.G.J. van (1950) The delimitation of Malesia and its main plant geographical
divisions. Flora Malesiana 1(1): LXX—LXXV. Djakarta: Noordhoff-Kolff N.V.
Reprinted (1985) Koenigstein, Germany: Koeltz Scientific Books.
Steenis, C.G.G.J. van (1979) Plant-geography of east Malesia. Bot. J. Linn. Soc. 79:
97-178.
Thiers, B. (continuously updated). Index Herbariorum: A global directory of public
herbaria and associated staff. New York Botanical Garden’s Virtual Herbarium
(http://sweetgum.nybg.org/ih/)
Tsai, L.M. (1991) Biomass production and biomass relationship of young Callicarpa
pentandra. Pertanika 14: 281-285.
Welzen, P.C. van & Slik, J.W.F. (2009) Patterns in species richness and composition of
plant families in the Malay Archipelago. Blumea 54: 166-174.
Gardens’ Bulletin Singapore 63(1 & 2): 299-306. 2011 299
How did Magnolias (Magnoliaceae: Magnolioideae)
reach Tropical Asia?
Hans P. Nooteboom
Netherlands Centre for Biodiversity — Naturalis,
Section National Herbarium of The Netherlands, Leiden University,
PO Box 9514, 2300 RA Leiden, The Netherlands
nooteboom@nhn.leidenuniv.nl
ABSTRACT. Extant magnolias (Magnoliaceae, Magnolioideae) have a classic disjunct
distribution in southeast Asia and in the Americas between Canada and Brazil. Molecular
analyses reveal that several North American species are basal forms suggesting that magnolias
originated in North America, as indicated by their fossil record. We recognise four elements in
their evolution: (1) Ancestral magnolias originated in the Late Cretaceous of North America
in high mid-latitudes (45°-60°N) at low altitudes in a greenhouse climate. (2) During the
exceptionally warm climate of the Eocene, magnolias spread eastwards, via the Disko Island
and Thulean isthmuses, first to Europe, and then across Asia, still at low altitudes and high
mid-latitudes. (3) With mid-Cenozoic global cooling, they shifted to lower mid-latitudes (30°—
45°N), becoming extinct in Europe (Yu/ania was still present less than 2 mya.) and southern
Siberia, dividing a once continuous distribution into two, centered in eastern Asia and in North
America. (4) In the late Cenozoic, as ice-house conditions developed, magnolias migrated
southward from both centres into moist warm temperate upland sites in the newly uplifted
mountain ranges of South and Central America, southeast Asia, and the High Archipelago,
where they diversified. Thus the late Cenozoic evolution of magnolias is characterised by
impoverishment of northern, and diversification of southern species, the latter being driven
by a combination of high relief and climate oscillations, and neither of the present centres of
diversity is the centre of origin. Magnolioideae appear to consist of only the genus Magnolia.
Keywords. Magnoliaceae, Magnolia, fossils, DNA distribution, paleoclimate
Introduction
Understanding the present distribution (Fig. 1) and fossil occurrence (Fig. 2) of
an ancient land-based group such as magnolias (subfamily Magnolioideae in the
family Magnoliaceae) has to be done in relation to the historical disposition of land
and global climate. For his Ph.D. (1956-59) Neil Opdyke studied the relationships
between paleomagnetically determined latitudes and paleoclimates as inferred from
temperature-sensitive deposits and from wind directions in eolian sandstones [Opdyke
1961a,b]. By placing such geological deposits at their original latitude, he was able to
discuss possible past changes in climatic zones and in the positions of continents. Being
able to do so is the key to unravelling the origin and paleogeographical dispersal of
plant groups and individual species, in a way not available to earlier phytogeographers.
The genus Magnolia was described by Linnaeus in 1753 from material collected
in what were then colonies of British North America. The type species is Magnolia
300 Gard. Bull. Singapore 63(1 & 2) 2011
Present distribution genus Magnoli
HM wmerous species
@ few isolated species
Fig. 1. Extant magnolias have a classic disjunct distribution in Southeast Asia and in the
Americas between Canada and Brazil, and nowhere in between. (Source: Hebda & Irving 2004)
Fig. 2. Well established fossil occurrences of Cenozoic magnolias and Cretaceous ancestral
taxa. Numbers in parentheses are approximate ages in millions of years. (Source: Hebda &
Irving 2004)
Magnolia in tropical Asia 301
virginiana L., the Laurel magnolia or Sweet bay. Magnolias are members of the
Paleogene “boreotropical” flora. Magnolias flourished throughout the Cenozoic, and
there are ancestral forms in the Late Cretaceous Epoch connecting them to one of the
oldest lineages of flowering plants.
Taxonomy of modern Magnolias
Until recently, magnolias in the broad sense were divided among about half a dozen
genera based on morphology and geographic isolation (Nooteboom 1993). There
was much discussion concerning whether all these should be recognised as separate
genera, the relationships among them, and the assignment of species to different genera
(Nooteboom 1996, 2000). Recent studies of the position of flowers, the presence and
absence of floral stipes (Figlar 2000), and especially of chloroplast DNA (cpDNA),
have clarified relationships within the genus. Specifically, Figlar’s study showed that
morphological differences, once thought to separate Magnolia and Michelia L. are not
valid, and cpDNA studies demonstrated that the variation within the genus Magnolia,
as it was originally narrowly accepted, is greater than the variation among other closely
related “genera” (Azuma et al. 2001, Kim et al. 2001).
This was later confirmed by study of nuclear DNA (Nie et al. 2008). For
example, species in the old genus Michelia cluster closely with Magnolia species
(Azuma et al. 2001, Kim et al. 2001). On the other hand, several species in the
old section Riytidospermum Spach, long considered to be indisputably magnolias,
are, in fact, not closely related. Also, from the perspective of cpDNA as well as of
nuclear DNA, western hemisphere (especially North American) species in the section
Rhytidospermum appear to be the most diverse and basal in the cpDNA and in the
nuclear DNA trees, suggesting that they could be relicts of the ancestral stock. Because
there is evidence of ancient hybridisation. for instance several Magnolia species are
polyploid, deep phylogeny cannot be deduced from chloroplast DNA, only from
nuclear DNA.
Recent and historical geographical distribution of Magnolias
Molecular analyses reveal that several North American species are basal forms,
suggesting that magnolias originated in North America, as indicated by their fossil
record. We recognise four elements in their evolution.
(1) Ancestral magnolias originated in the Late Cretaceous of North America (Fig. 2) in
the high mid-latitudes (45°-60°N), at low altitudes in a greenhouse climate.
Continental drift cannot be the sole factor in determining the distribution of
extant magnolias, because reconstructing continents for Late Cretaceous and Paleogene
epochs does not bring together the group’s two principal zones of modern occurrence.
Thus, Hebda & Irving (2004) made use of previous ideas concerning large changes
302 Gard. Bull. Singapore 63(1 & 2) 2011
Antarctic Ice
(a) super
greenhouse
<_>
o
=
> | terminal
© | Cretaceous East
© bolide Antarctica
= marked Ice
2 increase West
in CO, rapid 4 Antarctica
mid-Tertiary Ice
cooling
greenhouse ==> + << icehouse increasing
Arctic Ice
Compiled by E. Irving
60 40 20 (millions of years)
(b)
collision of India, progressive
4— development of mountains —>
in SE Asia and Americas
possible Antillean
< connection >
circum-Antarctic
current starts
A |
opening of creation o
Drake Passage and HA, separation o
Tasman Gateway __ Pacific and Indian
tte ee ee
London Clay Clarno Brandon
K ,Paleocene Eocene Oligocene Miocene
Temperature
60 40 20 (millions of years)
Fig. 3. Climatic (a) and paleogeographical (b) events during the past 70 Ma. HA = High
Archipelago on the Sunda platform, Pl = Pliocene, Q = Quaternary. (London Clay, Clarno,
Brandon and Clarkia are notable fossil magnolia localities). Compiled by E.I. Irving from
various sources.
Magnolia in tropical Asia 303
in the past distribution of moist warm temperate climate, and in the development of
upland habitat.
These changes are linked to the Mid- and Late Cenozoic evolution of global
climate from a non-glacial (greenhouse) to a glacial (ice-house) regime and to the
contemporaneous creation of regions of moist, warm temperate climate in the newly
uplifted mountains of the Americas, southeast Asia, and the High Archipelago on the
Sunda Platform (Fig. 3).
The history of land masses is shown in Fig. 4, based on plate tectonics (for
positions of continental lithosphere), paleomagnetic evidence (for geographical grid),
and the distribution of terrestrial and marine sediments.
(2) During the exceptionally warm climate of the Eocene (Fig. 4b), magnolias spread
eastwards, via the Disko Island and Thulean isthmuses, first to Europe, and then across
Asia, still at low altitudes and high mid-latitudes.
The Thulean land-bridge remained almost continuous until the Miocene, and
the remnant Disko Island volcanoes may have reduced the obstacle of the Labrador
Sea. Cross-Atlantic migration routes may, therefore, have been open but were less
hospitable than earlier.
(3) With mid-Cenozoic global cooling, they shifted to lower mid-latitudes (30°—
45°N), eventually becoming extinct in Europe and southern Siberia, dividing a once
continuous distribution into two, and becoming centred in eastern Asia and in North
America.
(4) In the late Cenozoic, as ice-house conditions developed, magnolias migrated
southward from both centres into moist warm temperate upland sites in the newly
uplifted mountain ranges of South and Central America, southeast Asia, and the High
Archipelago (Fig. 4e).
Thus the late Cenozoic evolution of magnolias is characterised by
impoverishment of northern, and diversification of southern, species, the latter being
driven by a combination of high relief and climate oscillations, and neither of the
present centres of diversity is the centre of origin.
The evidence from DNA
The divergence times for various major clades derived from a molecular (nuclear
DNA) phylogenetic analysis of magnolias have been estimated by Nie et al. (2008).
The relationship among the different groups 1s shown in Fig. 5. The major divergences
include:
— Liriodendron L. and the rest of Magnoliaceae diverging c. 93 mya (Liriodendron
then divided into two species c. 14 mya).
— The American section 7Zalauma Baill., which is the base of all other magnolias,
began diverging c. 54 mya. (The American species Magnolia tripetala L. is ancestral
304
Gard. Bull. Singapore 63(1 & 2) 2011
(a) Ancestral magnolias (AM) at about 80 to 90 Ma (b) About 60 Ma
: \—_>
(c) About 40 Ma
(e) Quaternary
and Recent
Fig. 4. a. Possible latest Cretaceous northward migration of Vancouver Island shown. b.
Distribution of possible magnolia sites with respect to land bridges. ¢. Slight cooling about 40
million years ago and a beginning of magnolia migration southwards. d. About 20 million years
ago, magnolias have become nearly extinct in Europe, but Y/ania Spach. seeds are found until
c. 2 mya (Hebda & Irving 2004, v.d. Hammen et al. 1971). e. The deciduous, temperate species
are derived from evergreen warm-temperate species as an adaptation to the climate. (Source:
Hebda & Irving 2004). BE = Bering land bridge; DG = De Geer Landbridge; FF = Fossil
Forest, Axel Heiberg Island; HA = High Archipelago on the Sunda platform; IP = Isthmus of
Panama; M = possible Magnolia site; TH = Thulean landbridge; TS = Turgai Street; * = fossil
Magnolia; crosses indicate fewer (often more isolated) species; black areas indicate numerous
species present; stippled areas are maximum glaciated cover.
Magnolia in tropical Asia 305
South American Talauma
Section Oyama (Asia)
54.12 10.57 Magnolia tripetala (Am. Rhytidospermum)
Asian Rhytidospermum)
Manglietia (Asia)
Section Kmeria (Asian)
Section Gwillimia (incl. section Blumiana)
(Asian)
Section Gynopodium (Asian)
30.22
Section Magnolia (American)
Section Auriculata (American)
Section Macrophylla (American)
Magnolia acuminata
28.29 (American section Yulania)
Chinese section Yulania
Section Michelia incl. sect. Elmerrillia
(Asia)
Fig. 5. A cladogram of the sections of which the nuclear DNA is known. Adapted from Fig. 6
of Nie et al. (2008).
to the other (Chinese) species of section Rhytidospermum, which began to diverge
around 10 mya.)
— The Chinese Magnolia sinica (Y.W. Law) Noteboom and Magnolia nitida W.W.
Sm. (section Gynopodium) are basal to section Magnolia, which diverged c. 30 mya.
(This split happened probably in America. The American M. virginiana L. is basal to
the rest of section Magnolia, which diverged as two groups c. 18 mya.)
— The American Magnolia fraseri Walter and M. macrophylla Michx., basal to both
Yulania and Michelia, are clearly ancestral. Magnolia acuminata (L.) L., the only
American Yulania species, is basal to both the rest of Yu/ania and Michelia, and
diverged about 28 mya.
From such analyses, the resulting cladograms lead to the conclusion that
Magnolioideae consists of only one genus, Magnolia. Even the subdivision into three
subgenera is now obsolete. Apart from subg. Magnolia, only one other subgenus may
be arguably recognised, Yu/ania Spach, consisting of five sections: the basal (American)
sections Auriculata Figlar & Noot. and Macrophylla Figlar & Noot.; Tulipastrum
(Spach) Figlar & Noot. with only the American M. acuminata; section Yulania; and
section Michelia. Here the problem is that sections Auriculata and Macrophylla do not
group morphologically with Michelia and Yulania.
ACKNOWLEDGEMENTS. I thank Dr. R.J. Hebda and Dr. Hiroshi Azuma for letting me use
their work. I thank the organisers of the 8th International Flora Malesiana symposium for
inviting me.
306 Gard. Bull. Singapore 63(1 & 2) 2011
References
Azuma, H., Garcia-Franco, J.G., Rico-Gray, V. & Thien, L.B. (2001) Molecular
Phylogeny of the Magnoliaceae: The biogeography of tropical and temperate
disjunctions. Amer. J. Bot. 88: 2275-2285.
Azuma, H., Rico-Gray, V., Garcia-Franco, J.G., Toyota, M., Asakawa, Y. & Thien, L.B.
(2004) Close relationship between Mexican and Chinese Magnolia (subtropical
disjunct of Magnoliaceae) inferred from molecular and floral scent analyses. Acta
Phytotax. Geobot. 55(3): 167-180.
Figlar, R.B. (2000) Proleptic branch initiation in Michelia and Magnolia subgenus
Yulania provides basis for combinations in subfamily Magnolioideae. In: Liu. Y.,
Fan, H., Chen, Z., Wu, Q. & Zeng, Q. (eds) Proceedings of the International
Symposium on the Family Magnoliaceae, pp. 26-37. Beijing: Science Press.
Hammen, T. van der, Wymstra, T.A. & Zagwijn, W.H. (1971) The floral record of the
late Cenozoic of Europe. In: Turekian, K.K. (ed) The Late Cenozoic Glacial Ages,
pp. 391-424. New Haven: Yale Univ.
Hebda, R.J. & Irving, E. (2004) On the origin and distribution of Magnolias: Tectonics,
DNA and climate. Geophys. Monogr. 145: 1-15.
Kim, S., Park, C.W., Kim, Y-D. & Suh, Y. (2001) Phylogenetic relationships in family
Magnoliaceae inferred from NDHF sequences. Amer: J. Bot. 88: 717-728.
Lang, G. (1994) Quartare Vegetationsgeschichte Europas. Methoden und Ergebnisse.
Jena, Stuttgart, New York: Gustav Fischer Verlag.
Linnaeus, C. (1753) Species Plantarum 1: 535.
Nie, Z.-L., Wen, J., Azuma, H., Qiu, Y.L., Sun, H., Meng, Y., Sun, W.-B. & Zimmer,
E.A. (2008) Phylogenetic and biogeographic complexity of Magnoliaceae in the
Northern Hemisphere inferred from three nuclear data sets. Molec. Phylogen.
Evol. 48: 1027-1040.
Nooteboom, H.P. (1993) Magnoliaceae. In: Kubitzki, K. (ed) The Families and Genera
of Vascular Plants, Vol. 2., pp. 391—401.Berlin: Springer Verlag.
Nooteboom, H.P. (1996) The tropical Magnoliaceae and their classification. In: D.
Hunt (ed) Magnolias and their Allies, pp. 71-80. Proceedings of an International
Symposium, Royal Holloway, University of London, Egham, Surrey, U.K. 12—13
April 1996.
Nooteboom, H.P. (2000) Different look at the classification of the Magnoliaceae. In:
Liu, Y., Fan, H., Chen, Z., Wu, Q. & Zeng, Q. (eds) Proceedings of the International
Symposium on the Family Magnoliaceae, pp. 26-37. Beijing: Science Press.
Opdyke, N.D. (1961a) The impact of paleomagnetism on paleoclimatic studies. /nt. J.
Bioclimatol. Biometeorol. 3: 1-6.
Opdyke, N.D. (1961b) The climatological significance of desert sandstones. In: Nairn,
A.E.M. (ed) Descriptive Palaeoclimatology, pp. 45—60.London: Interscience.
Wu, C.-I. & Li, W.H. (1985) Evidence for higher rate of nucleotide substitution in
rodents than in man. Proc. Natl. Acad. Sci. U.S.A. 82: 1741-1745.
Gardens’ Bulletin Singapore 63(1 & 2): 307-327. 2011 307
Spatial and temporal diversification of Tetrastigma
(Vitaceae)
Pingting Chen'*, Jun Wen? and Longging Chen!
'Key Laboratory of Horticultural Plant Biology (Ministry of Education),
College of Horticulture and Forestry Science,
Central China Agricultural University, Wuhan 430070, P. R. China
"Institute of Agricultural Quality Standards and Testing Technology Research,
Hubei Academy of Agricultural Science, Wuhan 430064, P. R. China.
celiacpt@yahoo.com
‘Department of Botany, National Museum of Natural History,
MRC166, Smithsonian Institution, Washington, D.C. 20013-7012, U.S.A.
wenj(@s1.edu (corresponding author)
ABSTRACT. The spatial and temporal diversification of 7etrastigma Planch., a genus of
the grape family Vitaceae with a wide distribution throughout subtropical and tropical Asia
to Australia, was examined through phylogenetic and biogeographic analyses. The times of
divergence within 7etrastigma were estimated with the Bayesian approach based on sequence
data from four plastid (atpB-rbcL, rps16, trnL-F, and psbA-trnH) markers using the computer
program BEAST. The divergence time between 7efrastigma and its closely relative Cayratia
was estimated as the early Eocene around 50.6 million years ago (mya), with 95% HPD: 36.3—
65.3 mya. The age of the crown group of 7etrastigma was dated to be late Eocene (36.9 mya,
with 95% HPD: 25.7-49.3 mya). Biogeographic analyses using LAGRANGE suggested that
the Sino—Himalayan region (and the adjacent Indochina) was the most likely ancestral area for
Tetrastigma. Most Tetrastigma species sampled from the Malesian region were nested within
clades of the Sino—Himalayan and Indochina region. A few Malesian species primarily from
SE Sulawesi, the Philippines and New Guinea are not associated with the Sino—Himalayan and
Indochina species and formed separated clades. The results suggest that 7efrastigma species
in the Malesian region have complex biogeographic origins and continental Asia served as an
important source area for the Malesian members of the genus.
Keywords. BEAST, biogeography, dating, LAGRANGE, Tetrastigma, Vitaceae
Introduction
Tetrastigma Planch. is one of the 14 currently recognised genera of the grape family
Vitaceae (Wen 2007). The genus contains approximately 95 species and has a wide
distribution in tropical and subtropical Asia, extending to Australia, ranging from
India to China, across SE Asia eastward to Fiji (Fig. 1; Chen et al. 2011). Tetrastigma
is characterised by unbranched to digitately branched tendrils, a dioecious sexual
system, and 4-lobed stigmas in female flowers (Wen 2007). Tetrastigma 1s well-known
in Southeast (SE) Asia for being the host plants for all three genera of Rafflesiaceae,
which contains the largest flower of the world (Meier 1997, Barkman et al. 2004).
308 Gard. Bull. Singapore 63(1 & 2) 2011
ge BY IG
Sn ES ES
QF
fine => = (9 ae SS x,
Tn gin Sey, = sh, OA
“RGB v3 ee
~ Ls \.
7 Rome
in
d: New World
Fig. 1. The geographic distribution of 7etrastigma in Asia and Australia (gray area). “‘a—d”
indicates four areas of endemism used in the LAGRANGE analysis: a, the Malesian region; b,
Sino—Himalayan and Indochina region; c, Africa; and d, New World.
Recently the genus attracted more scientific attention with the reported horizontal
gene transfer of mitochondrial genes (e.g., nad/B-C and apt/) from the Tetrastigma
host to the parasite of Rafflesia R.Br. (Davis & Wurdack 2004, Barkman et al. 2007).
Several species of the genus are widely cultivated as climbing ornamentals (Wen
2007). Tetrastigma hemsleyanum Diels & Gilg is an important Chinese folk medicine
for treating hepatitis, fever, pneumonia, rheumatism, and sore throat (Liu et al. 2002).
Southeast Asia and the West Pacific have long attracted the attention of
biogeographers. The waters of SE Asia contain the highest marine faunal diversity
in the world (Briggs 1974; Paulay 1997). Biodiversity in this region shows several
patterns of origin between two complicated palaeocontinents separated for a
considerable period of time. Two main patterns have often been recognised: (1) a
pattern of Southeast Asian elements (perhaps of Laurasian origin) and (2) a pattern of
Australian elements (perhaps of Gondwanan origin). Some recent phylogenetic and
biogeographic studies have shed important insights into the evolution of plant, insect,
fish, and animal distribution patterns in this region (see Butlin et al. 1998; Marwick
2009; Clouse & Girbet 2010; Esselstyn & Oliveros 2010; Renner et al. 2010). Yet the
Malesian flora comprises an estimated 42,000 plant species (Roos 1993), with only
about 15% being revised during the last 50 years. Thus biogeographic analyses of
more plant genera or lineages 1n this region are very much needed.
Several lines of evidence have shown that good dispersers such as some
plants and insects from Eurasia can dominate the entire SE Asia and extend to the
Pacific islands (Gressitt 1982, Baker et al. 1998). Zetrastigma is a good disperser,
bearing berries and often dispersed by fruit-eating birds, bats and mammals (Tiffney &
Biogeography of Tetrastigma 309
Barghoorn 1976, Moran et al. 2009). The genus thus provides another opportunity to
reconstruct the biogeographic diversification of plants in SE Asia and the West Pacific.
Tetrastigma is widely distributed in the Sino—Himalayan / Indochina region
and the Malesian region. The Malesian region comprises west Malesia (the Sunda
shelf: Malay Peninsula, Sumatra, Borneo, and West Java), central Malesia (Wallacea:
part of Java, the Philippines, Sulawesi, the Lesser Sunda Islands, and the Moluccas),
and east Malesia (the Sahul Shelf: New Guinea) as subunits (van Welzen et al. 2005).
The Wallace’s line separates the west and the east Malesian flora, the transition area
between west and east Malesia is known as Wallacea, i.e., central Malesia. The tectonic
history of the Malesian region is complicated because of the two waves of Australian
slivers moving towards the Eurasian plate. The first wave formed SE Asia and the west
part of the Malay Archipelago to the Wallace’s or the Weber’s line around 90 million
years ago (mya). The second wave made SE Asia collide with the West Pacific in the
middle Miocene (van Welzen et al. 2003). Besides this complicated tectonic history
and cycles of glacials, the origins of plant taxa may also be complex. It is still poorly
known which taxa in this region dispersed from the Sino-Himalayan region or from
the West Pacific region. Because species of Tetrastigma occupy the entire Malesian
region, it seems a good model to test the dispersal pathways in the Malesian region.
The fossil records of the Vitaceae in the Northern Hemisphere during the
Tertiary are well documented. Pollen, fruit, seed and leaf fossils have been recorded
from the late Paleocene to the Pleistocene (Collinson 1983, Cervallos-Ferriz &
Stockey 1990, Taylor 1990, Wheeler & Lapasha 1994). Fossil seeds of Vitaceae
are commonly found in many Tertiary beds in the Northern Hemisphere (Reid &
Chandler 1933; Kirchheimer 1939; Miki 1956; Dorofeev 1957, 1963; Chandler 1925,
1957, 1961a, 1961b, 1962, 1963, 1964, 1978; Tiffney & Barghoorn 1976; Chen &
Manchester 2007). Because the fossils resemble extant seeds, these fossil seeds have
usually been identified to extant genera (Chen & Manchester 2007). Many of those
fossils were designated to the genus Vitis and others were included in Ampelopsis,
Ampelocissus, Parthenocissus; yet only a few fossils were reported to be Jetrastigma
and Cayratia (Tiffney & Barghoorn 1976, Chen & Manchester 2007, Chen 2009). In
general, fossil seeds of Vitaceae were relatively diverse in Europe (Chen 2009). About
13 fossil species of Tetrastigma have been reported from the early Eocene to Pliocene
(Kirchheimer 1938: Miki 1956; Chandler 1925, 196lab, 1962; Reid & Chandler 1933:
Teodoridis 2003). Most of them varied morphologically and were from Europe in
the early Tertiary and only two species (7 japonica Miki and T. tazimiensis Miki)
were from Asia (Japan) from the Pliocene. Chen & Manchester (2007) carefully
examined the fossils of Ampelocissus concerning seed morphological characters.
They questioned the generic assignment of most Jetrastigma fossil seeds, which
resemble those of Ampelocissus and Ampelopsis. They transferred two Tetrastigma
fossils to Ampelocissus: with T. lobatum Chandler now being Ampelocissus lobatum
(Chandler) Chen & Manchester, and Tetrastigma chandleri now as Ampelocissus
chandleri (Kirchheimer) Chen & Manchester (Chen & Manchester 2007). There may
be no reliable Tetrastigma fossil seeds so far (S. Manchester, pers. comm.). Moreover,
the oldest fossil in Vitaceae was Ampelocissus parvisemina Chen & Manchester from
310 Gard. Bull. Singapore 63(1 & 2) 2011
the late Paleocene of North Dakota and the early-middle Eocene of Oregon (Chen &
Manchester 2007).
Chen et al. (2011) conducted a phylogenetic analysis of 7etrastigma using
four plastid markers. Yet the biogeography of the genus has never been explored. The
objectives of this study are to (1) estimate the divergence times of 7etrastigma clades,
and (2) infer the biogeographic diversification history of the genus.
Materials and methods
Estimation of divergence times
Representatives from the entire grape family plus Leea were sampled to help date the
ages of Tetrastigma and its relatives with both direct fossil and secondary calibrations
m Vitales. Sequences used in the dating analysis are shown in Appendix A and were
derived from a phylogenetic analysis of the genus by Chen et al. (2011). We used
the Bayesian dating method based on a relaxed-clock model to estimate divergence
times (Thorne et al. 1998; Thorne & Kishino 2002; Drummond et al. 2006). The
Bayesian coalescent approach to estimate the times of each clade in Tetrastigma and
their credibility intervals was implemented in the computer program BEAST version
1.4.7 (Drummond & Rambaut 2007), which employs a Bayesian Markov chain Monte
Carlo (MCMC) to co-estimate topology, substitution rates and node ages. All analyses
were performed using the GTR model of nucleotide substitution with a gamma and
invariant sites distribution with four rate categories. The tree prior model (constant
size) was implemented in the analysis, with rate variation across branches assumed
to be uncorrelated and lognormally distributed (Drummond et al. 2006). The final
estimates were obtained using the model that yielded the highest posterior probability.
Posterior distributions of parameters were approximated using two independent
MCMC analyses of 20,000,000 generations with 10% burn-in. Samples from the two
chains which yielded similar results were combined and convergence of the chains
was checked using the program Tracer 1.3 (Rambaut & Drummond 2004).
Ancestral area analyses
Reconstruction of ancestral areas on a phylogeny is important to understand the
biogeographic diversification history of a lineage, as it permits the inference of
the place of origin and dispersal routes of organisms. Dispersal-vicariance (DIVA;
Ronquist 1996, 1997) analysis is often used to construct the biogeographic history. But
for DIVA analyses the phylogeny of a group needs to be well resolved. The procedure
used in the computer program LARANGE differs from the DIVA methods in that
it allows a broader range of speciation models and also incorporates any available
temporal information such as divergence times and dispersal opportunities. We thus
employed the maximum likelihood-based method LAGRANGE (Ree et al. 2005,
Ree & Smith 2008) to reconstruct the diversification of 7etrastigma and its close
relatives. Four areas of endemism were defined according to their distribution: (a)
the Malesian region, (b) the Sino—Himalayan and Indochina region, (c) Africa, and
Biogeography of Tetrastigma Bat
(d) the New World. The maximum number of areas in ancestral ranges was set as two
in LAGRANGE, as no species of Vitaceae is distributed in more than two areas of
endemism.
Vitaceae fossil constraints
We constrained the ages of two nodes in the phylogeny of TJetrastigma and its
close relatives in Vitaceae (Fig. 2). The stem lineage of Ampelocissus and Vitis was
constrained to be 60 += 5.0 mya old (node) in Fig. 2 based on fossil seeds of Ampelocissus
parvisemina from the late Paleocene of North Dakota and the early to middle Eocene
of Oregon (Chen & Manchester 2007). Wikstrém et al. (2001) reported the estimated
divergence time between Leea and Vitis as 80-92 mya. Recently Magallon & Castillo
(2009) suggested an older age from 90.65 (90.47—90.84) to 90.82 (90.64-91) mya for
the origin of Vitaceae using different relaxed or constraint schemes, although there
is no direct credible fossil evidence for Vitaceae or Leeaceae in the Cretaceous. It is
curious that the distinctive seed morphology of this clade, readily observable in many
Paleogene localities, is missing from well studied Cretaceous deposits. The inferences
from Magallon & Castillo (2009) and Wikstrém et al. (2001) are close, although the
latter were criticised for the nonparametric rate smoothing method and for calibrating
their tree using only a single calibration point. Bell et al. (2010), however, suggested a
time ranging from 65 (45-81) to 48 (21-79) Ma for the stem age of Vitaceae. Although
their estimates were based on 36 fossil calibrations and a relaxed approach in dating
the whole angiosperm groups, they seem to have underestimated the age for Vitaceae
because the estimates were younger than the age suggested by fossil evidence (e.g.,
in Chen & Manchester 2007). We thus constrained the stem of Vitaceae to be 85 =
5.0 mya (Fig. 2), a strategy similar to that of Nie et al. (2010). We did not use any
Tetrastigma fossils from the Tertiary (Kirchheimer 1938; Miki 1956, Chandler 1925,
196lab, 1962: Reid & Chandler, 1933: Teodoridis 2003), as Chen & Manchester
(2007) questioned the placement of all these fossils in the genus.
Results
The chronogram of TJetrastigma and its relatives from Vitaceae based on combined
plastid atpB-rbcL, rps 16, trnL-F, and psbA-trnH data is shown in Fig. 2. The age of the
Tetrastigma stem was estimated to be 50.6 mya (with a 95% highest posterior density
[HPD] interval of 36.4—65.3 mya) in the early Tertiary. The crown age of Tetrastigma
was estimated to be 36.94 mya (with a 95% HPD: 25.7-49.3 mya) approximately in the
late Eocene. Node ages of eight major clades (clades A—H) are shown in Table 1. Clade
B formed a well-supported clade with the large Clade A based on Bayesian analysis
(Fig. 2). Three species (7. glabratum Wen10670, T. lawsoni Wen7505, T. tuberculatum
Wen10280) in Clade B were collected from Java, Singapore, and SE Sulawesi. The
crown age of Clade D was estimated to be 7.8 mya (95% HPD: 2.2—14.5 mya, node
D in Fig. 2) including two endemic species of the Philippines (T. ellipticum, and T.
laxum) and a new undescribed species collected from SE Sulawesi. The small Clade
312 Gard. Bull. Singapore 63(1 & 2) 2011
Leea spinea Wen9575
Leea monticola Wen9569
Leea indica Wen10910
Cayratia mollissima Wen8403
Cayratia geniculata Wen10275
Cayratia cordifolia Wen 10548
Cayratia pedata Wen7428
Tetrastigma pedunculare Wen10281
Tetrastigma brunneum Wen8240
Tetrastigma trifoliolatum Wen8&350
Tetrastigma papillosum Wen8401]
Tetrastigma obtectum Nie& Meng433
Tetrastigma triphyllum Wen106S5
Tetrastigma yunnanense Nie2003 104
Tetrastigma napaulense Nie&ZhuS48
Tetrastigma serrulatum Wen7429
Tetrastigma napaulense Tibet225
Tetrastigma serrulatum Wen 10856
— Tetrastigma loheri Wen10202
Tetrastigma pisicarpum Wen10185
Tetrastigma annamense Wen! 1034
Tetrastigma pyriforme Wen! 1006
Tetrastigma laevigatum Wen10131
Tetrastigma siamense Wen7485
“t Tetrastigma delavayi Wen7443
Tetrastigma lenticellatum Wen10597
Tetrastigma ceratopetalum Wen10870
Tetrastigma rumcispermum Tibet2003
Tetrastigma ellipticum Wen8260
Tetrastigma laxum Wen8314
Tetrastigma sp. nov. Deden976
Tetrastigma campylocarpum Wen10517
Tetrastigma curtisii Wen10277
Tetrastigma glabratum Wen10670
Tetrastigma lawsoni Wen7505
Tetrastigma cf, tuberculatum Wen10280
Tetrastigma hemsleyanum Wen10792
Tetrastigma bioritsense Wen9451
Tetrastigma laoticum Wen 10969
Tetrastigma voinierianum Wen7320
Tetrastigma sichouense Wen]0547
Tetrastigma obovatum Wen!0567
Tetrastigma wangii Wen&455
Tetrastigma funingense Wen!0579
Tetrastigma lanyuense Wen9404
Tetrastigma pachyphyllum Wen8319
Tetrastigma cruciatum Wen7486
Tetrastigma retinervium Wen10920
Tetrastigma tonkinense Wen7401
Tetrastigma cauliflorum Wen10521
Tetrastigma sp. Wen8465
Tetrastigma heterophyllum Wen10926
Tetrastigma planicaule Wen]0904
Tetrastigma jinghongense Wen8471
Tetrastigma beauvaisii Wen7419
Tetrastigma hookeri Wen8381
Tetrastigma tuberculatum Wen6668
Tetrastigma tuberculatum Wen7319
Tetrastigma tuberculatum Wen8335
Tetrastigma godefroyanum Wen6S75
— Tetrastigma henryi var. mollifolium Wen10S32
Tetrastigma henryi var. henryi Wenl0S18
Tetrastigma garrettii Wen7490
Tetrastigma gaudichaudianum Wen10939
Tetrastigma eberhardtii Wen10945
Tetrastigma apiculatum Wen10940
— Tetrastigma erubescens Wenl0604
Tetrastigma apiculatum Wen10S70
Tetrastigma caudatum Wen]0812
Tetrastigma diepenhorstii Wen8261
Tetrastigma strumarum Wen10757
Tetrastigma sp. Wen10768
Cayratia japonica Shui8 1847
vratia trifolia Wen10167
phostemma maranguense 19790047
phostemma simalans Gernathsn
Cyphostemma horombense Wen9506
Cayratia imerinensis Wen9S71
Cayratia triternata Wen9664
Cissus nodosa Wen10713
Cissus wenshanensis Shui81897
sus repens Shui81807
Cissus erosa Wen8S86
Cissus verticillata Wen8698
Cissus incisa Wen9727
Cissus subtetragona Wen1092)
Cissus hastata Wen10993
Vitis rotundifolia Wen9972
Vitis popenoei Wen8724
Vitis flexuosa Wen 10647
Vitis aestivalis Wen 10004
Ampelocissus elephantina Wen9S83
Parthenocissus vitacea Wen 0488
Parthenocissus quinquefolia Nie& Meng394
Ampelopsis cantaniensis Wen10242
ricocne | tocre__fooocere] wiccee | oF
90.0 80.0 700 60.0 500 400 300 200 100 #£0.0
8545 |mya
Fe os oe
Se eee ee)
Fig. 2. Chronogram of 7efastigma and its relatives from Vitaceae based on combined plastid
data (atpB-rbcL, rps16, trnL-F, and psbA-trnH) inferred from BEAST. Gray bars represent the
95% high posterior density credibility interval for node ages. Calibration points are indicated
with black asterisks.
Biogeography of Tetrastigma 38)
E was sister to Clade D with good Bayesian posterior probability support (node 5 in
Fig. 2). Clade D only contained central Malesian species of 7etrastigma. However, the
sister Clade E contained two biogeographically disjunct species, 7 campylocarpum
Planch. distributed from India to the Sino—Himalayan and Indochina region and 7.
curtisii (Ridl.) Suess. & Suess. of the Malesian region. The crown age of Clade E was
estimated to be 11.0 mya (95% HPD: 3.1—20.8 mya, node E in Fig. 2). The ancestral
area of Clade E was inferred to be widespread in the Sino—Himalayan and Indochina
region and the Malesian region (node E in Fig. 3). Clade F included seven species
from the Sino—Himalayan and Indochina region and three species from the Malesian
region (7: loheri Wen 10202 and T. pisicarpum Wen 10185 from SE Sulawesi and T.
laevigatum Wen 10131 from West Java). The east and central Malesian 7: diepenhorstii
Wen 10812 and west Malesian endemic T. strumarum Wen 10757 and T. sp. Wen10768
constituted Clade H (Fig. 3). The crown age of Clade H was estimated to be 14.8 mya
(95% HPD: 4.2—25.6 mya, node H in Fig. 2) in the middle Miocene.
Reconstruction of ancestral areas with LAGRANGE suggested an ancestral
distribution and early diversification of Tetrastigma in the Sino—Himalayan and
Indochina region in Fig. 3. Subsequently, 7etrastigma species were dispersed from
continental Asia to the Malesian region. The colonisation of the Sino—Himalayan and
Indochina region occurred at nodes A and C and the colonisations of the Malesian
region were supported by several nodes: B, D, G, and H (Fig. 3). The most widespread
ancestral range appeared at nodes E and F (Fig. 3).
Table 1. Prior probability and posterior distribution estimates for 7efrastigma within the
phylogenetic framework of Vitaceae. Mean dates were used as the divergence time of the nodes.
Node constrained
Posterior distribution Mean 95% HPD
and estimated (mya) (mya)
Meenici ~36 363-653 —
Node 2 36.9 25.7-49.3
Node 3 31.0 21.1-42.2
Node 4 25.5 16.1—36.1
Node 5 19.1 9.5-29.5
Node A 20.2 12.5—28.8
Node B DY) 2.5-18.5
Node C 23.0 13.8-32.8
Node D 7.8 2.2-14.5
Node E 11.0 9.4-29.5
Node F 20.3 11.6—29.5
Node G 15.4 5.4-27.0
Node H 14.8 4.2-25.6
314 Gard. Bull. Singapore 63(1 & 2) 2011
Leea spinea Wen9575
Leea monticola Wen9569
Leea indica Wen10910
Cayratia mollissima Wen&403
Cayratia geniculata Wen10275
Cayratia cordifolia Wen 0548
Cayratia pedata Wen7428
Tetrastigma pedunculare Wen]0281
Tetrastigma brunneum Wen8240
Tetrastigma trifoliolatum Wen8350
Tetrastigma papillosum Wen8401
Tetrastigma obtectum Nie& Meng433
Tetrastigma triphyllum Wen 10655
Tetrastigma yunnanense Nie2003104
Tetrastigma napaulense Nie&ZhuS48
Tetrastigma serrulatum Wen7429
Tetrastigma napaulense Tibet225
Tetrastigma serrulatum Wen10856
Tetrastigma loheri Wen!0202
Tetrastigma pisicarpum Wen]0185
Tetrastigma annamense Wen/ 1034
Tetrastigma pyriforme Wen! 1006
Tetrastigma laevigatum Wen/0131
Tetrastigma siamense Wen7485
Tetrastigma delavayi Wen7443
Tetrastigma lenticellatum Wen10597
Tetrastigma ceratopetalum Wen/0870
Tetrastigma rumcispermum Tibet2003
Tetrastigma ellipticum Wen8260
Tetrastigma laxum Wen8314
Tetrastigma sp. nov. Deden976
Tetrastigma campylocarpum Wen10517
Tetrastigma curtisii Wen 10277
Tetrastigma glabratum Wen10670
Tetrastigma lawsoni Wen7505
Tetrastigma cf. tuberculatum Wen10280
Tetrastigma hemsleyanum Wen10792
abla Tetrastigma bioritsense Wen9451]
———. Tetrastigma laoticum Wen10969
Tetrastigma voinierianum Wen7320
aT Tetrastigma sichouense Wen10547
Tetrastigma obovatum Wen10567
Tetrastigma wangii Wen8455
Tetrastigma funingense Wen]0579
4 alb Tetrastigma lanyuense Wen9404
Tetrastigma pachyphyllum Wen8319
Tetrastigma cruciatum Wen7486
Tetrastigma retinervium Wen!0920
Tetrastigma tonkinense Wen740]
Tetrastigma cauliflorum Wen10521
Tetrastigma sp. Wen8465
Tetrastigma heterophyllum Wen!10926
Tetrastigma planicaule Wen/0904
Tetrastigma Jinghongense Wen8471
weoogo
cs
oo
cs
gocccccrc cseocccococcococowwoseownwneaooroocwcocrweocoroscooocweod
AS Tetrastigma beauvaisii Wen7419
bib Tetrastigma hookeri Wen8381 ab
Tetrastigma tuberculatum Wen6668 ab
ala Tetrastigma tuberculatum Wen7319 ab
Tetrastigma tuberculatum Wen8335 ab
Tetrastigma godefroyanum Wen6375
Tetrastigma henryi var. mollifolium Wenl0532
Tetrastigma henryi var. henryi Wen /0S18
Tetrastigma garrettii Wen7490
Tetrastigma gaudichaudianum Wen!0939
Tetrastigma eberhardtii Wen10945
Tetrastigma apiculatum Wen10940
Tetrastigma erubescens Wen10604
Tetrastigma apiculatum Wen10570
Tetrastigma caudatum Wen/0812
Tetrastigma diepenhorstii Wen826/
H ala Tetrastigma strumarum Wen10757
Tetrastigma sp. Wen/0768
—— Cayratia japonica Shui8 1847
Cayratia trifolia Wen10167
Cyphostemma maranguense 19790047
: CL Cyphostemma simalans Gernathsn
Cyphostemma horombense Wen9S06
Cayratia imerinensis Wen9571
Cayratia triternata Wen9664
Cissus nodosa Wen!0713
Cissus wenshanensis Shui8 1897
Cissus repens Shui8 1807
Cissus erosa Wen8S86
Cissus verticillata Wen8698
Cissus incisa Wen9727
Cissus subletragona Wen!/092]
Cissus hastata Wen10993
Vitis rotundifolia Wen9972
Vitis popenoei Wen8724
Vitis flexuosa Wen 10647
Vitis aestivalis Wen10004
Ampelocissus elephantina Wen9583
coocccocoeccso
oc
Parthenocissus vitacea Wen 10488
Parthenocissus quinquefolia Nie& Meng394
Ampelopsis cantaniensis Wen10242
caanacraaccraaaecrennnnnag aoe
Fig. 3. Results of the LAGRANGE analysis of 7etrastigma and its close relatives. The tree
was based on a 50% majority-rule consensus tree of a Bayesian Markov chain Monte Carlo
(MCMC) analysis of the combined plastid data set. Estimated ancestral areas of the clades
(Clades A-H) within Tetrastigma are shown on the nodes using circles (see details in the
Discussion). The four areas of endemism are: a, the Malesian region; b, Sino—Himalayan and
Indochina region; c, Africa; and d, New World. A slash in the result of LAGRANGE indicates
the split of areas into two daughter lineages, i.e., left/right, where “up” and “down” are the
ranges inherited by each descendant branch.
Biogeography of Tetrastigma 315
Discussion
Ancestral area and divergence times of Tetrastigma
Tetrastigma \s distinctive in comparison with other genera of Vitaceae in its four-lobed
stigma in female flowers and dioecious sexual system. Our phylogenetic analyses
have suggested that 7efrastigma is nested within the genus Cayratia Juss. (also see
Chen et al. 2011). Tetrastigma was estimated to have diverged from its closest relative
in the early Eocene (50.6 mya, 95% HPD: 36.3—65.3 mya, node 1 in Fig. 2). The
genus Cayratia is widely distributed in the Old World from Africa throughout Asia
to Oceania. Cayratia has been shown to be paraphyletic (Rossetto et al. 2001, 2002,
2007; Ingrouille et al. 2002; Soejima & Wen 2006; Wen et al. 2007; Ren et al. 2011).
Our analyses has divided the genus into three geographic clades: the African clade, and
two clades in temperate to tropical Asia extending to Australia. 7etrastigma is sister to
the clade including Cayratia japonica (Thunb.) Gagnep. and C. trifolia (L.) Domin,
largely from temperate to tropical Asia. Thus, 7etrastigma may have originated in the
Sino—Himalayan and Indochina region in the early Eocene.
The stem group of 7etrastigma is estimated to be 50.6 mya (95% HPD: 36.3—
65.3 mya, node | in Fig. 2) approximately in the early Eocene. The crown of extant
Tetrastigma 1s estimated to be 36.9 mya (95% HPD: 25.7-49.3 mya, node 2 in Fig. 2).
The divergence of 7etrastigma species between the Sino—Himalayan and Indochina
region and the Malesian region may have begun in the late Eocene. In our Bayesian
analyses, at the beginning of Tetrastigma diversity, there are two Malesian clades
(clades G and H in Fig. 2), and the Sino—Himalayan and Indochina and Malesian
clades (clades A-F in Fig. 2). Improved understanding of SE Asia’s physical history
(Hall 2001, 2002; Metcalfe 2001; van Welzen 2005) suggests that only west Malesia
(Malaysia, Sumatra, Borneo, Java, part of Sulawesi) was above water before the early
Eocene. By the Miocene, the other islands (such as part of Sulawesi, Moluccas, part
of the Lesser Sunda islands, and New Guinea) in the Malesian region got into their
approximate present-day positions. With Jetrastigma fruits being dispersable berries,
the migrations of 7etrastigma between the Sino—Himalayan and Indochina region and
the Malesian region could have occurred at least three times, according to our results.
Biogeographic evolution of Tetrastigma in the Malesian region
Tetrastigma species from the Malesian region are scattered in six major
clades (clades B, D, E, F, G, and H in Fig. 2). Some of the west and central Malesian
Tetrastigma samples are nested within the clades of Sino—Himalayan and Indochina
region (clades B, E, and F in Fig. 3). Thus, migrations of 7etrastigma species between
the Sino—Himalayan and Indochina region and west Malesia may have occurred at
least three times from the late Oligocene to the middle Miocene (Fig. 2). During these
periods with climatic and sea level changes, species of 7etrastigma seem to have
dispersed from the Sino—Himalayan / Indochina region to the Malesian region and
also from the Malesian region back to the Sino—Himalayan / Indochina region.
Three species (7: glabratum Wen 10670, T. lawsoni Wen 7505, T. tuberculatum
Wen 10280) in Clade B are widespread in Indochina and both west and central
316 Gard. Bull. Singapore 63(1 & 2) 2011
Malesia. The divergence time between these two disjunctive clades is estimated to
be 25.5 mya (95% HPD: 16.1—36.1 mya, node 4 in Fig. 2) in the late Oligocene and
early Miocene. Throughout the Oligocene, a land connection between Borneo and
Indochina was hypothesised to have existed (Pupilli 1973, Lloyd 1978). Tetrastigma
species of the Sino-Himalayan and Indochina region may have dispersed into west
and central Malesia through the land bridge or via direct long-distance dispersal.
Furthermore, Clade F included seven species from the Sino—Himalayan and Indochina
region as well as species from west and central Malesia (7. loheri Wen 10202 and T.
pisicarpum Wen 10185 from SE Sulawesi and 7: laevigatum Wen10131 from West
Java). Although the phylogeny did not well resolve the relationships within Clade F,
the clade apparently represents another biogeographc connection between continental
Asia and the Malesian region. The Oligocene and earliest Miocene were periods with
much drier and cooler climates; a major climatic change occurred in the early Miocene
(Morley & Flenley 1987), a period with markedly warm and moist climatic conditions
through a large part of SE Asia and East Asia (also see Morley 1998). During this
period, the ancestor of Clade A may have diversified rapidly in the Sino—Himalayan /
Indochina region, with 50% of Tetrastigma species now endemic to this region (see Li
1998: Chen et al. 2011).
Dispersal of 7etrastigma ancestors between west and central Malesian region
and the Sino—Himalayan / Indochina region may also be shown in Clade G and node 3
(Fig. 3). In Clade G, the basally branched taxon 7: papillosum (BI.) Planch. is widely
distributed from southwestern China, Thailand, Southeast Asia to New Guinea. The
other three species in this clade are distributed in three areas of Malesia. Species in Clade
D included two endemic species of the Philippines (7. ellipticum and T. laxum) and a
new species collected from SE Sulawesi. The clade thus showed a close biogeographic
connection between the Philippines and Sulawesi. During the Neogene, a volcanic
arc existed along the north arm of Sulawesi and possible island chain connections
may have existed to the Philippines (Moss & Wilson 1998). Moreover, many authors
(Duffels 1990, Musser 1987, Balgooy 1987) have noted that the flora and fauna of the
south arm of Sulawesi are different from the rest of Sulawesi, which may be explained
by the geologic evidence that during much of the Tertiary, South Sulawesi was below
sea level. But the rifting and rotating of South Sulawesi (including SE Sulawesi at
present) and its accretion to North Sulawesi just occurred 10 mya ago (Hall 1998).
With a global warm phase during the middle Miocene (Fulthorpe & Schlanger 1989),
it is possible for migration through those island chains and connection lands between
the Philippines and Sulawesi, i.e., west Malesia and central Maleisa. Species endemic
to central Malesia formed Clade D, sister to Clade E (Fig. 3) and nested with the Sino—
Himalayan and Indochina region species. The ancestral area of Clade E is inferred
to be the Sino—Himalayan and Indochina region and Malesian region (node E in Fig.
3). The crown age of Clade E is estimated to be 11.0 mya (95% HPD: 3.1—20.8 mya,
node E in Fig. 2) in the middle Miocene. Two species are included in Clade E (Fig.
3), TZ. campylocarpum Planch. widely distributed from India to the Sino—Himalayan /
Indochina region and 7 curtisii (Ridl.) Suess. & Suess. occurring in west and central
Malesia. Thus, the possible dispersal route of Zetrastigma species is most likely from
Biogeography of Tetrastigma aM 7/
the Sino—Himalayan/Indochina region to west Malesia in the early Oligocene (node 3
in Fig. 2) and then to central Malesia after the middle Miocene (node 5 in Fig. 2).
Clade H contains the east Malesian and Papuan species of Jetrastigma
(Fig. 3). The basally branching species of Clade H is T. deipenhorstii (Miq.) Latiff,
a widespread species in west Malesia. It is probable that the east Malesian species
dispersed from west Malesia. Our results are consistent with routes from the Sino—
Himalayan / Indochina region to west Malesia and from west to central Malesia.
We need to test our hypotheses of 7Jetrastigma biogeography with additional
sampling in both the Sino-Himalayan / Indochina region and the Malesian region and
with improved resolution of the phylogeny and finer division of the areas of endemism.
Nevertheless, our initial results add another case study on the complex biogeographic
origins of Malesian plant taxa, and support the idea that continental Asia served as an
important source area for the Malesian members of the genus.
ACKNOWLEDGEMENTS. We thank W.-H. Chen, J. Gerrath, Z.-L. Nie, Y. Meng, Deden
Girmansyah, R. Li, H. Li, Y.-M. Shui, W.-D. Zhu, and Y.-F. Deng for help with collecting
leaf material, and Z.-L. Nie, H. Sun, S.-L. Zhou, Y.-M. Shui, W.-H. Chen, Nguyen Tiep Hiep,
Nguyen Quang Hieu, Leng-Guan Saw, Elizabeth Widjaja, Harry Wiriadinata, Scott Hoover,
Abdul Kartonegoro, Deden Girmansyah, Arief Hidayat, and Eko Walujo for facilitating field
work and/or providing field assistance. Support for the study was provided by the National
Science Foundation (DEB 0743474 to S.R. Manchester and J. Wen), the Smithsonian
Endowment Grant Program, the Small Grant Program of National Museum of Natural History
of the Smithsonian Institution, the John D. and Catherine T. MacArthur Foundation, and the
College of Horticulture and Forestry of Central China Agricultural University.
References
Baker, W.J., Coode, M.J.E., Dransfield, J., Dransfield, S., Harley, M.M., Hoffmann, P.
& Johns, R.J. (1998) Patterns of distribution of Malesian vascular plants. In: Hall,
R. & Holloway, J.D. (eds) Biogeography and Geological Evolution of SE Asia,
pp. 243-258. Leiden, The Netherlands: Backhuys Publishers.
Balgooy, M.M.J. van (1987) A plant biogeographical analysis of Sulawesi. In:
Whitmore, T.C. (ed) Biogeographical Evolution of the Malay Archipelago. Oxford
Monographs on Biogeography No. 4, pp. 94-102. Oxford: Clarendon Press.
Barkman, T.J., Lim, S.H., Salleh, K.M. & Nais, J. (2004) Mitochondrial DNA sequences
reveal the photosynthetic relatives of Rafflesia, the world’s largest flower. Proc.
Natl. Acad. Sci. U.S.A. 101: 787-792.
Barkman, T.J.. McNeal, J.R., Lim, S.H., Coat, G., Croom, H.B., Young, N.D. &
dePamphilis, C.W. (2007) Mitochondrial DNA suggests at least 11 origins of
parasitism in angiosperms and reveals genomic chimerism in parasite plants.
BMC Evol. Biol. 7: 248.
318 Gard. Bull. Singapore 63(1 & 2) 2011
Bell, C.D., Soltis, D.E. & Soltis, P.S. (2010) The age and diversification of the
angiosperms re-revisited. Amer. J. Bot. 97: 1296-1303.
Briggs, J.C. (1974) Marine Zoogeography, p. 475. New York: McGraw-Hill.
Butlin, R.K., Walton, C., Monk, K.A. & Bridle, J.R. (1998) Biogeography of Sulawesi
grasshoppers, genus Chitaura, using DNA sequence data. In: Hall, R. & Holloway,
D.J. (eds) Biogeography and Geological Evolution of Southeast Asia, pp. 355—
359. Leiden, the Netherlands: Backhuys Publishers.
Cevallos-Ferriz, S.R.S. & Stockey, R.A. (1990) Permineralized fruits and seeds from
the Princeton chert (Middle Eocene) of British Columbia: Vitaceae. Canad. J. Bot.
68: 288-295.
Chandler, M.E.J. (1925) The Upper Eocene Flora of Hordle, Hants. London:
Palaeontographical Society.
Chandler, M.E.J. (1957) The Oligocene flora of the Bovey Tracey Lake Basin,
Devonshire. Bull. Brit. Mus. (Nat. Hist.) 3: 103—106.
Chandler, M.E.J. (1961a) The Lower Tertiary Floras of Southern England. I. Paleocene
floras, London Clay Flora (Suppl.). London: British Museum (Natural History).
Chandler, M.E.J. (1961b) Flora of the Lower Headon Beds of Hampshire and the Isle
of Wight. Bull. Brit. Mus. (Nat. Hist.) 5: 133-135.
Chandler, M.E.J. (1962) The Lower Tertiary Floras of Southern England. II. Flora
of the Pipe-clay Series of Dorset (Lower Bagshot). London: British Museum
(Natural History).
Chandler, M.E.J. (1963) The Lower Tertiary Floras of Southern England. III. Flora of
Bounrnemouths Beds, the Boscombe, and the Highcliffe Sands. London: British
Museum (Natural History).
Chandler, M.E.J. (1964) The Lower Tertiary Floras of Southern England. IV. A
Summary and Survey of Findings in Light of Recent Botanical Observations.
London: British Museum (Natural History).
Chandler, M.E.J. (1978) Supplement to the Lower Tertiary Floras of Southern England,
Part 5, p. 47. London: The Tertiary Research Group.
Chen, I. (2009) History of Vitaceae Inferred from Morphology-based Phylogeny and
the Fossil Record of Seeds. Doctoral dissertation, Univ. of Florida, Gainesville.
Chen, I. & Manchester, S.R. (2007) Seed morphology of modern and fossil Ampelocissus
(Vitaceae) and implications for phytogeography. Amer. J. Bot. 94: 1534-1553.
Chen, P.T., Chen, L.Q. & Wen, J. (2011) The first phylogenetic analysis of Tetrastigma
(Miq.) Planch., the host of Rafflesiaceae. Taxon 60: 499-512.
Clouse, R.M. & Giribet, G. (2010) When Thailand was an island — the phylogeny and
biogeography of mite havestmen (Opiliones, Cyphophthalmi, Stylocellidae) in
Southeast Asia. J. Biogeogr. 2010: 1114-1130.
Collinson, M.E. (1983) Fossil Plants of the London Clay. London: Palaeontol. Assoc.
Davis, C.C. & Wurdack, K.L. (2004) Host-to-parasite gene transfer in flowering plants:
phylogenetic evidence from Malpighiales. Science 305: 676-678.
Dorofeey, P.I. (1957) Seeds of Ampelopsis from the Tertiary deposits of the territory of
USSR. Bot. Zurn. (Moscow & Leningrad) 42: 643-648.
Biogeography of Tetrastigma 319
Dorofeev, P.I. (1963) Tertiary Floras of Western Siberia. Moscow: Academy of
Sciences of the U.S.S.R., Komarov Botanical Institute.
Drummond, A.J. & Rambaut, A. (2007) BEAST: Bayesian evolutionary analysis by
sampling trees. BMC Evol. Biol. 7: 214.
Drummond, A.J., Ho, S.Y.W., Phillips, M.J. & Rambaut, A. (2006) Relaxed
phylogenetics and dating with confidence. PLoS Biol. 4: 699-710.
Duffels, J.P. (1990) Biogeography of Sulawesi cicadas (Homoptera: Cicadoidea). In:
Knight, W.J. & Holloway, J.D. (eds) /nsects and the Rain Forests of South East
Asia (Wallacea), pp 63—72. London: Royal Entomological Society.
Esselstyn, J.A. & Oliveros, C.H. (2010) Colonization of the Philippines from Taiwan:
a multilocus test of the biogeographic and phylogenetic relationships of isolated
populations of shrews. J. Biogeogr. 37: 1504-1514.
Fulthorpe, C.S. & Schlanger, S.O. (1989) Paleooceanographic and tectonic settings of
early Miocene reefs and associated carbonates of offshore Southeast Asia. Amer
Assoc. Petrol. Geol. Bull. 73: 729-756.
Gressitt, J.L. (1982) Pacific-Asian biogeography with examples from the Coleoptera.
Entomol. General. 8: 1-11.
Hall, R. (1998) The plate tectonics of Cenozoic SE Asia and the distribution of land and
sea. In: Hall, R. & Holloway, J.D. (eds) Biogeography and Geological Evolution
of SE Asia, pp. 99-131. The Netherlands, Leiden: Backhuys Publishers.
Hall, R. (2001) Cenozoic reconstructions of SE Asia and the SW Pacific: changing
patterns of land and sea. In: Metcalfe, I., Smith, J.M.B., Morwood, M. & Davidson,
I. (eds) Faunal and Floral Migrations and Evolution in SE Asia-Australasia, pp.
35-56. Lisse: A.A. Balkema.
Hall, R. (2002) Cenozoic geological and plate tectonic evolution of SE Asia and the
SW Pacific: computer-based reconstructions, model and animations. J. Asian
Earth Sci. 20: 353-431.
Ingrouille, M.J., Chase, M.W., Fay, M.F., Bowman, D., Bank, M.V.D. & Bruijn, A.D.E.
(2002) Systematics of Vitaceae from the viewpoint of plastid rbcL DNA sequence
data. J. Linn. Soc., Bot.138: 421-432.
Kirchheimer, F. (1938) Beitraége zur néheren Kenntnis von Vitaceen-Samenformen
tertiaren Alters. Planta 28: 582-598.
Kirchheimer, F. (1939) Rhamnales I: Vitaceae. In: W. Jongmans (ed) Fossilium
Catalogus II: Plantae. Naturwissenschaften 24: 1-174.
Li, C.L. (1998) Vitaceae. In: Flora Reipublicae Popularis Sinicae, vol. 48, pp. 1-177.
Beijing: Science Press.
Liu, D., Ju, J.H., Lin, G., Xu, X.D., Yang, J.S. & Tu, G.Z. (2002) New C-glycosylflavones
from Tetrastigma hemsleyanum (Vitaceae). Acta Bot. Sin. 44: 227-229.
Lloyd, A.R. (1978) Geological evolution of the South China Sea. SEAPEX Proc. 4:
95-137.
Magallon, S. & Castillo, A. (2009) Angiosperm diversification through time. Amer. J.
Bot. 96: 349-365.
Marwick, B. (2009) Biogeography of Middle Pleistocene hominins in mainland
Southeast Asia: A review. Quatern. Int. 202: 51-58.
320 Gard. Bull. Singapore 63(1 & 2) 2011
Meyer, W. (1997) Rafflesiaceae. F/ora Malesiana, ser. 1, 13: 1-42.
Metcalfe, I. (2001) Late Palaeozoic and Mesozoic tectonic and palaeogeographical
evolution of SE Asia-Australasia. In: Metcalfe, I., Smith, J.M.B., Morwood, M.
& Davidson, I. (eds) Faunal and Floral Migrations and Evolution in SE Asia-
Australasia, pp. 15-34. Lisse: A.A. Balkema.
Miki, S. (1956) Seed remains of Vitaceae in Japan. J. Inst. Polytech. Osaka City Univ.
7: 247-271.
Moran, C., Catterall, C.P. & Kanowski, J. (2009) Reduced dispersal of native plant
species as a consequence of the reduced abundance of frugivore species in
fragmented rainforest. Biol. Conser. 142: 541.
Morley, R.J. (1998) Palynological evidence for tertiary plant dispersals in the SE
Asia region in relation to plate tectonics and climate. In: Hall, R. & Holloway, J.
D. (eds) Biogeography and Geological Evolution of SE Asia, pp. 211-234. The
Netherlands, Leiden: Backhuys Publishers.
Morley, R.J. & Flenley, J.R. (1987) Late Cainozoic vegetational and environmental
changes in the Malay archipelago. In: Whitmore, T.C. (ed) Biogeographical
Evolution of the Malay Archipelago. Oxford Monographs on Biogeography No.
4, pp. 50-59. Oxford: Clarendon Press.
Moss, S.J. & Wilson, E.J. (1998) Biogeographic implications from the Tertiary
palaeogeographic evolution of Sulawesi and Borneo. In: Hall, R. & Hoiloway, J.
D. (eds) Biogeography and Geological Evolution of SE Asia, pp. 133-164. The
Netherlands, Leiden: Backhuys Publishers.
Musser, G.G. (1987) The mammals of Sulawesi. In: Whitmore, T.C. (ed)
Biogeographical Evolution of the Malay Archipelago. Oxford Monographs on
Biogeography No. 4, pp. 73-94. Oxford: Clarendon Press.
Nie, Z.L., Sun, H., Chen, Z.D., Meng, Y., Manchester, S.R. & Wen, J. (2010) Molecular
phylogeny and biogeographic diversification of Parthenocissus (Vitaceae) disjuct
between Asia and North America. Amer. J. Bot. 97: 1342-1353.
Paulay, G. (1997) Diversity and distribution of reef organisms. In: Birkeland, C.E. (ed)
Life and Death of Coral Reef, pp. 298-352. London: Chapman & Hall.
Pupilli, M. (1973) Geological evolution of South China Sea area, tentative
reconstruction from borderland geology and well data. In: Indonesian Petroleum
Association Proceedings 2nd Annual Convention |: 223-241.
Rambaut, A. & Drummond A.J. (2004) Tracer v1.3. http://beast.bio.ed.ac.uk/Tracer
Ree, R.H. & Smith, S.A. (2008) Maximum likelihood inference of geographic range
evolution by dispersal, local extinction, and cladogenesis. Syst. Biol. 57: 4-14.
Ree, R.H., Moore, B.R., Webb, C.O. & Donoghue, M.J. (2005) A likelihood framework
for inferring the evolution of geographic range on phylogenetic trees. Evolution
592229923:
Reid, E.M. & Chandler, M.E.J. (1933) The London Clay Flora. London: British
Museum (Natural History).
Ren, H., Lu, L.-M., Soejima, A., Luke, Q., Zhang, D.-X., Chen, Z.-D. & Wen, J. (2011).
Phylogenetic analysis of the grape family (Vitaceae) based on the noncoding
plastid trnC-petN, trnH-psbA, and trnL-F sequences. Taxon 60: 629-637.
Biogeography of Tetrastigma 321
Renner, S.S., Strijk, J.S., Strasberg, D. & Thebaud, C. (2010) Biogeography of the
Monimiaceae (Laurales): A role for East Gondwana and long-distance dispersal,
but not West Gondwana. J. Biogeogr. 37: 1227-1238.
Ronquist, F. (1996) DIVA. Ver. 1.1. Computer Program and Manual. Available by
Anonymous FTP (ftp.sysbot.uu.se) from Uppsala University, Uppsala.
Ronquist, F. (1997) Dispersal—vicariance analysis: A new approach to the quantification
of historical biogeography. Syst. Biol. 45: 195-203.
Roos, M.C. (1993) State of affair regarding Flora Malesiana: progress in revision work
and publication schedule. F/. Males. Bull. 11: 133-142.
Rossetto, M.B., Jackes, R., Scott, K.D. & Henry, R.J. (2001) Intergeneric relationships
in the Australian Vitaceae: New evidence from cpDNA analysis. Genet. Res. Crop
Evol. 48: 307-314.
Rossetto, M., Crayn, D.M., Jackes, B.R. & Porter, C. (2007) An updated estimate of
intergeneric phylogenetic relationship in the Australian Vitaceae. Canad. J. Bot.
85: 722-730.
Rossetto, M., Jackes, B.R., Scott, K.D. & Henry, R.J. (2002) Is the genus Cissus
(Vitaceae) monophyletic? Evidence from plastid and nuclear ribosomal DNA.
Syst. Bot. 27: 522-533.
Soejima, A. & Wen, J. (2006) Phyogenetic analysis of the grape family (Vitaceae)
based on three chloroplast markers. Amer. J. Bot. 93: 278-287.
Taylor, D.W. (1990) Paleobiogeographic relationships of angiosperms from the
Cretaceous and early Tertiary of the North American area. Bot. Rev. 56: 279-417.
Teodoridis, V. (2003) Early Miocene carpological material from the Czech part of the
Zittau Basin. Acta Palaeobot. 43: 9-49.
Thorne, J.L. & Kishino, H. (2002) Divergence time estimation and rate evolution with
multilocus data sets. Syst. Biol. 51: 689-702.
Thorne, J.L., Kishino, H. & Painter, I.S. (1998) Estimating the rate of evolution of the
rate of molecular evolution. Molec. Biol. Evol. 15: 1647-1657.
Tiffney, B.H. & Barghoorn, E.S. (1976) Fruits and seeds of the Brandon Lignite. I.
Vitaceae. Rev. Palaeobot. Palynol. 22: 169-191.
Welzen, P.C. van, Turner, H. & Hovenkamp, P. (2003) Historical biogeography of
Southeast Asia and the West Pacific, or the generality of unrooted area networks
as historical biogeographic hypotheses. J. Biogeogr. 30: 181-192.
Welzen, P.C. van, Slik, J.W. & Alahuhta, J. (2005) Plant distribution patterns and plate
tectonics in Malesia. Biol. Skr. 55: 199-217.
Wen, J. (2007) Vitaceae. In: Kubitzki, K. (ed) The Families and Genera of Vascular
Plants, vol. 9, pp. 466-478. Berlin: Springer-Verlag.
Wen, J., Nie, Z. L., Soejima, A. & Meng, Y. (2007) Phylogeny of Vitaceae based on the
nuclear GA// gene sequences. Canad. J. Bot. 85: 731-745.
Wheeler, E.A. & Lapasha, C.A. (1994) Woods of the Vitaceae—Fossil and modern.
Rev. Palaeobot. Palynol. 80: 175-207.
Wikstrém, N., Savolainen, V. & Chase, M.W. (2001) Evolution of the angiosperms:
Calibrating the family tree. Proc. Roy. Soc. London, B, Biol. Sci. 268: 2211-2220.
322 Gard. Bull. Singapore 63(1 & 2) 2011
Appendix A. Taxa and accessions used for dating and biogeographic analysis of Tetrastigma
and the outgroup taxa of Vitaceae and Leeaceae with their GenBank numbers. “—” indicates
missing data. All voucher specimens are deposited at the US National Herbarium (US).
Collection localities
Species GenBank accessions
(and vouchers) atpB-rbcL, rps16, trnL-F, psbA-trnH
Ampelocissus Madagascar, HMS585516, HM5S85792, HMS585932,
elephantina Planch.
Ampelopsis cantoniensis
Planch.
Cayratia cordifolia C.Y.
Wu ex CLL. Li
Cayratia geniculata
Gagnep.
Cayratia imerinensis
(Baker) Desc.
Cayratia Japonica
(Thunb.) Gagnep.
Cayratia mollissima
Gagnep.
Cayratia pedata
Gagnep.
Cayratia trifolia (L.)
Domin
Cayratia triternata
(Baker) Desc.
Cissus erosa Rich.
Cissus hastata Planch.
Cissus incisa Des Moul.
Cissus nodosa Blume
Cissus repens Lam.
Cissus subtetragona
Planch.
Cissus verticillata (L.)
Nicolson & C.E.
Jarvis
Antsiranana (J. Wen
9583)
Indonesia, SE Sulawesi
(J. Wen 10242)
China, Yunnan (J. Wen
10548)
Indonesia, SE Sulawesi
(J. Wen 10275)
Madagascar,
Antsiranana (J. Wen
9571)
China, Yunnan (¥.-M.
Shui 81836)
Malaysia, Pahang (J.
Wen 8403)
Thailand, Chiang Mai
(J. Wen 7428)
Indonesia, SE Sulawesi
(J. Wen 10167)
Madagascar,
Antsiranana (J. Wen
9664)
Peru, Arequipa (J. Wen
8586)
Vietnam, Danang (J.
Wen 10993)
USA, Texas (J. Wen
9727)
Indonesia, Papua (/.
Wen 10713)
China, Yunnan (¥-M.
Shui 81807)
Vietnam, Ninh Binh (J/.
Wen 10921)
Mexico, Chiapas (J.
Wen 8698)
HMS585659
HM585517, HM585793, HM585933,
HM585660
HMS585518, HM585794, HM585934,
HM585661
HM585519, HM585795, HM585935,
HM585662
HM585520, HM585796, HM585936,
HM585663
HM585521, HM585797, HM585937,
HM585664
HM585522, HM585798, HM585938,
HMS585665
HMS585523, HM585799, HM585939,
HMS585524, HM585800, HM585940,
HM585666
HMS585525, HM585801, HM585941,
HMS585667
HM585526, HM585802, HM585942,
HMS585668
HM585527, HM585803, HM585943,
HM585669
HMS585528, HM585804, HM585944,
HM585670
HMS585529, HM585805, HM585945,
HM585671
HM585530, HM585806, HM585946,
HMS585672
HMS585531, HM585807, HM585947,
HM585673
HMS585532, HM585808, HM585948,
HMS585674
Biogeography of Tetrastigma
Cissus wenshanensis
ey. Li
Cyphostemma
horombense Desc.
Cyphostemma
maranguense (Gilg)
Desc.
Cyphostemma simulans
(C.A.Sm.) Wild &
R.B. Drumm.
Leea indica (Burm.f.)
Merr.
Leea monticola J. Wen
Leea spinea Desc.
Parthenocissus vitacea
(Knerr) Hitche.
Parthenocissus vitacea
(Knerr) Hitche.
Tetrastigma annamense
Gagnep.
Tetrastigma apiculatum
Gagnep.
Tetrastigma apiculatum
Gagnep.
Tetrastigma beauvaisii
Gagnep.
Tetrastigma bioritsense
(Hayata) Hsu & Kuoh
Tetrastigma brunneum
Merr.
Tetrastigma
campylocarpum
Planch.
Tetrastigma caudatum
Merr. & Chun
Tetrastigma cauliflorum
Merr.
China, Yunnan (¥.-M.
Shui 81897)
Madagascar,
Fianarantsoa (J. Wen
9506)
National Botanic
Garden of Belgium
(cult.) (19790047)
USA, Iowa (cult.) (/.
Gerrath s.n.)
Vietnam, Ninh Binh (/.
Wen 10910)
Madagascar,
Antsiranana (J. Wen
9569)
Madagascar,
Antsiranana (J. Wen
9575)
China, Yunnan (cult.)
(Z.-L. Nie & ¥. Meng
394)
Canada, Quebec (J. Wen
10488)
Vietnam, Lam Dong (J.
Wen 11034)
Vietnam, Hoa Binh (/.
Wen 10940)
China, Yunnan (/. Wen
10570)
Thailand, Mae Hong
Son (J. Wen 7419)
China, Taiwan (J. Wen
9451)
Philippines, Luzon (/.
Wen 8240)
China, Yunnan (J. Wen
10521)
Vietnam, Vinh Phuc (./.
Wen 10812)
China, Yunnan (J. Wen
10521)
HM585533, HM585809,
HM585675
HM585534, HM5858 10,
HMS585676
HM585535, HM585811,
HM585677
HM585536, HM585812,
HM585678
HM585537, HM585813,
HM585679
HM585538, HM585814,
HM585539, HM585815,
HM585540, HM5858 16,
HM585680
HM585541, HM585817,
HM585681
HM585543, HM585819,
HMS585683
HM585546, HM585822,
HMS585686
HMS585544, HM585820,
HM585684
HM585547, HM585823,
HM585687
HMS585548, HM585824,
HM585688
HMS585549, HM585825,
HM585689
HMS585552, HM585828,
HMS585692
HMS585551, HM585827,
HM585691
HMS585552, HM585828,
HM585692
323
HM585949,
HM585950,
HM585951,
HM585952,
HM585953,
HM585954,
HM585955,
HM585956,
HM585957,
HMS585959,
HMS585962,
HM585960,
HM585963,
HMS585964,
HMS585965,
HM585968,
HM585967,
HM585968,
Tetrastigma
ceratopelatum C.Y.
Wu
T. cf. tuberculatum
Tetrastigma cruciatum
Craib & Gagnep.
Tetrastigma curtisit
(Ridl.) Suesseng.
Tetrastigma delavayi
Gagnep.
Tetrastigma
diepenhorstii (Miq.)
Latiff
Tetrastigma eberhardtii
Gagnep.
Tetrastigma ellipticum
Merr.
Tetrastigma erubescens
Planch.
Tetrastigma funingense
Ci
Tetrastigma garrettii
Gagnep.
Tetrastigma glabratum
Planch.
Tetrastigma
godefroyvanum Planch.
Tetrastigma
gaudichaudianum
Planch.
Tetrastigma
hemsleyanum Diels &
Gilg
Tetrastigma henryi
(Gagnep.) var. henryi
Gagnep.
Tetrastigma henryi
(Gagnep.) var.
mollifolium W.T.
Wang
Tetrastigma
heterophyllum
Gagnep.
Vietnam, Lao Cai (/.
Wen 10870)
Indonesia, SE Sulawesi
(J. Wen 10280)
Thailand, Chiang Mai
(J. Wen 7486)
Indonesia, SE Sulawesi
(J. Wen 10277)
Thailand, Chiang Mai
(J. Wen 7443)
Philippines, Luzon (J.
Wen 8261)
Vietnam, Ninh Binh (/.
Wen 10945)
the Philippines, Luzon
(J. Wen 8260)
China, Yunnan (J. Wen
10604)
China, Yunnan (J. Wen
10579)
Thailand, Chiang Mai
(J. Wen 7490)
Indonesia, West Java (J.
Wen 10670)
China, Hainan (J. Wen
6575)
Vietnam, Hoa Binh (/.
Wen 10939)
Vietnam, Ninh Binh (/.
Wen 10792)
China, Yunnan (J. Wen
10518)
China, Yunnan (J. Wen
10532)
Vietnam, Ninh Binh (./.
Wen 10926)
Gard. Bull. Singapore 63(1 & 2) 2011
HMS585558, HM585834, HM585974,
HM585698
HM585559, HM585835, HM585975,
HMS585699
HM585562, HM585838, HM585978,
HMS585702
HM585563, HM585839, HM585979,
HM585703
HMS585565, HM585841, HM585981,
HM585705
HMS585567, HM585843, HM585983,
HMS585707
HMS585568, HM585844, HM585984,
HM585708
HM585569, HM585845, HM585985,
HMS585709
HM585570, HM585846, HM585986,
HM585710
HM585574, HM585850, HM585990,
HM585714
HMS585578, HM585854, HM585994,
HMS585718
HM585581, HM585857, HM585997,
HM585719
HMS585583, HM585859, HM585999,
HM585721
HMS585585, HM585861, HM586001,
HM585723
HM585586, HM585862, HM586002,
HMS585724
HM585587, HM585863, HM586003,
HM585725
HMS585588, HM585864, HM586004,
HMS585726
HM585580, HM585856, HM585996,
Biogeography of Tetrastigma
Tetrastigma hookeri
Planch.
Tetrastigma
Jinghongense C.L. Li
Tetrastigma laevigatum
Gagnep.
Tetrastigma lanyuense
C.E. Chang
Tetrastigma laoticum
Gagnep.
Tetrastigma lawsoni
(King) Burkill
Tetrastigma laxum Mert.
Tetrastigma
lenticellatum C.Y. Wu
ex W.T. Wang
Tetrastigma loheri
Gagnep.
Tetrastigma napaulense
Ce )-C.L. Li
Tetrastigma napaulense
we.) C.L. Li
Tetrastigma obovatum
Gagnep.
Tetrastigma obtectum
(Wall.) Planch.
Tetrastigma
pachyphyllum
(Hemsl.) Chun
Tetrastigma papillosum
Planch.
Tetrastigma pedunculare
Planch.
Tetrastigma pisicarpum
(Mig.) Planch.
Tetrastigma planicaule
Gagnep.
Tetrastigma pyriforme
Gagnep.
Tetrastigma retinervium
Planch.
Malaysia, Pahang (J.
Wen 8381)
China, Yunnan (J. Wen
8471)
Indonesia, West Java (J.
Wen 10131)
China, Taiwan (J. Wen
9404)
Vietnam, Guangnam (/.
Wen 10969)
Singapore (J. Wen 7505)
Philippines, Luzon (J.
Wen 8314)
China, Yunnan (J. Wen
10597)
Indonesia, SE Sulawesi
(J. Wen 10202)
China, Xizang (Tibet
225)
Nepal, Kathmandu (Z.-
L. Nie & W-D. Zhu
548)
China, Yunnan (J. Wen
10567)
China, Guizhou (Z.-L.
Nie & Y. Meng 433)
Philippines, Luzon (J.
Wen 8319)
Malaysia, Pahang (J/.
Wen 8401)
Indonesia, SE Sulawesi
(J. Wen 10281)
Indonesia, SE Sulawesi
(J. Wen 10185)
Vietnam, Ninh Binh (J.
Wen 10904)
Vietnam, Lam Dong (J.
Wen 11006)
Vietnam. Vinh Phuc (J.
Wen 10920)
HM585589, HM585865,
HM585727
HM585590, HM585866,
HM585728
HM585591, HM585867,
HM585729
HM585593, HM585869,
HM585731
HM585595, HM585871,
HM585733
HM585599, HM585874,
HM585737
HM585602, HM585877,
HM585740
HM585604, HM585879,
HM585742
HM585605, HM585880,
HM585743
HM585607, HM585882,
HM585745
HM585606, HM585881,
HM585744
HM585608, HM585883,
HM585746
HMS585612, HM585886,
HM585750
HMS585616, HM585891,
HM585753
HM585617, HM585892,
HM585754
HMS585620, HM585895,
HM585757
HM585621, HM585896,
HM585758
HM585622, HM585897,
HM585759
HMS586005,
HM586006,
HMS586007,
HM586009,
HM586011,
HM586015,
HM586018,
HM586020,
HM586021,
HMS586023,
HM586022,
HM586024,
HMS586027,
HM586032,
HM586033,
HM586036,
HM586037,
HMS585623, —, HM586038,
HM585760
HMS585625, HM585899,
HMS585762
HM586040,
326
Tetrastigma
rumicispermum
Planch.
Tetrastigma serrulatum
Planch.
Tetrastigma serrulatum
Planch.
Tetrastigma siamense
Gagnep. & Craib
Tetrastigma sichouense
Cia
Tetrastigma sp. nov.
Tetrastigma sp.
Tetrastigma sp.
Tetrastigma strumarum
Gagnep.
Tetrastigma tonkinense
Gagnep.
Tetrastigma
trifoliolatum Merr.
Tetrastigma triphyllum
(Gagnep.) W.T. Wang
Tetrastigma
tuberculatum (Blume)
Latiff
Tetrastigma
tuberculatum (Blume)
Latiff
Tetrastigma
tuberculatum (Blume)
Latiff
Tetrastigma
voinierianum Pierre
ex Pit.
Tetrastigma wangii J.
Wen
Tetrastigma yunnanense
Gagnep.
Vitis aestivalis Michx.
China, Yunnan (7ibet
2003)
Vietnam, Lao Cai (J.
Wen 10856)
Thailand, Chiang Mai
(J. Wen 7429)
Chiang Mai (J. Wen
7485)
China, Yunnan (J. Wen
10547)
Indonesia, SE Sulawesi
(G. Deden 976)
China, Yunnan (J. Wen
58465)
Indonesia, West Papua
(J. Wen 10768)
Indonesia, Papua (J.
Wen 10757)
Thailand, Chiang Mai
(J. Wen 7401)
Malaysia, Selangor (J.
Wen 8350)
China, Yunnan (J. Wen
10655)
USA, Missouri Bot.
Gard. (cult.) (J. Wen
6668)
USA, Illinois (cult.) (/.
Wen 7319)
Malaysia, Selangor (J.
Wen 8335)
USA, Illinois (cult.) (/.
Wen 7320)
China, Yunnan (J. Wen
8455)
China, Yunnan (cult.)
(Z.-L. Nie 2003104)
USA, South Carolina (/.
Wen 10004)
Gard. Bull. Singapore 63(1 & 2) 2011
HM585626, HM585900, HM586041,
HM585763
HM585628, HM585902, HM586043,
HM585765
HM585629, HM585903, HM586044,
HM585766
HMS585630, HM585904, HM586045,
HM585767
HMS585631, HM585905, HM586046,
HMS585768
HMS585640, HM585913, HM586055,
HM585777
HM585639, HM585912, HM586054,
HM585776
HM585637, HM585910, HM586052,
HM585774
HMS585641, HM585914, HM586056,
HMS585778
HM585642, HM585915, HM586057,
HM585779
HMS585644, HM585917, HM586059,
HMS585781
HM585648, HM585921, HM586063,
HM585783
HMS585649, HM585922, HM586064,
HM585784
HM585650, HM585923, HM586065,
HM585651, HM585924, HM586066,
HM585785
HMS585652, HM585925, HM586067,
HM585786
HMS585653, HM585926, HM586068,
HM585787
HMS85654, HM585927, HM586069,
HM585655, HM585928, HM586070,
HMS585788
Viti flexuosa Thunb. China, Yunnan (J. Wen
10647)
Vi itis popenoei J.L. Mexico, Chiapas (J.
Vitis rotundifolia Michx. USA, Virginia (J. Wen
9972)
a2
HM585656, HM585929, HM586071,
HM585789
HM585657, HM585930, HM586072,
HM585790
HMS585658, HM58593 1, HM586073,
HM585791
; —— S77 Se
7) >
ie = 7 NS
Gardens’ Bulletin Singapore 63(1 & 2): 329-339. 2011 329
The floristic position of Java
Peter C. van Welzen! and Niels Raes
Netherlands Centre for Biodiversity — Naturalis,
Section National Herbarium of The Netherlands, Leiden University,
PO Box 9514, 2300 RA Leiden, The Netherlands
'welzen@nhn.|leidenuniv.nl (corresponding author)
ABSTRACT. Floristically, Java was always considered to be west of Wallace’s line together
with the Malay Peninsula, Sumatra, Borneo, the Philippines and Sulawesi. Recently, statistical
analyses of rough geographic data per species (presence or absence on islands or island groups)
showed that Java is part of the central Wallacean area in Malesia (together with also the
Philippines, Sulawesi, the Lesser Sunda Islands, and the Moluccas) rather than of Sundaland
(restricted to the Malay Peninsula, Sumatra, and Borneo). More precise distribution maps for
Java with collecting localities show that most species are widespread over Java or show a
(more) western distribution; few species show an eastern distribution. The distributions show
strong correlations with altitude (mountain species) and with precipitation (roughly wet in the
west, dry in the east). The expectation was to find mainly species with a drought preference
(Wallacean). However, most species show a preference for a wet distribution, which is related
to a Sunda distribution. The fact that the statistical tests used for the first database show a
Wallacean connection for Java probably is the result of the relative values these test use instead
of absolute numbers, e.g., the resemblance between, especially, the flora of the Lesser Sunda
Islands with Java is very high.
Keywords. Flora Malesiana, floristics, Java, Malesia
Introduction
In 1859 Wallacea introduced his famous zoological boundary, Wallace’s line (Huxley
1868), that divided the Malay Archipelago (or Malesia: Steenis 1950; Raes & Welzen
2009) into an eastern and western part. Wallace’s line runs east of the Philippines, then
either west (Wallace 1859, 1863—1876) or east (Wallace 1860, 1910) of Sulawesi (also
known as Celebes), and ends between Bali and Lombok in the Lesser Sunda Islands.
Wallace discussed the position of Sulawesi in his book ‘Island life’ (Wallace 1880),
in which he calls Sulawesi an ‘anomalous island’ with no continental connections as
Sulawesi lacks Sundaic groups and contains (old) endemic and Australasian species. A
more complete historical overview is presented in Simpson (1977) and George (1981),
who both show that a number of variants of Wallace’s line have been proposed based
on the study of different groups of organisms (Fig. 1). The area encompassed by these
lines is often called Wallacea, a term coined by Dickerson (1928), for an area already
delimited by Wallace in 1863. The areas to the west (Malay Peninsula, Sumatra, Java,
Borneo) and to the east (New Guinea) are referred to as the Sunda Shelf and the Sahul
Shelf, respectively.
330 Gard. Bull. Singapore 63(1 & 2) 2011
ie aay fol
Wallace's lines
Merrill-Dickerson
a / Huxley line
Zollinger’s line
Weber's line ‘
Lydekker’s line
|
i i border Wallacea —
Sahul Shelf
!
I
1
I
Pag
77
Pe
.
CT le he ees a LD
pete PTT
border Wallacea —
Sunda Shelf
Fig. 1. Interpretations of Wallace’s lines around Sulawesi and various alternatives are depicted
in shades of blue. Red lines show the borders between the Sunda Shelf, Wallacea and the Sahul
Shelf resulting from analyses discussed in the text.
Almost only zoological data were used to distinguish the lines (George 1981),
and unsurprisingly, most botanist did not use the various boundaries. Steenis (1950)
used distributions of plant genera when he contemplated the limits of the Malesian
region, and based on the same data Wallace’s line did not appear to be a distinct boundary
in plant distributions. Welzen et al. (2005) used a limited floristic species database
to show that all lines do form distinctive boundaries in plant distributions, each line
stopping at least twice as many taxa as passing. Welzen & Slik (2009) even indicated
which families are mainly responsible for the distribution patterns in Malesia. Only
20 families determine the patterns. Dipterocarpaceae, Fagaceae, and Nepenthaceae
have their centre of diversity on the Sunda Shelf; Ericaceae, Monimiaceae, and
Sapindaceae are typical for the Sahul Shelf; Araliaceae, Boraginaceae, Convolvulaceae,
Cyperaceae, Dioscoreaceae, Lamiaceae, Loranthaceae, Mimosaceae, and Moraceae
are predominantly Wallacean, and Burseraceae, Caesalpiniaceae, Flacourtiaceae,
Meliaceae, and Myristicaceae show no distinct centre.
Recently, Welzen et al. (2011) analysed the distribution data per island
group of all species published in Flora Malesiana series 1 (Angiosperms) and the
Malesian orchids in Orchids Monograph (for island groups see Fig. 1). Various
phenetic techniques were used: Principal Components Analysis, Fig. 2A; Non-metric
Multidimensional Scaling analysis, Fig. 2B; a cluster analysis, Unweighted Pair
Group Method with Arithmetic mean, Fig. 2C; and Kroeber’s coefficient, the mean
Floristic position of Java 33]
floral similarity between pairs of areas, Fig. 2D. All results show that Java is not west
of Wallace’s line, but east of it (Fig. 2). The tests were repeated for 100 randomly
drawn matrices with equal contributions of all areas (500 species per area), which
should nullify the effect of different island sizes (large islands like Borneo and New
Guinea harbour far more (endemic) species than other areas and are then automatically
separated from the rest in the analyses). Also, the analyses of the 100 matrices showed
the same result for Java, not a part of the Sunda Shelf areas, but part of Wallacea. Thus,
the borders between the three areas, Sunda Land/Shelf, Wallacea and Sahul Land/
Shelf, have to be redrawn a bit (red lines in Fig. 1).
The Kroeber’s Coefficient (Fig. 2D) nicely shows that in fact the botanical
relationship between Java and Wallacea just wins in the other three analyses (PCA,
NMS, UPGMA) from a Sunda relationship, because Java has a high floral mean
resemblance (Fig. 3) with on the one hand Sumatra and the Malay Peninsula (Sunda
Land) and on the other hand a slightly higher mean resemblance with the Lesser
Sunda Islands and Sulawesi (Wallacea). The floral resemblance with Borneo (Sunda)
and the Philippines and Moluccas (Wallacea) is somewhat less. Welzen et al. (2011)
Huxley Line
Fs Huxley Line
Merrill-Dickerson Line Merrill-Dickerson Line
Borneo
Axis 2
New Guinea
Lippi
—_— | Philippines &
Axis 1 ® Malay Peninsula
Malay Pen. © Sumatra
Axis 2
Java 4 i
Sulawesi A . Axis 3 4
4
Lesser Sunda Isl. &
E / New Guinea
” | |
Moluccas
B
Malay Peninsula Sumatra — = 30-39%
New Guinea
Lydekker's Line
Philippines
Moluccas
Sulawesi
Lesser Sunda Isl.
Java
Huxley Line ss
Merrill-Dickerson Line Borneo
Malay Peninsula
Sumatra
& Sorensen’s coefficient —> D
Philippines
Fig. 2. The results of various phenetic tests on a database with presence/absence data per island
group for indigenous species revised in Flora Malesiana and Orchid Monographs. All tests
place Java together with Wallacean areas. A. Principle Components Analysis (PCA). B. Non-
metric Multidimensional Scaling analysis (NMS). C. Unweighted Pair Group Method with
Arithmetic mean (UPGMA). D. Kroeber’s coefficients with a mean floral similarity between
pairs of areas indicated by various thicknesses of the lines connecting the pairs.
332 Gard. Bull. Singapore 63(1 & 2) 2011
a
Fig. 3. Kroeber’s coefficients between Java and the Sunda and Wallacean areas showing a high
resemblance, on the one hand, with Sumatra and the Malay Peninsula on the Sunda Shelf; and,
on the other hand, with the Lesser Sunda Islands and Sulawesi within Wallacea. Blue lines
depict the borders of Wallacea.
discuss that the floral resemblances can largely be explained by the supposed savannah
corridors running from the Malay Peninsula along and over Sumatra to Java during
glacial periods. Via this corridor species that prefer a yearly dry period could disperse
and are now still found in areas with a yearly dry monsoon.
The purpose of this paper is threefold: - establish why plant distributions more
strongly point to a Wallacean relationship for Java; - check which kind of phenetic
distribution patterns exist witin Java; and - explain these.
Materials and methods
In previous studies (Welzen et al. 2005, Welzen & Slik 2009, Welzen et al. 2011), a
database was used whereby the presence and absence of indigenous species revised in
Flora Malesiana series | and Orchid Monographs was noted for the Malay Peninsula,
Sumatra, Java, Borneo, the Philippines, Sulawesi, Lesser Sunda Islands, Moluccas, and
New Guinea. The areas used necessitated two alterations for two lines. The Merrill-
Dickerson variant of the Huxley line runs officially between Palawan and the rest of
the Philippines. This line is now considered to run between Palawan and Borneo (Fig.
1). Also, Wallace’s line ends between Bali and Lombok in the Lesser Sunda Islands;
this is redrawn between Java and Bali (Fig. 1). These redrawn lines are used in this
study.
For this study, a database was created with collecting localities of species that
are represented by digitised herbarium specimens from Java. Up to now, the database
Floristic position of Java 333
contains 97 families and 447 genera. The families starting with A or B are fully digitised,
plus a part of C (26 families), some of the other families have also been digitised (e.g.,
Euphorbiaceae, Rhamnaceae, Sapindaceae, Vitaceae), but most other families are only
partly digitised. The localities were georeferenced with the websites http://earth-info.
nga.mil/gns/html/index.html and http://www. fallingrain.com/world/. Only the species
with five or more different localities were included in the analysis, fewer localities
were considered too incomplete to infer a distribution pattern. In total 808 species
were sorted visually into different distribution patterns. This appeared to be rather
straightforward. Four patterns were discriminated: 1) widespread over the island; 2)
a western distribution (subdivided into 2a purely west, 2b west plus a few dots in
central Java, and 2c west up to central Java; these three were not always distinctive);
3) a (central to) eastern distribution; and 4) a west-and-east-only distribution. All
distributions were divided into low and high (above 1000 m) altitude.
The dot maps were produced with MapInfo Professional 7.0 (© MapInfo
Corporation), the altitudinal and precipitation maps were made with Manifold GIS
(Manifold.net), for which the datasets were obtained from the WorldClim 30 arc-
second dataset (www.worldclim.org).
Results
Table 1 shows how many species were found per pattern. A large part, 49%, of the
sampled plant species is widespread over Java (Fig. 4A). This group was not used in
further analyses as they did not convey any information concerning a possible split
of Java into the Sunda or Wallacean realms. It is possible to split off from this group
plants typical for mangroves and beaches/dunes. A total of 273 species (Table 1) show
a predominantly western distribution (Fig. 4B—D), some only in the extreme west (123
species, Fig. 4B), others with a few specimens in central Java (97 species, Fig. 4C)
Table 1. Numbers of species per pattern, divided for low, high or all altitudes. The western
pattern is subdivided into three sub-patterns (shown in italics). The percentage is the percentage
of 808 species.
Pattern = 1000 alt. > 1000 m alt. All altitudes Total % | Fig.
ieee 6|0~«i C(<i‘(z OO!
West — Central 175 64 34 213 34%
West 76 BiG 10 123 15% ~~ 4b
West (- Central) 61 18 18 97 IPG, We
West — Central 38 9 6 3 7% 4d
West and East 57 14 li 88 11% 4e
Central — East 36 10 5 51 6% 4f
334 Gard. Bull. Singapore 63(1 & 2) 2011
or spread from west up to central Java (53 species, Fig. 4D). The opposite pattern,
(central to) east Java also exists (51 species, Fig. 4F). The most curious distribution 1s
perhaps the west and east distribution (88 species, Fig. 4E), whereby the species are
absent in central Java.
Fig. SA shows the mean amounts of yearly precipitation; western Java up to
the central part, with the exception of the northern rim, plus the areas around the
mountains in east Java are relatively wet, whereas the northern rim and the eastern
“%
(
re ATK, frm _F~ oar
J a vy |
9) : [ee 2 eee O
aAc_. Hn S ren ~
ar, Ne a(R ce Oe
lia eee Cn ~~ See —
= I
aa 3 & > Seg > |
eS ys s Da
Fig. 4. Examples of the different distribution patterns, all low altitude. — A. Maranthes
corymbosa Blume (Chrysobalanaceae/Rosaceae): widespread. B. Pometia pinnata J.R.Forst.
& G.Forst. (Sapindaceae): west. C. Kibara coriacea (Blume) Tul. (Monimiaceae): West (—
Central). D. Cayratia japonica (Thunb.) Gagnep. (Vitaceae): West — Central. E. Epipogium
roseum (D.Don) Lindl. (Orchidaceae): West and East. F. Capparis pubiflora DC. (Capparaceae):
Central — East.
Floristic position of Java 335
half are much drier. Fig. 5B shows the mountains on Java. There is a row of volcanoes
along the central axis of the island with a concentration of mountains in especially the
western part. The low-altitude distributions correlate very well with the precipitation
map (Fig. 5A). The western distributions are present in the wetter areas (demonstrated
by the red dots of Kibara coriacea; Fig. SA). The same correlation with wet areas is
shown by the west and east distributions (grey dots of Epipogium roseum; Fig. 5A), in
the west they are in the wet areas, in the east in the wet areas around the mountains. The
eastern patterns show a correlation with low amounts of rain (blue dots of Capparis
pubiflora; Fig. 5A).
The high-altitude distributions show a good correlation with the altitudinal
map. Fig. 5B shows with red dots the distribution of the widespread Sarcococca
pruniformis Lindl., and with blue dots the west and east distribution of Dendrobium
tetraedre (Blume) Lindl. These species are present on the slopes of the mountains and
thus also in areas with a higher precipitation.
ioe aoe f0e" fO9° aoe Tio iz" ais 114°
oa
. 5189
| of , > Mae 4135
| 3081
° °
rm | 2 |) 2027
973
° °
<a ©
°o
pe ee ee? 6 OO «6? 6s?’ 64%
10625 1072 <1ose 109° 110 D1? 2a Js? wWi4°
io ane" is” “we® To? 11" 12° Ts" mW. B
Fig. 5. A. Map of Java showing the mean annual precipitation in mm, ranging from 973 mm per
year (brownish yellow) to 5189 mm (dark green). Red dots represent the distribution of Kibara
coriacea (Blume) Tul. (Fig. 4C), green dots are Capparis pubiflora DC. (Fig. 4F); and grey
dots are Epipogium roseum (D.Don) Lindl. (Fig. 4E). B. Map of Java showing altitude in m;
red dots: Sarcococca pruniformis Lindl. (Buxaceae); blue dots: Dendrobium tetraedre (Blume)
Lindl. (Orchidaceae); both are species of higher altitudes.
886 Gard. Bull. Singapore 63(1 & 2) 2011
Discussion
The Sunda Shelf is characterised by an everwet climate, while most areas in Wallacea
have a dry monsoon during part of the year (Steenis 1979). Thus, the distributions
correlating with high(er) precipitation, the western (Fig. 4B—D; 5A: red dots), west
and east (Fig. 4E, 5A: grey dots) and higher altitudinal distributions (Fig. 5B) point
at a Sunda element in the flora of Java, while the distributions correlated with a
dry monsoon, the eastern distributions (Fig. 4F, 5A: red dots), show the Wallacean
influence. From Table | it follows that most species show a Sunda distribution pattern,
while in the introduction it was explained that the highest floristic bonds are with the
Wallacean areas (Fig. 2). How can we explain this apparent discrepancy?
The sample used might be too small. Up to now only the Javanese collections
of the families starting with “A” or “B” have been fully digitised and georeferenced.
These families might be biased towards a western distribution. The database will be
extended in the future to cover all Javanese collections. The bias is probably absent,
because Araliaceae and Boraginaceae, representing typical Wallacean elements
(Welzen & Slik 2009; see introduction), are already included in the sample.
Dot maps only show places that were visited by collectors. All distributions are
incomplete, because many places have never been visited or sampled. This problem can
be overcome by applying species distribution or ecological niche modelling (e.g., Raes
2009), whereby environmental variables are correlated with the conditions present at
the collecting localities and these are extrapolated to possibly suitable, non-sampled
areas. These models may perhaps show a more Wallacean correlation. However, this
is not expected as the dot maps already infer such a high correlation with precipitation
and altitude.
One might argue that redrawing Wallace’s line between Java and Bali, instead
of between Bali and Lombok, caused the close floristic bonds between Java and the
Lesser Sunda Islands. A possible high floral resemblance between Java and Bali may
obscure the gap in floral elements between Bali and the rests of the Lesser Sunda
Islands. However, this is not the case, because Bali is (as far as plants are concerned)
much under-sampled in comparison to Java and the two other provinces in the Lesser
Sunda Islands. The specimen database in the Leiden herbarium (L) shows 44,038
specimens for Java, only 728 for Bali, and 6,180 for the rest of the Lesser Sunda
Islands.
Table 2 provides another explanation. Java shares a very high percentage of its
flora with Sumatra and the Malay Peninsula (74% with each), but from the viewpoint
of Sumatra and the Malay Peninsula this is far less (48% and 39%, respectively). For
the Lesser Sunda Islands and Sulawesi this is different, they share 78% and 56% of
their flora, respectively, with Java, while slightly more than 50% of the Javanese flora
is present in the Lesser Sunda Islands and Sulawesi. Thus, based on percentages the
shared flora between Java and the Wallacean areas is higher than with Sunda Land.
However, when total numbers of species are compared (first column of Table 2), then
Java shares more species with Sumatra and the Malay Peninsula than with Wallacea.
The statistical tests discussed in the introduction (Welzen et al. 2011; Fig. 2) use
Floristic position of Java 337
Table 2. Floristic overlap between Java and the various other islands: the first column shows
the number of shared species between Java and one of the other regions; the second column the
percentage overlap from the perspective of the other region; and the third column the percentage
overlap from the perspective of Java. Thus Java and Sumatra share 999 species, which is 48% of
Sumatra’s flora and 74% of Java’s flora. The numbers of species were obtained from a database
with presence/absence data per island group for all Malesian indigenous species published in
Flora Malesiana ser. I and in Orchid Monographs.
Region No. = % Region % Java
Sumatra 999 7 48
Malay Peninsula 831 39
Borneo 173 28 57
Lesser Sunda Islands 702 . 78 2
Sulawesi 683 56 51
Philippines 762 4] 57
Moluccas 490 52 36
New Guinea ; 581 ai 20 ; | 43 :
relative numbers in their analyses. Therefore, it is not surprising that the tests placed
Java in the Wallacean realm, while total numbers point at a Sunda connection.
The West and East disjunct distributions are not easily explained. They may
be a result of glacial—interglacial cycles. During glacial periods the sea levels dropped
and altitudinal floral zones on mountains became much lower, probably providing
continuous ranges or stepping stones for dispersal and as a result, continuous, non-
disjunct distributions. During interglacial periods, like present day, sea levels are
much higher, just like the altitudinal floral zones on mountains. Seemingly, especially
in central Java, species have disappeared, perhaps due to adverse conditions on the
central mountains. This may also have happened to species now only restricted to
higher altitudes in west Java: these may have been widespread during glacial periods,
but disappeared in central and east Java during interglacial periods. An alternative
explanation might be that the human influence in especially central Java was much
higher than in east and west (probably not realistic as most people live in west Java) or
that there has been less sampling in especially central Java.
Conclusions
Java shows a kind of floristic Janus head. On the one hand, based on total numbers of
shared species, Java clearly has a Sunda Shelf relationship with especially Sumatra and
the Malay Peninsula (Table 2, column 1). On the other hand, when relative numbers
are used, then the higher resemblance with especially the flora of the Lesser Sunda
338 Gard. Bull. Singapore 63(1 & 2) 2011
Ip
Islands and, to a lesser extent, that of Sulawesi, places Java in the Wallacean realm.
The distributions in Java mainly show a western or eastern component.
The western distributions correlate with a high(er) amount of rainfall, the eastern
distributions with a preference or tolerance for a drier climate. The higher altitudinal
distributions correlate with the presence of volcanoes on Java.
ACKNOWLEDGEMENTS. The first author thanks the organisers of the 8th Flora Malesiana
Symposium for the opportunity to present this paper and the Netherlands Centre for Biodiversity
Naturalis for funding the visit to Singapore during the symposium. The reviewers are sincerely
acknowledged for their suggested improvements.
References
‘Dickerson, R.E. (1928) Distribution of Life in the Philippines. Philippine Bureau of
Sciences Monograph 21.
George, W. (1981) Wallace and his line. In: Whitmore, T.C. (ed) Wallace's Line and
Plate Tectonics. Oxford: Clarendon Press.
Huxley, T.H. (1868) On the classification and distribution of the Alectoromorphae and
Heteromorphae. Proc. Zool. Soc. London, pp. 214-319.
Raes, N. (2009) Borneo, a quantitative analysis of botanical richness, endemicity
and floristic regions based on herbarium records. Thesis, Nationaal Herbarium
Nederland, Leiden (https://openaccess.leidenuniv.nl/handle/1887/13470).
Raes, N. & Welzen, P.C. van (2009) The demarcation and internal division of Flora
Malesiana: 1857—present. Blumea 54: 6-8.
Simpson, G.G. (1977) Too many lines; the limits of the oriental and Australian
Zoogeographic regions. Proc. Amer. Philos. Soc. 121: 107-120.
Steenis, C.G.G.J. van (1950) The delimitation of Malaysia and its main plant
geographical divisions. In: Steenis, C.G.G.J. van (ed) Flora Malesiana Ser. 1, 1:
Ixx—Ixxv. Djakarta: Noordhoff-Kolff n.v.
Steenis, C.G.G.J. van (1979) Plant-geography of east Malesia. Bot. J. Linn. Soc. 79:
97-178.
Wallace, A.R. (1859) Letter from Mr. Wallace concerning the geographical distribution
of birds. /bis 1: 449-454.
Wallace, A.R. (1860) On the zoological geography of the Malay archipelago. J. Linn.
Soc. 14: 172-184.
Wallace, A.R. (1863) On the physical geography of the Malay Archipelago. J. Roy.
Geogr. Soc. 33: 217-234.
Wallace, A.R. (1876) The Geographical Distribution of Animals. 2 vol. London:
MacMillan & Co.
Wallace, A.R. (1880) /sland life. London: MacMillan & Co.
Wallace, A.R. (1910) The World of Life. London: Chapman & Hall.
Floristic position of Java 339
Welzen, P.C. van, Parnell, J.A.N. & Slik, J.W.F. (2011) Wallace’s Line and plant
distributions: two or three phytogeographical areas and where to group Java? Biol.
J. Linn. Soc. 103: 531-545.
Welzen, P.C. van & Slik, J.W.F. (2009) Patterns in species richness and composition of
plant families in the Malay Archipelago. Blumea 54: 166-171.
Welzen, P.C. van, Slik, J.W.F. & Alahuhta, J. (2005) Plant distribution patterns and
plate tectonics in Malesia. Plant diversity and complexity patterns. In: Fries, I. &
Balslev, H. (eds) Local, regional and global dimensions. Biol. Skr. 55: 199-217.
=]; enheanmere ts
é reses ié
Lia)
-_ of ise pe i
i
a
, 7 a
j ; & “Sox i =
: -
“i pers payhs
pe
Gardens’ Bulletin Singapore 63(1 & 2): 341-356. 2011 34]
Colonisation and diversity of epiphytic orchids
on trees in disturbed and undisturbed forests
in the Asian tropics
Mohammed Kamrul Huda’ and Christopher C. Wilcock?
‘Department of Botany, University of Chittagong, Chittagong 4331, Bangladesh
mkhuda70@hotmail.com
*School of Biological Sciences, University of Aberdeen, Aberdeen AB24 3UU, UK
ABSTRACT. Orchids are the most diverse group of epiphytes with more than two thirds of all
their species being epiphytic, yet they are comparatively little studied. Colonisation, diversity
and distribution of epiphytic orchids and their phorophytes (supporting trees) at 72 sites in
disturbed and undisturbed natural forests of south-eastern Bangladesh were studied within
a 21,070 km? area. No evidence of phorophyte specificity was found, but some phorophyte
species were preferred. Most phorophytes (76%) bore only a single species of orchid in one
clump. Both orchid and phorophyte species were diverse but 30% of epiphytic orchid species
were restricted to a single tree. Larger trees and trees in richer orchid areas accumulated more
orchids. Colonisation by an epiphytic orchid is a rare and random event, the presence of one
orchid neither attracting nor repelling others on the same phorophyte. The data suggest that
the frequency of colonisation by epiphytic orchids is primarily a function of the age of the
phorophyte (with greater age allowing both more time and more surfaces to accumulate seeds
on them) and the existing orchid richness of an area (allowing for a higher colonisation rate
from local seed input). Selective logging of the oldest trees in an area would therefore cause a
decline in epiphytic orchid abundance and further loss in orchid richness of the area.
Keywords. Colonisation, diversity, epiphytes, orchids, tropical forests
Introduction
It has been estimated that 24,000 or more vascular plant species are epiphytes (Kress
1986). Epiphytes predominantly occur in tropical and subtropical regions (Benzing
1986). According to Huda et al. (1999), there are 106 epiphytic and 56 terrestrial
orchids in Bangladesh and later on, it was estimated 116 epiphytic, 60 terrestrial, one
saprophytic and one amphibious species of orchid were found (Huda 2007). Orchids
are the most diverse group of epiphytes, about 70% of their species being epiphytic
(Gentry & Dodson 1987, Gravendeel et al. 2004).
The availability of suitable phorophytes (supporting trees) is believed to
have a strong influence on orchid distribution (Withner 1974). Orchid populations are
generally small and composed of scattered or clustered individuals (Ackerman 1986).
The numbers of orchids based on the number of different clumps on phorophytes
have been studied by Johansson (1974) and Catling et al. (1986) but they did not
consider the size of the clump or the height at which they grew. Phorophyte age and
342 Gard. Bull. Singapore 63(1 & 2) 2011
area available for colonisation may also affect epiphyte abundance and community
composition (Catling et al. 1986). Moreover, phorophyte architecture such as canopy
structure, branching pattern, leaf size, etc., may have strong influences on the
germination and establishment of epiphytes (Benzing 1986). More recently, Werneck
& Espirito-Santo (2002) and Garcia-Suarez et al. (2003) studied the structure,
distribution and abundance of vascular epiphytes at different heights and in relation
to the ‘diameter at breast height’ (dbh) of the phorophyte. However, the height of
epiphytic orchids on phorophytes may be an important factor related to pollinator
visitation. Some publications have specifically studied the richness and distribution
of epiphytic species (Madison 1977, Cardelus et al. 2006) but while these studies
include some orchids, to our knowledge, no recent detailed studies of epiphytic orchids
have been reported. Several other researchers have included some orchids in general
studies of epiphytes (Lebrun 1937, Johansson 1974, Catling et al. 1986, Michaloud
& Michaloud-Pelletier 1987, Benzing 1990, Bogh 1992, Zimmerman & Olmsted
1992, Tremblay et al. 2006). Among the most detailed studies of epiphytic orchids
‘are those of Went (1940), Hosokawa (1957) and Johansson (1974) in the forests of
Indonesia, Micronesia and West Africa, respectively. The distribution, colonisation
and association of epiphytic orchids in Belize were studied by Catling et al. (1986),
and orchid-phorophyte relationships in a forest watershed in Puerto Rico were studied
by Migenis & Ackerman (1993).
The aims of the present work were to detail the diversity and distribution of
epiphytic orchids over a wider geographical area in both disturbed and undisturbed
natural forests and to determine their colonisation pattern and the diversity of
phorophytes.
Materials and methods
Study sites
The study area extends over 21,070 sq. km and a general field survey was conducted in
72 different sites of Chittagong, Cox’s Bazar, Rangamati, Bandarban and Khagrachari
districts of south-east Bangladesh. Occurrences on roadside trees were excluded in
this study. Three different forest types were also specially studied. One of the forests
(Sitapahar reserve forest in Rangamati district) was a natural mixed-species forest,
while the others have been replanted after logging. Dariardighi in Cox’s Bazar district
was planted in the 1940s and Perachar in Khagrachari district was planted 1n the 1980s;
both these sites also had some residual mature trees from before the last logging. In
both forests, trees were being logged since the last replanting.
Methods
Epiphytic orchid diversity and frequency of occurrence were generally studied in 72
sites, as well as three different one-hectare quadrats in south-east Bangladesh. Epiphytic
orchid species were recorded from orchid-bearing trees at all sites. The studied sites
were selected on the basis of the number of reserved forests and relatively undisturbed
Colonisation and diversity of epiphytic orchids 343
vegetation units in the respective districts, which resulted in uneven sample sizes.
Both big areas with less studied sites (e.g., Bandarban) and smaller areas with more
studied sites (e.g., Cox’s Bazar) were included.
Access to tree crowns was achieved with a ladder or by climbing the trees
with help of local tree climbers. Inaccessible branches were also examined with the
help of binoculars. Specimens of all orchid species were collected from the study areas
during fieldwork. Collected specimens were identified by consulting the literature
and known specimens available at K, E and ABD. Due to difficulties in determining
sterile specimens, small samples were taken from each and grown in the Orchidarium
of Chittagong University until flowering. Voucher specimens of all species were
deposited in the Herbarium of Chittagong University (HCU). The epiphytic orchid-
bearing phorophytes (supporting trees) were identified in the field and specimens of
unidentified phorophyte trees were collected for later identification.
The epiphytic orchid diversity index for a particular area was calculated by using the
following formula, following the IUCN/SSC Orchid Specialist Group (1996):
Total number of orchid species
Orchid diversity index = = == — x 1000
Total area (sq. km)
Numbers of clumps / individual phorophyte and clump sizes of epiphytic
orchids were recorded in the field. Due to the irregular structure and size of the
different orchids, clumps were also divided into three groups, 1) small clumps of 1—5
pseudobulbs or stem shoots, 2) medium clumps of 6—20 pseudobulbs or stem shoots,
and 3) large clumps of more than 20 pseudobulbs or stem shoots (Huda 2000). Even
very large 1-2 m spreading clumps were considered as a single large clump when all
parts remained connected.
The clump height on the phorophyte tree above the forest floor and diameter at
breast height (dbh) were also recorded for each individual phorophyte. Clump height
of orchids was measured by using a clinometer. The lower portion of the clump was
considered in measuring the height. In hilly areas, metre-graduated bamboo sticks
were used to measure height above the ground. The dbh was measured using diameter
tape.
In the three one-hectare quadrats, all trees >5 m tall were sampled. Within
each quadrat, all trees and orchid species were identified either in the field or in the
laboratory. The number of clumps / individual phorophyte, clump size, height of orchid
clumps above the ground and dbh for trees within the quadrats were also recorded
from all three forest areas.
Analyses
Statistical analyses including the ¥? test, Yates correction, Poisson distribution, t-test,
correlation and ANOVA by Post hoc test (LSD and Tukey test) were performed with
Microsoft Excel and SPSS version 13.
344 Gard. Bull. Singapore 63(1 & 2) 2011
Results
Distribution of epiphytic orchids
A total of 41 different epiphytic orchid species were recorded from the 72 sites. The
scientific name authorities of orchids are given in Fig. 1, those of phorophyte tree
species in Fig. 3. The distribution of the study sites, the total number of different
epiphytic orchids encountered, and the orchid diversity index of each district are
presented in Table 1. In spite of uneven sample sizes, Cox’s Bazar was the richest
area for epiphytic orchids with the highest diversity index and was much more diverse
than the average for the whole region; whereas Khagrachari (a slightly bigger district
with a similar number of sites studied) was the poorest area with the lowest diversity
index. This may give some indication that the richness sampled has not been unduly
influenced by area or number of study sites.
Orchid diversity and colonisation
The 41 orchid species were found on a total of 287 trees in the 72 studied sites. Most
orchid species are infrequent in their occurrence on phorophytes. Aerides odoratum,
Cymbidium aloifolium and Bulbophyllum lilacinum were the most frequently
encountered on 66, 59 and 39 phorophytes, respectively, and 12 orchid species were
found to be present on only one individual phorophyte (Fig. 1). Only 8% of the orchid
species were found on less than 10 phorophyte trees. Among them, Aerides odoratum
and Cymbidium aloifolium, are the commonly occurring species.
The distribution of number of epiphytic orchid clumps per tree among the
phorophyte trees is illustrated in Fig. 2. Most phorophytes bore only a single orchid
Table 1. Total number of epiphytic orchid species found to occur in the 72 studied sites in 5
different districts of south east Bangladesh. The epiphytic orchid diversity index was calculated
according to the IUCN/SSC Orchid Specialist Group (1996).
District name
Parameters Gone ; Total
Chittagong ; Rangamati Bandarban Khagrachari
Bazar
Total no. of om) %6 17 16 07 4]
orchid species
Total area (sq.
Tea 5283 2492 6116 4479 2700 21070
Negus 14 13 Dp I 12 72
studied
Epiphytic
Sita 4.16 10.43 2.78 3.57 2.59 3.42
diversity
index
Colonisation and diversity of epiphytic orchids 345
Aerides odoratum Lour.
Cymbidium aloifoilum (L.) Sw.
Bulbophyllum lilacinum Ridl.
Dendrobium aphyllum (Roxb.) C.E.C Fischer
Acampe praemorsa (Roxb.) Blatt. & McC.
Rhyncostylis retusa (L.) Blume
Papilionanthe teres (Roxb.) Schltr.
Pholidota pallida Lindl.
Acampe papillosa (Lindl.) Lindl.
Micropera pallida (Roxb.) Lindl.
Thrixspermum trichoglottis (Hook. f.) Kuntze
Luisia trichorhiza (Hook. f.) Blume
Aerides multiflorum Roxb.
Smitinandia micrantha (LindI.) Holtum
Acampe ochracea (Lindl.) Hochr.
Oberonia rufilabris Lindl.
Dendrobium palpebrae Lindl.
Dendrobium lindleyi Steud.
Robiquettia succisa (Lindl.) Seidenf. & Garay
Pomatocalpha undulatum (Rchb. f.) J.J.Sm-.
Coelogyne cristata Lindl.
Pelatanthena insectifera (Rchb. f.) Ridl.
Micropera rostrata (Roxb.) Balakr.
Luisia zeylanica Lindl.
Eria tomentosa (Koen.) Hook. f.
Dendrobium parishii Rchb. f.
Dendrobium densiflorum Wall. ex Lindl.
Cleisomeria lanatum Lindl.
Bulbophyllum sessile (Koen.) J.J.Sm.
Cleisostoma appendiculatum Benth. & Hook. f.
Thrixspermum centipeda Lour.
Staurochilus ramosus (Lindl.) Seidenf.
Saccolabiopsis pusilla (Lind].) Seidenf. & Garay
Phalaenopsis comu-cervi (Breda) Par. & Rchb.
Oberonia falconeri Hook. f.
Luisia micrantha Hook. f.
Era pubescens (Hook. f.) Lindl.
Eria bractescens Lindl.
Dendrobium tortile Lindl.
Dendrobium farmeri Paxton
Unknown sp
0 10 20 30 40 50 60 70
Total number of host trees
Fig. 1. Number of phorophyte trees associated with the different epiphytic orchid species.
250
S 20
©
2 150
eed
lo)
g
2 10
j=}
Z
50
0
feet 2S oe 8 mo 0 i a IS i
No. of orchid clumps / tree
Fig. 2. Distribution of number of epiphytic orchid clumps per tree among the phorophyte trees.
346 Gard. Bull. Singapore 63(1 & 2) 2011
clump. Only seven trees had more than seven clumps, and the maximum number of
clumps recorded on a single phorophyte was 16.
Phorophyte diversity
A wide diversity of phorophyte species were colonised by epiphytic orchids. Both
native (51 species) and exotic (6 species) phorophyte trees were recorded and
phorophytes can be either evergreen (35 species) or deciduous (15 species) or semi-
evergreen (5 species) but orchid density on evergreen phorophyte species was greater
than deciduous (2.12, df=16, P=0.008 for quadrats, and 2.01, df/—48, P=0.05 for
72 sites). Orchids were found on 287 phorophytes of 57 different species (Fig. 3). A
total of 30 different phorophyte species were recorded only once. The most frequently
encountered phorophytes were Mangifera indica (49 trees) followed by Artocarpus
heterophyllus, Syzygium grandis, Tectona grandis and Lagerstroemia speciosa.
Mangifera indicaL. EE
Artocarpus heterophyllus Lam.
Syzygium grandis (Wight.) Walp.
Tectona grandis L.
Lagerstroemia speciosa (L.) Pers.
Ficus benghalensis L.
Stereospermum chelonoides (L.) DC
Dipterocarpus turbinatus Gaertn.
Albizia saman (Jack.) F. Muell.
Albizia procera (Roxb.) Benth.
Psidium guajava L. &
inknow n
Artocarpus chaplasha Roxb.
Ziziphus mauritiana Lam
Tamarindus indica L
Toona ciliata M. Roem.
Terminalia bellirica (Gaertn.) Roxb.
Lannea coromandelica (Houtt) Merr &
Bischofia javanica Blume
Syzygium cumini (L.) Skeels. Za
Mangifera sylvatica Roxb.ex Wall Fag
Albizia labbeck (L.) Benth.
Syzygium sp
Hibiscus thhaceus L.
Dillenia scabrella Roxb.. ex Wall. &
Dillenia indica L,
Casuarina equisitifolia Forst. E&
Adina sessilifolia Hook, f. &
Woodfordia fruticosa Salisb.
Vitex glabrata R. Br
Trichilia connaroides (Wight. & Arn.) Bentv.
Terminalia chebula Retz.
Syzygium fruticosum DC.
Syzygium formosum (Wall.) Masam.
Syzygium claviflorum (Roxb.) Wall. &
Swintonia floribunda Gniff
Quercus acuminata Roxb. f
Plumeria rubra L.
Phyllanthus sikimensis Muell. Arg.
Phoenix sylvestris Roxb. §
Oreocnide integrifolia (Gaud.) Mig
Olea dioica Roxb. B
Machilus bambycina King. [
Jatropha curcas L,
Holarrhena pubescens Wall. ex G. Don
Grewia disperma Rottb. —
Garcinia cowa Roxb. ex DC
Ficus religiosa L. &
Cinnamomum dubium Nees. &
Callicarpa arborea Roxb
Bursera serrata Colebr. &
Brownlowia elata Roxb
Bombax ceiba L.
Boemeria glomerulifera Miq.
Bambusa tulda Roxb, @
Aegle mermelos (L.) Correa.
Unknown climber §
0 10 20 30 40 50 60
Number of host trees
Fig. 3. Number of orchid-bearing trees per phorophyte tree species, ranked by abundance.
Colonisation and diversity of epiphytic orchids 347
Orchid colonisation and phorophyte specificity
Number of orchid species and phorophyte tree density were positively correlated
(7=0.902, df=55, P<0.001). A scatter plot of this data with 99% confidence intervals
showed that the diversity of epiphytic orchids on the majority of phorophyte species
was within expectation of the frequency of occurrence of the phorophyte (Fig. 4). Three
native species Syzygium cumini, Stereospermum cheolonoides and Syzygium grandis
had a wider diversity of orchids on them than expected indicating heterogeneity.
No phorophyte species had lower levels of heterogeneity indicating an absence of
phorophyte specificity. There was no indication of a levelling off in the relationship
between phorophyte tree density and orchid diversity, indicating that saturation point
had not been reached.
Colonisation pattern
The total number of orchid clumps recorded on an individual phorophyte species was
correlated with the number of phorophytes of that species (7=0.944, df=55, P<0.001).
Two phorophyte species, A/bizia procera and Albizia saman supported more clumps
than expected (Fig. 5) indicating increased suitability as phorophytes. The greatest
number of clumps was recorded on Mangifera indica.
Number of orchid species
r=0.902 df=55 p<0.001
T T
0 10 20 30 40 50
Number of host trees/ phorophyte species
Fig. 4. The relationship between number of orchid species and phorophyte tree density. The
regression line is fitted with 99% confidence intervals; 39: Stereospermum chelonoides, 42:
Syzygium cumini, 45: Syzygium grandis.
348 Gard. Bull. Singapore 63(1 & 2) 2011
Total number of clumps
r=0.944 df=55 p<0.001
0 10 20 30 40 50
Number of host trees / phorophyte species
Fig. 5. The relationship between total number of epiphytic orchid clumps found and phorophyte
tree density. The regression line is fitted with 99% confidence intervals; 4: Albizia procera, 5:
Albizia saman.
10 @ 72 sites O 3 Quadrats b
4
| rf
0
b a’ b
Average height from ground, m
Small Medium Large
Clump Size
Fig. 6. Relative distribution of small, medium and large orchid clumps at different heights from
the ground (mean+ SE).
Colonisation and diversity of epiphytic orchids 349
™72 sites 03 Quadrats
Average DBH, cm
Small Medium Large
Clump size
Fig. 7. Relative distribution of small, medium and large orchid clumps on phorophyte trees of
different dbh sizes (Mean+SE).
The average height of occurrence of orchid clumps of different sizes is shown
in Fig. 6. Small clumps are found at lower heights than medium or large clumps of
epiphytic orchids (F=5.637, df=314, P=0.004; LSD and Tukey test showed significant
differences between small and medium + large) in the 72 sites, and also at lower
heights in the forest quadrats compared with large clumps (F=8.353, df= 41, P=0.001;
LSD and Tukey test showed significant differences between large and small + medium
clumps).
Clump sizes and average dbh of phorophytes were not significantly different
among the 72 sites (F=2.143, df= 226, P=0.12) but in the forest quadrats the dbh of
trees with large clumps was significantly higher than those for smaller and medium
clumps (F=8.389, df=38, P=0.001; LSD and Tukey test showed difference between
large and small; Fig. 7).
Colonisation of orchids in disturbed and undisturbed forest quadrats
Colonisation of trees by epiphytic orchids, phorophyte characteristics and area orchid
diversity for the three forest quadrats were studied. A test of association between
orchid presence on phorophytes and forest quadrats showed that orchids occur more
frequently on trees in the quadrat sited in the more orchid-rich district than those from
a quadrat in an area of lower orchid diversity index (y° =11.43, df=2, P=0.003; Fig. 8).
Within the forest quadrats, average dbh was associated with orchid occurrence
(3x2 contingency table for large, medium and small dbh of trees with orchid presence,
x =6.04, df=2, P=0.049). There was an association between clump size (large, medium
and small) and forest quadrats (y° =7.41 df=2, P=0.0246) indicating that more large
clumps were found on trees at Sitapahar than expected (Fig. 9).
350 Gard. Bull. Singapore 63(1 & 2) 2011
100 @ Orchid index
90
80 0 % of presence of
70 orchids on host
% of occurrence
Sitapahar Dariardighi Perachara
Name of the sites
Fig. 8. The orchid diversity index and relative commonness of orchids on phorophyte trees,
compared for the three |-ha quadrat study sites; the latter with 95% confidence intervals shown.
80 =
70
60 - O Disturbed (Dariardighi and Perachara)
O) Undisturbed (Sitapahar)
Number of clumps
Ses
S
Large Medium Small
Clump size
Fig. 9. Clump sizes of epiphytic orchids in the two different forest types.
Orchid colonisation in each of the forest quadrats follows the Poisson
distribution (Table 2) indicating that Orchid colonisation is a rare and random event
and one epiphytic orchid species does not attract or repel others on that phorophyte.
nN
—
Colonisation and diversity of epiphytic orchids 3
Table 2. Distribution of orchid clumps on trees above 5 m in height in 3 hectare forest quadrats
and the x” values obtained from a test of goodness of fit to values based on a Poisson distribution.
Places sees van Iclump 2clump 3clump 4clump y? values df P-values
no orchids
Sitapahar 22 10 3 l 0 0.22 2 0.9
Dariardighi ii 10 3 0 l 1.09 3 0.78
Perachara 17 3 0 0 0 0.02 ] 0.89
Total 46 23 6 3 l 0.47 3 0.93
Discussion
Epiphytic orchid diversity and distribution on phorophytes
The total number of epiphytic orchid species recorded during this project accounted
for 39% of the total epiphytic orchid flora of Bangladesh. But considering the size of
the study area (15% of Bangladesh), the epiphytic orchid species diversity of south-
east Bangladesh is therefore reasonably diverse. In India, 14 epiphytic orchid species
have been recorded from a tropical evergreen forest at Varagalaiar, Western Ghats
(Annaselvam & Parthasarathy, 2001) and 26 species from a moist lowland forest of
Eastern Himalaya (Padmawanthe et al. 2004). A total of 101 epiphytic orchid species
were recorded from West Africa rain forest (Johansson 1974), 41 epiphytic orchid
species from South Africa (Harrison 1972), 11 from a forest watershed of Puerto Rico
(Migenis & Ackerman 1993), 232 from West Tropical Africa (Hepper 1968) and 414
from Zaire (Nihoul et al. 1969). The epiphytic orchid species diversity in south-east
Bangladesh therefore appears to be only moderately rich in comparison with these
other forest areas.
Orchids were found infrequently on trees throughout the 72 sites with the
majority of species present in only one clump per phorophyte. A maximum of 16
clumps was found on an old tree of Ficus benghalensis. Thirty-six of the 41 epiphytic
orchid species were found on less than 14 phorophytes. Epiphytic orchid occurrences
on trees in south-east Bangladesh are therefore infrequent or rare and most do not form
tree populations of more than one clump.
Phorophyte diversity
Most studies of epiphytes have concentrated on the epiphytes themselves and not on
the phorophytes that support them. We have documented all the phorophytes for the
epiphytic orchids in this study, together with a record of some of their characteristics.
The 41 species of epiphytic orchids were found on 57 different phorophyte species
of wide taxonomic diversity. The epiphytic orchids were found on both shrubby
and small trees as well as very large trees, bamboos and date palms. The two most
frequently encountered phorophytes were Mangifera indica (Mango) and Artocarpus
Bp Gard. Bull. Singapore 63(1 & 2) 2011
heterophyllus (Jackfruit) and we found 16 and 12 different species of orchids on them,
respectively. These phorophytes are very common, especially near to forest villages,
as tree poachers do not cut these fruit-bearing trees.
Most of the epiphytic orchids were found on evergreen rather than deciduous
tree species. However, further examination of this discrepancy in the forest quadrat
showed no association between leaf persistency and orchid presence on the trees,
indicating that there is no preference for one type of tree. Nevertheless epiphytic
orchid density was greater in evergreen species than in deciduous ones, indicating
that leaf persistence may influence secondary colonisation. This might result from the
continuous presence of moisture throughout the year promoting seedling growth and/
or shade effects, which maintain the growth environment and microclimate. Large
openings in the canopy have been shown to reduce the abundance, distribution and
diversity of shade-requiring epiphytes (Hietz 1999). Conversely, Annaselvam &
Parthasarathy (2001) found that epiphyte density was greater on deciduous species
in India than on evergreens. Architecture may play a critical role in determining
phorophyte suitability (Migenis & Ackerman 1993).
Most individual phorophytes bore only one epiphytic orchid species on
them. This agrees with Annaselvam & Parthasarathy (2001) who showed that most
phorophytes supported only a single epiphyte on them. Benavides et al. (2005) found
one phorophyte carried an average of 2.2 epiphytic individuals and 1.8 epiphytic
species in the rain forests of Colombian Amazonia.
Phorophyte specificity and heterogeneity
No phorophyte specificity was found in the present study. Increase in the number of
phorophytes of a single species was correlated with numbers of native species and
exotic trees in line with expectation. Further, low levels of occurrence of epiphytic
orchid species per native phorophyte supported the lack of specificity. Although
Thrixspermum trichoglottis occurred most frequently on an exotic, Psidium guajava,
this species also occurred on three native species. Orchids also showed no phorophyte
specificity in the Bisley watershed (Migenis & Ackerman 1993). However, epiphytes
often exhibit a certain degree of phorophyte preference, often on trees shared by
co-occurring epiphytes, indicating the suitability of a tree for epiphyte colonisation
(Benzing 1990). Phorophytic preference is also a common phenomenon in Puerto
Rico and elsewhere in the Neotropics, but specificity is not common (Ackerman et
al. 1989, Allen 1959, Zimmerman & Olmsted 1992). Phorophyte specificity has been
shown, or at least surmised, for some Indonesian, Philippine and Puerto Rican orchids
(Went 1940, Sanford 1974, Tremblay et al. 1998) but as in many other tropical forests
(Johansson 1974, Todzia 1986, Zimmerman & Olmsted 1992) there was no evidence
of phorophyte specificity exhibited by orchids in the present study.
There was also no indication of a levelling off of the relationship between
phorophyte frequency and orchid diversity, suggesting that no species saturation had
been reached. The total number of orchid clumps found on most phorophyte species
increased with more records of the host in line with expectation. Two species of
Albizia procera and Albizia saman supported more clumps than expected, indicating
Colonisation and diversity of epiphytic orchids 353
some preference for, or ease of, colonisation. These results are similar to those of
Zimmerman & Olmsted (1992) who showed that commoner phorophytes bore more
epiphytic species.
Colonisation pattern
Our data show that large clumps were found at greater height on the phorophytes,
which might be expected as orchids age and increase in clump size with the growing
tree. However, it might indicate habitat preferences for different epiphytic orchids
within the phorophyte. No similar data directly match our findings due to lack of
previous work on epiphyte clump size. Werneck & Espirito-Santo (2002) reported
a higher abundance of epiphytes at intermediate heights on the phorophytes, but
different epiphytic species showed very contrasting vertical distributions. Benavides
et al. (2005) reported that epiphytic diversity was highest in the branches of crowns
and lowest on the stem bases.
Werneck & Espirito-Santo (2002) found that epiphytic species differed
significantly in their distribution along branch diameters of the phorophyte and our
data showed that larger clumps were found more frequently on trees with larger dbh in
the forest quadrats. Orchids tend to be restricted to larger phorophytes because of their
preference for larger diameter supports (Zimmerman & Olmsted 1992). Prosperi (1998)
also found a highly significant positive correlation between the rate of colonisation
and host diameter above 50 cm, and according to Flores-Palacios & Garcia-Franco
(2006) epiphytic communities are unsaturated, as the number of species increases with
tree size and do not reach a ceiling in Guyana.
Colonisation of orchids in disturbed and undisturbed forests
Our data illustrate that forests in areas of richer epiphytic orchid diversity can have
higher colonisation rates (even though disturbed) than forests in orchid-poor areas,
even when undisturbed. This suggests that epiphytic orchid-rich areas therefore
provide potential recruitment for all types of forests within that zone. However, our
data suggest that throughout all forest types orchid colonisation is a random and rare
event, with no attraction or repulsion shown by the presence of a single clump on a
tree. Larger trees within the forests accumulate more epiphytic orchids and more large
clumps, and undisturbed forests also have larger clumps. Our data suggest that logging
of the most mature trees in a forest selectively modify the epiphytic orchid flora by
combining the decimation of both the largest clumps of orchids, and the trees with the
greatest colonisation, which may have accumulated over many years of growth.
The occurrence of epiphytic orchids tends to be biased towards the larger
phorophytes, a pattern to be expected if large trees are more likely to accumulate
epiphytes, as a result of increased branch area or increased time available for epiphytic
colonisation (Zimmerman & Olmsted 1992). The rate of illegal felling of large trees
has probably already drastically affected the epiphytic orchid flora.
The data presented here show that the diversity of epiphytic orchids is now
based on a fragile phorophyte system because most orchids occur only singly on a
tree. In addition, a large number of epiphytic orchid species have been found only
354 Gard. Bull. Singapore 63(1 & 2) 2011
once (30%). So, careful management is essential to conserve the larger phorophytes
as well as the epiphytic orchids, otherwise one third of all the epiphytic orchids will
experience severe decline.
ACKNOWLEDGEMENTS. The authors would like to express their thanks to the British
Council, The Royal Society, U.K. and the Islamic Development Bank for financial support,
which enabled them to complete this work. The authors also thank the University of Aberdeen
for financial support for two years of fieldwork in south east Bangladesh.
References
Ackerman, J.D. (1986) Mechanism and evolution of food-deceptive pollination
systems in orchids. Lindleyana 1(2): 108-113.
_ Ackerman, J.D., Montalvo, A.M. & Vera, A.M. (1989) Epiphytic host specificity of
Encyclia krugii a Puerto Rican endemic orchid. Lindleyana 4: 74-77.
Allen, P.H. (1959) Orchid hosts in the tropics. Amer. Orchid Soc. Bull. 28: 243-244.
Annaselvam, J. & Parthasarathy, N. (2001) Diversity and distribution of herbaceous
vascular epiphytes in a tropical evergreen forest at Varagalaiar, Western Ghats,
India. Biodivers. Conserv. 10: 317-329.
Barthlott, W., Schmit-Neuerburg, V., Nieder, J. & Engwald, S. (2001) Diversity and
abundance of vascular epiphytes: a comparison of secondary vegetation and
primary montane rainforest in the Venezuelan Andes. P/. Ecol. 152: 145-156.
Benavides A.M.D., Duque, A.J.M., Duivenvoorden, J.F., Vasco, G.A. & Callejas,
R. (2005) A first quantitative census of vascular epiphytes in rain forests of
Colombian Amazonia. Biodivers. Conserv. 14: 739-758.
Benzing, D.H. (1986) The genesis of orchid diversity: emphasis on floral biology leads
to misconceptions. Lindleyvana 1: 73-89.
Benzing, D.H. (1990) Vascular Epiphytes: General Biology and Related Biota, 332 p.
U.S.A., Cambridge: Cambridge University Press.
Bogh, A. (1992) Composition and distribution of the vascular epiphyte flora of an
Ecuadorian Montane rain forest. Se/bvana 13: 25-34.
Cardelus, C.L., Colwell, R.K. & Watkins J.E., Jr. (2006) Vascular epiphyte distribution
patterns: explaining the mid-elevation richness peak. J. Ecol. 94: 144-156.
Catling, P.M., Brownell, V.R. & Lefkovitch, L.P. (1986) Epiphytic orchids in a Belizian
grapefruit orchard: distribution, colonisation, and association. Lindleyana 1(3):
194-202.
Flores-Palacios, A. & Garcio-Franco, J.G. (2006) The relationship between tree size
and epiphytic species richness: testing four different hypotheses. J. Biogeogr. 33:
323-330.
Garcia-Suarez, M.D., Rico-Gray, V. & Serrano, H. (2003) Distribution and abundance
of Tillandsia spp. (Bromeliaceae) in the Zapotitlan Valley, Puebla, Mexico. Pi.
Ecol. 166: 207-215.
Colonisation and diversity of epiphytic orchids 355
Gentry, A.H. & Dodson, C.H. (1987) Diversity and biogeography of Neotropical
vascular epiphytes. Ann. Missouri Bot. Gard. 74: 205-233.
Gravendeel, B., Smithson, A., Slik, F.J.W. & Schuiteman, A. (2004) Epiphytism and
pollinator specialization: drivers for orchid diversity? Philos. Trans., Ser. B. 359:
1523-1535.
Harrison, E.R. (1972) Epiphytic Orchids of South Africa - A Field Guide to the
Indigenous Species, p. 107. The Natural Branch of the Wildlife Protection and
Conservation of South Africa.
Hepper, F.N. (1968) Flora of West Tropical Africa 2nd ed., 3(1): 276. London: Crown
Agents.
Hietz, P. (1999) Diversity and conservation of epiphytes in a changing environment.
Proceedings of the International Conference on Biodiversity and Bio resources:
Conservation and Utilization, pp. 23-27. International Union of Pure and Applied
Chemistry (IUPAC).
Hosokawa, T. (1957) Outline of the mangrove and strand forests of Micronesian
Islands. Mem. Fac. Sci. Kyushu Univ. Ser. E. 2(3): 101-118.
Huda, M.K. (2000) Diversity, Ecology, Reproductive Biology and Conservation
of Orchids of Southeast Bangladesh. Ph.D. thesis. University of Aberdeen,
Department of Plant and Soil Science, U.K.
Huda, M.K. (2007) An up to date enumeration of the family Orchidaceae from
Bangladesh. J. Orchid Soc. India 21(1—2): 35-49
Huda, M.K., Rahman, M.A. & Wilcock, C.C. (1999) A preliminary checklist of orchid
taxa occurring in Bangladesh. Bangladesh J. Plant Taxon. 6: 69-85.
IUCN/SSC Orchid Specialist Group (1996) Orchids- Status Survey and Conservation,
p. 153. Action Plan. Gland Switzerland and Cambridge, UK: IUCN.
Johansson, D. (1974) Ecology of Vascular Epiphytes in West African Rain Forest.
Ph.D. thesis. University of Upsala, Sweden.
Kress, W.J. (1986) The systematic distribution of vascular epiphytes: an update.
Selbyana 9: 2-22.
Lebrun, J. (1937) Observations sur les epiphytes de la forest equatoriate congolaise.
Soc. Scientif. Bruxelles 57: 31-38.
Madison, M. (1977) Vascular epiphytes: their systematic occurrence and salient
features. Se/byana 2: 1-13.
Michaloud, G. & Michaloud-Pelletier, S. (1987) Ficus hemi-epiphytes (Moraceae) et
arbres supports. Biotropica 19(2): 125—136.
Migenis, L.E. & Ackerman, J.D. (1993) Orchid phorophyte relationship in a forest
watershed in Puerto- Rico. J. Trop. Ecol. 9: 231-240.
Nihoul, E., Schelpe, E.A. & Hunt, P.F. (1969) A provisional checklist of the orchids in
the Congo-Kinshasa. Am. Orchid Soc. Bull. 38: 578-584.
Padmawanthe, R., Qureshi, Q. & Rawat, G.S. (2004) Effects of selective logging on
vascular epiphyte diversity in a moist lowland forest of Eastern Himalaya, India.
Biol. Conserv. 119: 81-92.
Prosperi, J. (1998) Biologie du Development des Hemi-epiphytes Ligneux. Unpublished
Ph.D. thesis, University of Montpellier II, Montpellier.
356 Gard. Bull. Singapore 63(1 & 2) 2011
Sanford, W.W. (1974) The ecology of orchids. In: Withner, C.L. (ed) The Orchids:
Scientific Studies, pp. 1-100. New York: John Wiley & Sons.
Todzia, C. (1986) Growth habits, host tree species and density of hemi epiphytes on
Barro Colorado Island, Panama. Biotropica 18: 22—27.
Tremblay, R.L., Melendez-Ackerman, E. & Kapan, D. (2006) Do epiphytic orchids
behave as metapopulations? Evidence from colonisation, extinction rates and
asynchronous population dynamics. Biol. Conserv. 129: 70-81.
Tremblay R.L., Zimmerman, J.K., Lebron, L., Bayman, P., Sastre, I., Axelrod, F. &
Alers-Garcia, J. (1998) Host specificity and low reproductive success in the rare
endemic Puerto Rican orchid Lepanthes caritensis. Biol. Conserv. 85(3): 297—
304.
Went, F.W. (1940) Coping with the epiphytic existence: pollination strategies. Se/byana
9: 52-60.
Werneck, M.D.S. & Espirito-Santo, M.M.D. (2002) Species diversity and abundance
of vascular epiphytes on Vellozia piresiana in Brazil. Biotropica 34(1): 51—S7.
Withner, C.L. (1974) The Orchids; Scientific Studies. London: John Wiley & Sons.
Zimmerman, J.K. & Olmsted, I.C. (1992) Host tree utilization by vascular epiphytes
in a seasonally inundated forest (Tintal) in Mexico. Biotropica 24(3): 402-407.
Gardens’ Bulletin Singapore 63(1 & 2): 357-374. 2011 357
Habitat and ecological preferences of
Aydriastele costata (Palmae)
in Waigeo Island, West Papua
Didik Widyatmoko
Cibodas Botanic Gardens, Indonesian Institute of Sciences (LIPI),
Jl. Kebun Raya Cibodas, Sindanglaya, Cianjur 43253, Indonesia
didik_widyatmoko@yahoo.com
ABSTRACT. The research was conducted to test hypotheses about the significance and influence
of edaphic parameters and association patterns in governing the occurrence and abundance of
the karst palm Hydriastele costata and its co-occurrence with other plant species in Waigeo
Island, West Papua. The results indicate that a number of interrelating edaphic factors influence
the palm’s occurrence and abundance. The palm showed a preference for dry, well-drained soil,
with high magnesium (Mg” ) content. Most colonies occurred in localities where Mg” content
was very high. High alkaline concentrations also strongly corresponded to the presence of the
emergent palm. Six of 15 tropical plant species were positively associated while the rest were
negatively associated with H. costata. Four species (Casuarina rumphiana, Decaspermum
bracteatum, Baeckea frutescens, and Pinanga rumphiana were strongly associated with H.
costata, as indicated by their high association degrees using the Ochiai indices. The palm P.
rumphiana appears to have similar habitat requirements as H. costata. The xerophytic palm H.
costata tended to occupy sites with medium carbon/nitrogen (C/N) ratios where all sampled
populations occurred in habitats with average C/N values more than 10. Based on the 7-squared
values, exchangeable Mg” and calcium (Ca*) appeared to have more influence on plant density
and frequency than on crown and basal areas. The exchangeable Ca* contents showed a similar
pattern to Mg* concentrations. Curiously, potassium (K~), sodium (Na), aluminium (Al**)
and hydrogen (H’) contents did not show significant relationships with the palm abundance
parameters.
Keywords. Association, co-occurrence, habitat preferences, Hydriastele costata, palms,
Waigeo, West Papua
Introduction
Understanding the mechanisms for species co-occurrence and habitat preference
(specialisation) is crucial for habitat management (Begon et al. 1996, Ludwig &
Reynolds 1988, Mohler 1990, Nakashizuka 2001, Christie & Armesto 2003, van der
Heijden et al. 2003, Hall et al. 2004). Although the detection of co-occurrence or
association among or between species and environmental variables does not provide
a causal understanding (Morisita 1959, Schluter 1984, Silvertown et al. 1992, Real
& Vargas 1996), it can be used to generate hypotheses of possible underlying causal
factors.
358 Gard. Bull. Singapore 63(1 & 2) 2011
Palms often show local or regional patterns of co-occurrence and ecological
preferences (Tomlinson 1979, House 1984, Kahn & Mejia 1990, Moraes 1996, Svenning
1999). Some palms appear to be adapted to specific edaphic conditions, such as soil
quality, drainage and type (House 1984, Tomlinson 1990, Moraes 1996, Widyatmoko
& Burgman 2006). Most tropical rain forest tree species have strongly aggregated
spatial distribution patterns due to a high degree of habitat specialisation (Ashton
1998, Phillips 1998, Condit et al. 2000, Hubbell 2001). However, Duivenvoorden
(1995, 1996) argued that most trees of the well-drained upland habitat in Colombian
Amazonia are likely to be soil generalists rather than specialists, implying limited
importance of microhabitat specialisation for maintaining tree species richness.
There has been a lack of consensus about the importance of the correlations
between plant abundance and edaphic conditions at local and intermediate spatial
scales, e.g., at 1-100 km? (Gartlan et al. 1986; Swaine 1996; Clark et al. 1998, 1999;
Hall et al. 2004). Tropical soils are not homogeneous at regional, intermediate or even
local scales (Richter & Babbar 1991, Hall et al. 2004) and abrupt discontinuities in
edaphic conditions are common features (Clark et al. 1998). Regional or intermediate
spatial scales refer to strong environmental discontinuities (habitat types) while local
spatial scales refer to environmental conditions that vary at scales less than 10° m, such
as treefall gaps and local topographic variation (Svenning 1999).
Plant co-occurrence and abundance may be determined largely by nutrient
availability, heterogeneity of the biotic and abiotic environment, and microhabitat
specialisation (Silvertown & Law 1987, Ludwig & Reynolds 1988, Kahn & Mejia
1990, Hatfield et al. 1996, Clark et al. 1998, Svenning 1999, Webb & Peart 2000, van
der Heijden et al. 2003, Palmiotto et al. 2004). Some other studies have shown that
tropical plant species distributions and community composition are correlated with
soil nutrient status (Tucker 1992, Poulsen 1996, Clark et al. 1998, Svenning 2001,
Widyatmoko 2001, Widyatmoko & Burgman 2006, Widyatmoko et al. 2007) such as
magnesium and phosphor (Olsen & Sommers 1982, Vitousek & Sanford 1986, Baillie
et al. 1987, Suarez 1996, Sollins 1998, Tiessen 1998, Potts et al. 2002, Hall et al. 2004,
Palmiotto et al. 2004) as well as calcium, potassium, and sodium contents (Suarez
1996, Widyatmoko & Burgman 2006).
Spatial distribution patterns may also be determined by complex relationships
within and between species, including seed dispersal (Bell 2000), competition for
pollinators (Armbruster 1995, Svenning 1999), recruitment and regeneration (Harms
et al. 2000, Christie & Armesto 2003, Widyatmoko et al. 2005), density dependence
(Webb & Peart 2000), intermediate disturbance (Molino & Sabatier 2001) or variation
in topography and soil water (Campbell 1985, Swaine 1996, Davie & Sumardja 1997,
Clark et al. 1998, Svenning 2001). Very little information is available about the roles
and influences of soil conditions and biotic associations on plant abundance and co-
occurrence (Higgs & Usher 1980, House 1984, Gentry 1988, Duivenvoorden 1995).
Hypotheses regarding species co-occurrence invoke equilibrium and non-
equilibrium explanations (Svenning 1999, Nakashizuka 2001, Groeneveld et al.
2002, Edmunds et al. 2003). Equilibrium hypotheses assume that species co-occur
by occupying different niches (niche partitioning), while non-equilibrium hypotheses
Hydriastele ecology in Waigeo, West Papua 359
emphasise local fluctuations, disturbance and chance events that do not determine
species composition, although they may result in expectations for relative species
abundances (Hubbell 2001, Chisholm & Burgman 2004). Both equilibrium and non-
equilibrium processes seem likely to contribute to the composition of most plant
communities (Nakashizuka 2001).
The interspecific association test is a simple species-based approach for
preliminarily defining community types that can be recognised by a small assemblage
of common species. If sets of species are found to co-occur, and the occurrence of these
sets can be related to habitat factors, such information will provide more compelling
evidence for niche processes structuring the community than does a single species
approach.
The objective of this research was to (1) test hypotheses about the significance
and influence of edaphic parameters in governing the occurrence and abundance of the
important karst palm Hydriastele costata F.M.Bailey in a tropical lowland rain forest
of Waigeo Island, West Papua, and (2) assess the tendency of the palm to co-occur with
other plant species. We addressed three questions to answer. First, do local edaphic
conditions in different habitat types affect the occurrence and abundance of H. costata?
Second, does the palm species associate with other plant species in the Waigeo forest?
Third, if so, how strong is the association? Such information is required to support the
reserve management system, particularly through long-term monitoring of significant
populations of the species occurring on different habitat types. Long-term monitoring
programs will provide foundations for developing management prescriptions and
conservation priorities for the valuable species and its habitat.
Materials and methods
Study species
Hydriastele costata (Arecaceae) is a solitary, erect, straight, unarmed, tall (up to 20-30
m), pleonanthic, monoecious palm species (Uhl & Dransfield 1987, Widyatmoko et al.
2007). The palm species occurs in New Guinea, Bismarck Archipelago and northern
Australia, particularly in the humid lowland and hill tropical rain forests at altitudes
from 0 up to 500 m above sea level, occupying mainly the forest upper canopy,
preferring coastal calcareous soils. H. costata is not only of interest from an ecological
point of view, but also in terms of economic significance since it provides a number
of benefits. The palm is used as an ornamental plant, the stem is used for constructing
local houses (floors and laths), while the hard outer part of the stem is locally used for
making spears.
Study sites
The study focused on the Waifoi Forest and Kamtabae River located within the East
Waigeo Nature Reserve, the Raja Ampat Islands, West Papua (Fig. 1), at altitudes
ranging from 0 to 500 m above sea level, where H. costata mostly occurs. The camp
established at Kamtabae River (130'43°38.2”E 0'5’53.3”S) was used as the reference
360 Gard. Bull. Singapore 63(1 & 2) 2011
Bougainville Strait
Dampler Strait
: [Air ter
Bes
sg 1073-1074]
Fig. 1. The study area where coordinates and elevation of all sites studied (tree symbols) were
recorded; Waifoi (Wayfo1) village is at the center of the circle. Inset shows locality of Waifo1
on Waigeo.
point to explore the surrounding forests. Different directions comprising all four
aspects were covered in order to comprehensively cover the study area. Lowland and
hill forests of mixed-age indigenous vegetation, with slopes ranging from 30 to 70%,
dominated the inland nature reserve topography.
East Waigeo Nature Reserve was established in 1996 based on the decree of the
Indonesian Minister of Forestry No. 251/Kpts-II/1996 covering a total area of 119,000
hectares, located between 130°39°49”E and 130'55’54”E and between 0'02’27”S and
0.08°51”S. It has the ‘Af’ climate type, experiencing eight consecutive wet months.
All months have an average temperature above 18°C (ranging from 23°C to 32°C).
Waigeo only has small seasonal temperature variations of less than 3°C (the Koppen’s
System, Tarbuck & Lutgens 2004); with an average humidity of 85% during June
2007.
Hydriastele ecology in Waigeo, West Papua 361
Waigeo Island is one of the four major islands of the Raja Ampat Archipelago.
The waters and environment around Waigeo Island have been known as the most
biodiverse marine area in the world, especially in terms of coral reefs and fish species
(Webb 2005, Pemerintah Kabupaten Raja Ampat & Conservation International
Indonesia 2006). However, despite it being a biologically very rich area, little is
known about the Islands’ plant diversity and terrestrial resources (Badan Perencanaan
Pembangunan Nasional - Bappenas 2003, Webb 2005). Detailed surveys focusing on
the plant diversity will provide important baseline data for managing and conserving
the Islands’ biodiversity sustainably (Webb 2005).
Geologically, Waigeo Island is interesting, in having extensive karst
ecosystems, alluvium substrates, acid volcanic and ultrabasic rocks, with some
relatively high mountains (Jepson & Whittaker 2002, Webb 2005, Pemerintah
Kabupaten Raja Ampat & Conservation International Indonesia 2006). The flora must
be diverse according to the substrate and biogeographic reasons, as well as habitat
characteristics which range from submontane forests to sago swamps and mangroves
(via forests on karst and acid volcanics). Hill forests on volcanic substrates and karst
formations extensively occur on this island. The island ultrabasic scrub 1s also unique
and widely known for its endemic species (Webb 2005). Each island of the Raja Ampat
has its own characteristics, especially in terms of vegetation composition and habitat
types. Waigeo Island is botanically very important and valuable, despite its relatively
small size compared to the main island of Papua (Johns 1995, Johns 1997, CI 1999).
Selection of habitat types
In order to study H. costata habitat preferences, eight habitat types were selected:
coastal line (seashore), coastal hill slope, coastal hill top, inland river bank, inland hill
slope, inland hill top, disturbed forest (most native species present), and converted
forest (most native species removed). The characteristics of each habitat type (including
slope, soil formation, elevation soil pH, average humidity, and average temperature)
were described and recorded.
Vegetation structure and composition
A series of 24 belt transects (of 100 m x 10 m each) was established at the eight
habitat types selected (i.e., three transects on each habitat), stretching from the camp
at the Kamtabae River (13043°38.2”E 05753.3°S) to behind the Waifoi Village
(130°42°46.7°E 0°6°5.9"S). Locations of each habitat type and belt transect were
recorded using a Garmin Global Positioning System MAP 175. The major axes of all
transects were orientated north-south derived from a selected compass bearing (Krebs
1989, Cropper 1993). All stemmed individuals of H. costata within each transect were
counted. Damaged or dead individuals were not included. Land slopes were measured
using a clinometer (SUUNTO Optical Reading Clinometer PM-5), while soil pH and
humidity were measured using a soil tester Demetra patent no. 193478 Electrode
Measuring System. Soil profiles were sampled by using a soil sampler the Belgium
auger | m. Soil analyses were conducted at the Soil Research Center, Bogor. The level
362 Gard. Bull. Singapore 63(1 & 2) 2011
of forest (habitat) disturbance was determined on the basis of the proportion of the
remaining native species.
Interspecific association (co-occurrence)
Association patterns among co-occurring species were tested using the chi-square test
statistic by constructing the hypothesis that two species are not associated at some
predetermined probability level. Fifteen plant species were tested for association from
67 observed. The strength of each association was tested using the Ochiai Index (OI)
as recommended by Ludwig & Reynolds (1988):
a
Oa een
Va+bva+b
Where:
a = the number of plots where both species (H. costata and the paired species)
occur;
b = the number of plots where H. costata occurs, but not the paired species; and
c = the number of plots where the paired species occurs, but not H. costata.
Test of association. The palm was absent from one of the eight observed habitat types.
The site was in the Waifoi village and was regarded as non-natural forest area (1.e., the
coexisting plants have been planted with Theobroma cacao). Measures of interspecific
association were based on the presence and absence of species within quadrats
developed. A total of 144 quadrats of 5 m x 5 m each were sampled from the observed
sites within the reserve with different vegetation types and associations. Quadrats were
arranged systematically in an alternating pattern within the belt transects (of 100 m x
10 m each) in order to cover uniformly both sides of the axes (Mueller-Dombois &
Ellenberg 1974, Cox 1974, Sokal & Rohlf 1981, Ludwig & Reynolds 1988). The data
were then summarised in the form of a 2 x 2 contingency table.
The null hypothesis (H,) constructed was that the distribution of H. costata
is independent of the other species. To test the null hypothesis of independence, the
chi-square test statistic (x,) was used (Ludwig & Reynolds 1988). The significance
of the chi-square test statistic is determined by comparison with the chi-square
distribution (%,) for | df at a= 0.05. If x, > %, the null hypothesis is rejected. Rejecting
the null hypothesis indicates an association between H. costata and the paired species,
implying that the two species co-occur at a frequency greater than expected by random
association. Positive or negative associations were determined by comparing the value
of observed occurrences (O,,,) to that of expected occurrences (E_, ). If observed is
greater than expected, there is a positive association (the pair of species occurred
together more often than expected if independent).
Measure of the strength of association. The Ochiai Index (a measure of association)
was used to quantify the strength of association between the species tested (1.e., H.
costata and the paired species), as the association test can only determine whether the
Hydriastele ecology in Waigeo, West Papua 363
species tested are associated or not associated, but not the degree of the association.
The index was recommended by Janson & Vegelius (1981) and Hubalek (1982) as it
proved less biased. The value of the Ochiai Index is equal to 0 at no association and 1
at complete or maximum association.
Results
Habitat preferences
H. costata seemed to prefer specific habitat types. Highest densities occurred on
hill slopes and tops near the coastal area (Table 1, Fig. 2). In contrast, the palm was
suppressed at the shore and even absent from converted forest where most native
species have been removed and replaced by Theobroma cacao (Fig. 2). Although the
palm tolerated minor forest disturbance, the populations were generally low in this
type of habitat, indicating a tolerance of sub-optimal conditions. The highest density
was found on ultrabasic soil, on steep slopes and dry-open canopy gaps, where there
were 72.4 adult individuals ha‘! (Table 1).
Species co-occurrence
Sixty seven possible co-occurring species were analysed, of which 15 species were
tested for association with H. costata. Six of these 15 tested species were positively
associated while the rest were negatively associated (Table 2). For the six species
(Casuarina rumphiana Mig., Decaspermum bracteatum (Roxb.) A.J.Scott, Baeckea
frutescens L., Pinanga rumphiana (Mart.) J.Dransf. & Govaerts, Exocarpos latifolius
R.Br., and Myrsine rawacensis A.DC.), the association with H. costata was strong,
Abundance (Individuals ha ~')
>
oO
Coastal line Hill slope, Hilltop, River bank, Hill slope, Hill top, Disturbed Converted
coastal coastal inland inland inland forest forest
Habitat Type
Fig. 2. Population densities of Hydriastele costata at different habitat types within the East
Waigeo Nature Reserve, the Raja Ampat Islands, West Papua.
364 Gard. Bull. Singapore 63(1 & 2) 2011
Table 1. The abundance of H. costata at various habitat types within the East Waigeo Nature
Reserve, the Raja Ampat Islands, West Papua. Mean abundance + S.D. (95% Confidence
Interval).
Habitat Abund- Habitat characteristics Altitude SoilpH Av. Hum- Av.
type ance (m asl) idity (%) Temp
(ha'') (°C)
Coastal 6.2+1.6 0-30% slope, tidal 0-10 6.2 82+8.17 30.9
line/shore influences, mud
formation
Hill slope, 72.4+5.4 30-80% slope, dry, 20-100 7.2 69: 7.02, 31-9
coastal open (wide canopy
gaps), karst, ultrabasic
soil
_ Hill top, 48.3+5.3 30-60% slope, very 100-150 7.2 63+6.24 31.7
coastal dry, wide canopy gaps,
karst, ultrabasic soil
River bank, 17.5+3.9 0-30% slope, lowland, 30-40 6.0-6.9 89+8.29 28.5
inland alluvium, frequent
floods
Hill slope, 8.6+2.9 30-70% slope, hill 40-120 6.6-7.0 84+6.72 28.3
inland forest, alluvial deposit,
volcanic soils
Hill top, 3.4+1.1 30-60% slope, hill 120-170 6.1-6.2 82+9.02 27.6
inland forest, alluvial deposit
Disturbed DD 3= (0).3) 10-40%, most native 60-100 6468 81+648 29.1
forest species remained,
Lansium domesticum
planted
Converted 0 10-40%, most native 20-60 6.1-6.3 74+6.01 29.6
forest species removed, cacao
planted
indicated by their >0.5 indices. Although in some sites Livistona brevifolia Dowe &
Mogea and Licuala graminifolia Heatubun & Barfod were found together with H.
costata, their co-occurrence was not consistent. Surprisingly, the apparently closely
associated species Styphelia abnormis (Sond.) F.Muell. and Wendlandia buddlejacea
F.Muell. were negatively associated with H. costata. Unlike Orania regalis Blume ex
Zipp., which is a shade-tolerance species, H. costata prefers and occupies karst and
open coastal areas. As a consequence, their association degree was very low (Table 2).
A number of interrelating edaphic factors appeared to explain the occurrence
Hydriastele ecology in Waigeo, West Papua
365
Table 2. Results of the association tests using the chi-square test statistic (yt) between H. costata
and the fifteen co-occurring species. Values of the Ochiai Index are 0 at “no association” and |
at “complete (maximum) association”.
Paired species Result of Types of Strength of association
chi-square test Association (Ochiai Index)
Casuarina rumphiana Associated Positive 0.69
Decaspermum bracteatum Associated Positive 0.67
Baeckea frutescens Associated Positive 0.62
Pinanga rumphiana Associated Positive 0.61
Exocarpos latifolius Associated Positive 0.55
Myrsine rawacensis Associated Positive 0.52
Styphelia abnormis Associated Negative 0.43
Wendlandia buddlejacea Associated Negative 0.42
Ploiarium sessile Associated Negative 0.31
Livistona brevifolia Associated Negative 0.24
Decaspermum fruticosum Associated Negative 0.19
Licuala graminifolia Associated Negative 0.16
Pometia pinnata Associated Negative 0.15
Sommieria leucophylla Associated Negative 0.11
Orania regalis Associated Negative 0.07
2.00 -
ua ®
o
= 1.50 1
=
= |
=
w 1.00 |
® °
°
c
S y = 1.51x - 0.52
2 0.50 5 :
= © R* = 0.66; p < 0.0001
2
oo
0.00 - =
0.0 0.5 1.0 15 2.0
Mg?* Content (cmol(+) kg”)
Fig. 3. Relationship between Mg” content and density of Hydriastele costata within the East
Waigeo Nature Reserve, Waigeo. Mg® content values are Lo
Log,, individuals ha-1.
g and abundance values are
366 Gard. Bull. Singapore 63(1 & 2) 2011
2.00 -
i)
= 1.50 +
=
ao]
=
w 1.00 -
8
&
3 050 - y = 1.02x + 0.57
2 a" R? = 0.65; p < 0.0001
<
0.00 ai} T T 1
0.00 0.50 1.00 1.50
Ca”* Content (cmol(+) kg”)
Fig. 4. Relationship between Ca** content and density of H. costata within the East Waigeo
Nature Reserve, Waigeo. Ca content values are Log,, and abundance values are Log,,
individuals ha".
Table 3. Results of the soil analyses conducted at six different habitat types within the East
Waigeo Nature Reserve, Waigeo, West Papua. Hill slope, coastal: Horizon Ao (0-13 cm), Al
(13-48.5 cm), A2 (48.5—78.5 cm), AB (78.5—148 cm). Hill top, coastal: Horizon Ao (0-7 cm),
Al (7-18 cm), A2 (18-32.5 cm), AC (32.5—86.5 cm). Hill slope, inland: Horizon Ao (0-20
cm), Al (20-40 cm), A2 (40-60 cm), AB (60-80 cm). Hill top, inland: Horizon Ao (0-15 cm),
Al (15-35 cm), A2 (35-75 cm). River bank, inland: Horizon Ao (0-16 cm), Al (16-40 cm).
Coastal line: Horizon Ao (0-10 cm).
Para Hill slope, coastal Hill top, coastal Hill slope, inland Hilltop, inland =River — Coastal
meter bank, line
inland
Ao Al A2 AB Ao Al A2 AC AO Al A2 AB Ao Al A2 Ao Al Ao
pH 66 68 69 7.1 66 6:8 7:0) 7 “GON 6:1) (6:6) 26:9 Gi) Gl 62 GA mors 7.2
C/N 8 8 10 8 7 9 9 11 8 9 8 8 12 9 8 8 8 12
GaZE) 10) 8:08 16) 7s) 16 17 13 LO) WAS WS Pus 0:8 U7 20'S W072 o es: 33
AB+ 0:0 0:0 O!0 0:0) O10) (010) 0:0) 20:0" 0:0) D0 OO 00 0:0 00 SOOO Oss 0.0
H+ OO (OL 00 “OL ONT 0!) FOO OO) Os Onl Olle O00 Ol Ose Osteen Onl ane an 0.1
Hydriastele ecology in Waigeo, West Papua 367
and abundance of H. costata. This palm showed a preference for ultrabasic soils, steep
slopes, and dry-open canopy gaps, with high magnesium (Mg?’) and calcium (Ca**)
contents. The largest population occurred on hill slopes where the highest magnesium
content was recorded (Table 3). There was a strongly positive correlation between the
abundance of H. costata and the soil mineral Mg” content (Fig. 3). The three largest
populations (coastal hill slopes, coastal hill tops, and inland river bank) occurred
in sites where Mg’ contents were high (Table 3). According to the Soil Research
Center (1983), a Mg** content >8.0 cmol(+)/kg was categorised as “very high” (Table
4). To some extent, Ca** contents also influenced the occurrence of H. costata, 1.e.,
higher concentrations of Ca** corresponded with higher densities of the palm (Fig.
4), although the trend was not as clear as that for Mg**. Curiously, K*, Na*, Al* and
H* concentrations did not show significant relationships with H. costata abundance
parameters. The palm tended to occur in sites with lower C/N ratios (higher N
contents). All observed populations occurred in habitats with average C/N values <10.
Table 4. Classification and criteria for soil chemical properties as defined by the Soil Research
Center (1983).
Soil Properties “gerler ; iia aD Very Een
G (%) : <I 00 | 1.00-2.00 | 2.01-3.00 3.01-5.00 >5.00
N (%) <0.10 0.10-0.20 0.21-0.50 0.51-0.75 >0.75
C/N <5 5—10 11-15 16-25 >25
P,O, HCI (mg/100g) <10 10-20 21-40 41-60 >60
P.O. Bray | (ppm) <10 10-15 16—25 26-35 >35
P,O, Olsen (ppm) <10 10-25 26-45 45-60 >60
KO HCI 25% <10 10-20 21-40 41-60 >60
(mg/100g)
Cation Exchange <5 S16 17-24 25-40 >40
Capacity (cmol(+)/kg)
K* (cmol(+)/kg) <0.1 0.1—0:2 0.3—-0.5 0.6—-1.0 >1.0
Na* (cmol(+)/kg) <0.1 0.1—0.3 0.4-0.7 0.8-1.0 >1.0
Mg” (cmol(+)/kg) <0.4 0.4—1.0 1.1—2 2.1-8.0 >8.0
Ca** (cmol(+)/kg) <2 2-5 6-10 11-20 >20
Alkali Saturation (%) <20 20-35 36-50 51-70 >70
Alumin. Saturation (%) <10 10-20 21-30 31-60 >70
pH H,O <A 9-9) 5:0 — 6.5 6.6 — 7.5 7.6 — 8.5 >8.5
Acid Slightly Neutral Slightly Alkaline
Acid Alkaline
368 Gard. Bull. Singapore 63(1 & 2) 2011
The largest colony at the coastal hill slopes had an average C/N value of 8.5, followed
by the coastal hill tops colony with an average value of 9.0 (Table 3).
Based on the r-squared values, exchangeable Mg** appeared to have more
influence on plant density and frequency than on basal area and crown area (Table 5).
The exchangeable Ca’* concentrations showed a very similar pattern to Mg”* contents,
while C/N values seemed to have a negative correlation with frequency and density
(1.e., higher values of C/N correlated with lower plant densities). Soil pH appeared
to have more influence on plant density than on plant frequency. On the other hand,
K*, Na‘, Al’* and H* contents did not show significant relationships with the palm
abundance parameters, as indicated by their low correlation coefficients (Table 5).
Discussion
The positive association of H. costata with high contents of Mg** and Ca’* is similar
to that of the Papuasian palm Orania regalis (Widyatmoko 2009), the Malayan rain
forest bertam palm Eugeissona triste Griff. (Fong 1977) and the Amazonian palms
Phytelephas macrocarpa Ruiz & Pay. and Astrocaryum murumuru Wallace var.
murumuru (Vormisto 2002) which prefer higher soil mineral contents. On the other
hand, the association pattern of H. costata is different from that of the lipstick palm
Cyrtostachys renda Blume (Widyatmoko & Burgman 2006) and the bayas palms
Oncosperma horridum Scheff. and O. tigillarium (Jack) Ridl. (House 1984) which
prefer low levels of Ca**, Mg”* and K*. Widyatmoko & Burgman (2006) showed that
C. renda preferred sandy, well-drained soils with low mineral contents, while House
(1984) found that O. horridum and O. tigillarum did not avoid flooded areas and
poorly drained clay substrates.
Table 5. Values of correlation coefficient (7-squared) between edaphic parameters and
abundance of H. costata within the East Waigeo Nature Reserve, Waigeo. Notes: (+) indicates
a positive correlation; (-) indicates a negative correlation; * p < 0.0001; sample size (n) = 14.
Basal Area
Eeoehe Frequency Dees Canopy Gs
Parameters ( Individuals ha!) (m* ha’!) (m? ha!)
pH (-) 0.52 : (4) 0.61" (-) 0.31 ; (-)0.30 : ¥
C/N (-) 0.60* (-) 0.62* (-) 0.27 (-) 0.31
Exch. Ca** (+) 0.62* (+) 0.65* (+) 0.41 (+) 0.41
Exch. Mg?’ (+) 0.63* (+) 0.66* (+) 0.54 (+) 0.47
Exch. K’ G),0:32 (+) 0.38 G)0!33 (+) 0.29
Na’ (+) 0.41 (GE O).35) (+) 0.39 (+) 0.36
As 0.00 0.00 0.00 0.00
H’ (-) 0.34 (-) 0.39 (-) 0.28 (-) 0.24
Hydriastele ecology in Waigeo, West Papua 369
Unlike Licuala graminifolia which is a relatively shorter-lived opportunistic
species that rapidly colonises canopy gaps, H. costata is a slower-growing, long-lived
species constituting an emergent canopy layer. Unlike H. costata, L. graminifolia
occupies lower subcanopies, thus having different levels of sunlight exposure. Licuala
graminifolia is more widely distributed throughout Papua. To some extent, H. costata,
Pinanga rumphiana and Casuarina rumphiana may fill equivalent ecological roles and
share membership of the same ecological guild. Hydriastele costata and P. rumphiana
seem to share similar population establishment strategies, and both species naturally
regenerate from seeds but not from suckers and both species produce relatively
abundant seeds.
The abundance of H. costata seemed to increase with the cation exchange
capacity. Soil cation exchange potential is linked with soil drainage capacity and well-
drained soils contain high sand fractions (White 1997). The mean density of H. costata
on hill slopes adjacent to coastal area was 72.4 individuals ha', while on hill slopes
far away from coastal area (inland) it was only 8.6 individual ha'. The absence of H.
costata from converted forest is an indication that this species is intolerant of habitat
disturbance, in which growth is prevented. In addition to the apparent preference for
dry, well-drained soils, H. costata appeared to be more common in sites with higher
electrical conductivity and higher concentrations of major nutrients, especially Mg**
and Ca™*. Surprisingly, K* and Na’ contents did not correlate significantly with the
palm density and frequency. This may be due to the very low contents of these minerals
at various sites studied.
Hydriastele costata often forms a prominent component of the coastal Waigeo
vegetation. However, a high level of disturbance, such as forest clearance behind the
Waifoi forest, has caused some colonies to decline. In heavily shaded inland sites
of the reserve, the palm very scarcely occurs with only very few individuals found.
The palm is not a true gap exploiter and appears to be unable to take advantage of
unstable canopy conditions (i.e., slightly disturbed habitats) and to become established
in ecologically limited spaces.
Slope angle and vegetative cover affect moisture effectiveness by governing
the ratio of surface run-off to infiltration. As drainage deteriorates, the oxidised soil
profile of well-drained sites is transformed into the mottled and gleyed profile of a
wet soil. The influence of slope on soil texture and water holding capacity partly
determines the levels of available mineral nutrients, and thus the establishment and
spatial distribution of vegetation. Soils on slopes tend to be coarser and better drained
than those on flat ground where run-off creates accumulations of small soil particles
(House 1984, White 1997, Hall 2004).
It seems that generative propagation through seed germination is most
important for colony maintenance, while seed dispersal must be important for the
establishment of new colonies far removed from reproductive adults through water
transport (hydrochory). Seeds were sometimes seen to germinate in canopy gaps.
Curiously, seedlings were often absent beneath the crowns of mature individuals. As
light exposure is important for flowering and successful fruit set, and because the
crowns of this palm occupy mainly the upper canopy, it is not surprising that fertile
370 Gard. Bull. Singapore 63(1 & 2) 2011
adult plants were relatively abundant. No effective dispersers of H. costata seeds were
encountered during this study. Due to small seed size, long-distance travellers such as
frugivores and granivores (pigeons) are likely to be potential dispersal agents.
Conclusion
Relationship between Mg** and Ca** content and the occurrence and abundance of H.
costata was detected. Four species (Casuarina rumphiana, Decaspermum bracteatum,
Baeckea frutescens, and Pinanga rumphiana) were strongly associated with H. costata,
indicating the same habitat requirements and ecological preferences. However, it is
still unclear whether rapid drainage or intolerance to low nutrient content determines
the occurrence and abundance of H. costata and what factors drive the interspecific
association. If intolerance to low nutrients is the case, the absence of the palm from
sites with high nutrient contents may be due to rapid water shortage, particularly on
steep slopes and hill tops. Otherwise, it may be due to its slow intrinsic growth rates
during the seedling stages, excluding it from sites where plants with faster growth
rates predominate. All of these are possible explanations of the research findings and
thus further research is recommended. The information gleaned from this study will
be useful to reserve managers to quantify the palm’s occurrence and abundance in the
reserve, guide an effective monitoring program, and possible use of the palm as an
indicator of habitat conditions.
ACKNOWLEDGEMENTS. I thank Didit Okta Pribadi, Wihermanto, Saripudin, Sudarsono,
Supardi (Bogor Botanic Gardens), Rustandi (Cibodas Botanic Gardens), Deden Mudiana
(Purwodadi Botanic Gardens), and I Gede Tirta (Bali Botanic Gardens) for their able assistance
and cooperation. The map was drawn by Didit Okta Pribadi. I also greatly appreciate Dr.
Irawati and Dr. Hery Harjono for support and encouragement. Acknowledgements also go to
Ir. Kurung, M.M. for permission to enter the East Waigeo Nature Reserve, and Djefri Tibalia,
Alberth Nebore, Irman Meilandi, Kris and Husen (CI Sorong), Gustab Gaman and Sakeus
Dawa (Waifoi village) for their help and cooperation.
References
Armbruster, W.S. (1995) The origins and detection of plant community structure:
reproductive versus vegetative processes. Folia Geobot. Phytotax. 30: 483-497.
Ashton, P.S. (1998) Niche specificity among tropical trees: a question of scales.
In: Newbery, D.M., Prins, H.H.T. & Brown, N.D. (eds) Dynamics of Tropical
Communities. Oxford: Blackwell Science.
Badan Perencanaan Pembangunan Nasional - Bappenas (2003) /ndonesia Integrated
Biodiversity Strategy and Action Plan (IBSAP). Jakarta.
Baillie, I.C, Ashton, P.S., Court, M.N., Anderson, J.A.R., Fitzpatrick, E.A. & Tinsley, J.
(1987) Site characteristics and the distribution of tree species in mixed dipterocarp
forest on tertiary sediments in Central Sarawak. J. Trop. Ecol. 3: 201-220.
Hydriastele ecology in Waigeo, West Papua 371
Begon, M., Harper, J.L. & Townsend, C.R. (1996) Ecology: Individuals, Populations
and Communities, 3rd Ed. Oxford: Blackwell.
Bell, G. (2000) Neutral Macroecology. Science 293: 2413-2418.
Campbell, G.S. (1985) Soil Physics with BASIC: Transport Models for Soil — Plant
Systems. Amsterdam: Elsevier.
Chisholm, R.A. & Burgman, M.A. (2004) The unified neutral theory of biodiversity
and biogeography: Comment. Ecology 85(11): 3172-3174.
Christie, D.A. & Armesto, J.J. (2003) Regeneration microsites and tree species
coexistence in temperate rain forests of Chiloe Island, Chile. J. Ecol. 91: 776-784.
Clark, D.B., Clark, D.A. & Read, J.M. (1998) Edaphic variation and the mesoscale
distribution of tree species in a neotropical rain forest. J. Ecol. 86: 101-112.
Clark, D.B., Palmer, M.W. & Clark, D.A. (1999) Edaphic factors and the landscape-
scale distributions of tropical tree species. Ecology 80: 2662-2675.
Condit, R., Ashton, P.S., Baker, P., Bunyavejchewin, S., Gunatilleke, S., Gunatilleke,
N., Hubbell, S.P., Foster, R.B., Itoh, A., LaFrankie, J.V., Lee, H.S., Losos, E.,
Manokaran, N., Sukumar, R. & Yamakura, T. (2000) Spatial patterns in the
distribution of tropical trees. Science 288: 1414-1418.
Conservation International (1999) The Irian Jaya Biodiversity Conservation Priority-
Setting Workshop: Final Report. Washington, D.C.: Conservation International.
Cox, G.W. (1974) Laboratory Manual of General Ecology 2nd Ed. U.S.A.
Cropper, S.C. (1993) Management of Endangered Plants. Melbourne: Jenkin Buxton
Printers.
Davie, J. & Sumardja, E. (1997) The protection of forested coastal wetlands in Southern
Sumatra: a regional strategy for integrating conservation and development. Pacific
Conservation Biol. 3: 366-378.
Duivenvoorden, J.F. (1995) Tree species composition and rain forest — environment
relationships in the middle Caqueta, Colombia, NW Amazon. Vegetatio 120: 91—
113.
Duivenvoorden, J.F. (1996) Patterns of tree species richness in rain forests of the
middle Caqueta area, Colombia, NW Amazonia. Biotropica 28: 142-158.
Edmunds, J., Cushing, J.M., Costantino, R.F., Henson, S.M., Dennis, B. & Desharnais,
R.A. (2003) Park’s Tribolium competition experiments: a non-equilibrium species
coexistence hypothesis. J. Animal Ecol. 72: 703-712.
Fong, F.W. (1977) Edaphic conditions under Bertam (Eugeissona tristis Griff.) on
Seraya Ridge Forests. Trop. Ecol. 18: 60-70.
Gartlan, S., Newbery, D.M., Thomas, D.W. & Waterman, P.G. (1986) The influence
of topography and soil phosphorus on the vegetation of Korup Forest reserve,
Cameroon. Vegetatio 65: 131—148.
Gentry, A.H. (1988) Changes in plant community diversity and floristic composition
on environmental and geographical gradients. Ann. Missouri Bot. Gard. 75: 1-34.
Groeneveld, J., Enright, N.J., Lamont, B.B. & Wissel, C. (2002) A spatial model of
coexistence among three Banksia species along a topographic gradient in fire-
prone shrublands. J. Ecol. 90: 762-774.
B72 Gard. Bull. Singapore 63(1 & 2) 2011
Hall, J.S.. McKenna, J.J., Ashton, P.M.S. & Gregoire, T.G. (2004) Habitat
characterizations underestimate the role of edhapic factors controlling the
distribution of Entandrophragma. Ecology 85: 2171-2183.
Harms, K.E., Wright, S.J., Calderon, O., Hernandez, A. & Herr, E.A. (2000) Pervasive
density-dependent recruitment enhances seedling diversity in a tropical forest.
Nature 404: 493-495.
Hatfield, J.S., Link, W.A., Dawson, D.K. & Lindquist, E.L. (1996) Coexistence
and community structure of tropical trees in a Hawaiian montane rain forest.
Biotropica 28: 746-758.
Heijden, M.G.A. van der, Wiemken, A. & Sanders, I.R. (2003) Different arbuscular
mycorrhizal fungi alter coexistence and resource distribution between co-
occurring plant. New Phytol. 157: 569-578.
Higgs, A.J. & Usher, M.B. (1980) Should reserves be large or small? Nature 285:
568-569.
_ House, A.P. (1984) The Ecology of Oncosperma horridum on Siberut Island, Indonesia.
Principes 28(2): 85-89.
Hubalek, Z. (1982) Coefficients of association and similarity based on binary (presence-
absence) data: an evaluation. Biol. Rev. 57: 669-689.
Hubbell, S.P. (2001) The Unified Neutral Theory of Biodiversity and Biogeography:
Monographs in Population Biology. Princeton, NJ: Princeton University Press.
Janson, S. & Vegelius, J. (1981) Measures of ecological association. Oecologia 49:
371-376.
Jepson, P. & Whittaker, R.J. (2002) Ecoregions in Context: A critique with special
reference to Indonesia. Conservation Biol. 16: 42—57.
Johns, R.J. (1995) Malesia — An introduction. Curtis's Botanical Magazine 12(2):
53-62.
Johns, R.J. (1997) Background Papers for the Study of the Flora and Vegetation of the
N. E. Kepala Burung, Irian Jaya, Indonesia. Kew: Royal Botanic Gardens, Kew.
Kahn, F. & Mejia, K. (1990) Palm communities in wetland forest ecosystems of
Peruvian Amazonia. Forest Ecol. Managem. 33 & 34: 169-179.
Krebs, C.J. (1989) Ecological Methodology. New York: Harper & Row Publishers.
Ludwig, J.A. & Reynolds, J.F. (1988) Statistical Ecology, A Premier on Methods and
Computing. New York: John Wiley & Sons.
Mohler, C.L. (1990) Co-occurrence of Oak Subgenera: implications for niche
differentiation. Bull. Torrey Bot. Club 117: 247-255.
Molino, J.F. & Sabatier, D. (2001) Tree diversity in tropical rain forests: a validation of
the intermediate disturbance hypothesis. Science 294: 1702-1704.
Moraes, M.R. (1996) Diversity and distribution of palms in Bolivia. Principes 40(2):
75-85.
Morisita, M. (1959) Measuring of interspecific association and similarity between
communities. Mem. Fac. Sci. Kyushu Univ., Ser. E (Biology) 3(1): 65-80.
Mueller-Dombois, D. & H. Ellenberg (1974) Aims and Methods of Vegetation Ecology.
New York: John Wiley & Sons.
Hydriastele ecology in Waigeo, West Papua 373
Nakashizuka, T. (2001) Species coexistence in temperate, mixed deciduous forests.
Trends Ecol. Evol. 16(4): 205—210.
Olsen, S.R. & L.E. Sommers (1982) Phosphorus. In: Page, A.L. (ed) Methods in
Soil Analysis, Part 2 Chemical and Microbiological Properties. Madison, WI:
American Society of Agronomy.
Palmiotto, P.A., Davies, S.J., Vogt, K.A., Ashton, M.S., Vogt, D.J. & Ashton, PS.
(2004) Soil-related habitat specialization in dipterocarp rain forest tree species in
Borneo. J. Ecol. 92: 609-623.
Pemerintah Kabupaten Raja Ampat dan Conservation International Indonesia (2006)
Atlas Sumberdaya Wilayah Pesisir Kabupaten Raja Ampat, Provinsi Irian Jaya
Barat. Sorong: Pemerintah Kabupaten Raja Ampat dan CI Indonesia.
Phillips, V.D. (1998) Peatswamp ecology and sustainable development in Borneo.
Biodivers. Conserv. 7: 651-671.
Potts, M.D., Ashton, P.S., Kaufman, L.S. & Plotkin, J.B. (2002) Habitat patterns in
tropical rainforests: a comparison of 105 plots in Northwest Borneo. Ecology 83:
2782-2797.
Poulsen, A.D. (1996) Species richness and density of ground herbs within a plot of
lowland rainforest in north-west Borneo. J. Trop. Ecol. 12: 177-190.
Real, R. & Vargas, M. (1996) The probabilistic basis of Jaccard’s Index of Similarity.
Syst. Biol. 45: 380-385.
Richter, D.D. & Babbar, L.I. (1991) Soil diversity in the tropics. Advanced Ecol.
Restoration 21, 315-389.
Schluter, D. (1984) A variance test for detecting species associations with some
example applications. Ecology 65: 998-1005.
Silvertown, J. & Law, R. (1987) Do plants need niches? Some recent developments in
plant community ecology. Trends Ecol. Evol. 2: 24-26.
Silvertown, J., Holtier, S., Johnson, J. & Dale, P. (1992) Cellular automaton models
of interspecific competition for space — the effect of pattern on process. J. Ecol.
80: 527-534.
Soil Research Center (1983) Classifications and criteria for soil chemical properties.
Bogor, Indonesia: Ministry of Agriculture, Research and Development Agency
for Agriculture.
Sokal, R.R. & Rohlf, F.J. (1981) Biometry: The Principles and Practice of Statistics in
Biological Research. New York: W.H. Freeman & Co.
Sollins, P. (1998) Factors influencing species composition in tropical lowland rain
forest: does soil matter? Ecology 79: 23-30.
Suarez, D.L. (1996) Magnesium and calcium. In: Bigham, J.M. (ed) Methods of
Soil Analysis, Part 5 Chemical Methods. Madison, WI: Soil Science Society of
America. Pp. 575-602.
Svenning, J.C. (1999) Microhabitat specialization in a species-rich palm community
in Amazonian Ecuador. J. Ecol. 87: 55—65.
Svenning, J.C. (2001) On the role of microenvironmental heterogeneity in the ecology
and diversification of Neotropical rain forest palms (Arecaceae). Bot. Rev. 67:
1-53.
374 Gard. Bull. Singapore 63(1 & 2) 2011
Swaine, M.D. (1996) Rainfall and soil fertility as factors limiting forest species
distributions in Ghana. J. Ecol. 84: 419-428.
Tarbuck, E.J. & Lutgens, F.K. (2004) Earth Science 9th Ed. (Chapter 18. Climate).
Pearson Prentice Hall Companion Website.
Tiessen, H. (1998) Resilience of phosphorus transformations 1n tropical forests and
derived ecosystems. In: Schulte, A. & Ruhiyat, D. (eds) Soils of Tropical Forest
Ecosystems Characteristics, Ecology and Management. Berlin: Springer-Verlag.
Pp. 92-98.
Tomlinson, P.B. (1979) Systematics and ecology of the Palmae. Ann. Rev. Ecol. Evol.
Syst. 10: 85-107.
Tomlinson, P.B. (1990) The Structural Biology of Palms. Oxford: Clarendon Press.
Tucker, R. (1992) Experiences with Cyrtostachys renda. Mooreana 2: 11-16.
Uhl, N.W. & Dransfield, J. (1987) Genera Palmarum: A Classification of Palms Based
on the Work of Harold E. Moore, Jr. Lawrence, Kansas: L. H. Bailey Hortorium
& International Palm Society.
Vitousek, P. & Sanford, R.L. (1986) Nutrient cycling in moist tropical forests. Ann.
Rey. Ecol. Evol. Syst. 17: 137-167.
Vormisto, J. (2002) Palms as rainforest resources: how evenly are they distributed in
Peruvian Amazonia? Biodivers. Conserv. 11: 1025-1045.
Webb, C.O. (2005) Vegetation of the Raja Ampat Islands, Papua, Indonesia. 4 Report
to the Nature Conservancy. Revision 1.5.
Webb, C.O. & D.R. Peart (2000) Habitat associations of trees and seedlings in a
Bornean rain forest. J. Ecol. 88: 464-478.
White, R.E. (1997) Principles and Practice of Soil Science. The Soil as a Natural
Resource, 3rd Ed. Oxford: Blackwell Science.
Widyatmoko, D. (2001) Autecology and Conservation Management of a Rare Palm
Species: The Case Study of Lipstick Palm Cyrtostachys renda Blume in Kerumutan
Wildlife Sanctuary, Sumatra. Ph.D. Thesis. Bogor: Bogor Agricultural University.
Widyatmoko D. (2009) Interspecific association patterns and edaphic factors’
influences: a case study of Orania regalis Zippelius in Waigeo island, West Papua.
Berita Biologi 9(5): 595—607.
Widyatmoko, D. & Burgman, M.A. (2006) Influences of edaphic factors on the
distribution and abundance of a rare palm (Cyrtostachys renda) in a peat swamp
forest in eastern Sumatra, Indonesia. Austral Ecol. 31(8): 964-974.
Widyatmoko, D., Burgman, M.A., Guhardyja, E., Mogea, J.P., Walujo, E.B. & Setiadi, D.
(2005) Population status, demography and habitat preferences of the threatened
lipstick palm Cyrtostachys renda Blume in Kerumutan Reserve, Sumatra. Acta
Oecol. 28: 107-118.
Widyatmoko, D., Pribadi, D.O., Wihermanto, Saripudin, Sudarsono, Supardi,
Rustandi, Mudiana, D. & Tirta, I.G. (2007) Exploration and plant diversity study
in the East Waigeo Nature Reserve and Sorong Nature Park, West Papua: An
Expedition Report (unpublished). Bogor: Center For Plant Conservation, Bogor
Botanic Gardens & LIPI.
Gardens’ Bulletin Singapore 63(1 & 2): 375-383. 2011 375
Hybrid zone characteristics of the
intergeneric hybrid bamboo
x Gigantocalamus malpenensis (Poaceae: Bambusoideae)
in Peninsular Malaysia
K.M. Wong! and Y.W. Low
Singapore Botanic Gardens, | Cluny Road, Singapore 259569
'wkm2000@gmail.com
ABSTRACT. The natural intergeneric hybrid bamboo * Gigantocalamus malpenensis and its
parental species, Dendrocalamus pendulus and Gigantochloa scortechinii, were mapped along
a 4-km stretch of the Gombak valley in Peninsular Malaysia. Hybrid clumps were associated
only with D. pendulus, which implied the latter was the seed parent, and not with G. scortechinii,
which occupied only broader, gentler sites available for its establishment at the southwest and
mid-southern parts of a southwest-to-northeast trending corridor of disturbance associated with
expressway construction some three decades before. The effects of landscape-level disturbance
on vegetation and plant demographic changes, gene flow and breeding dynamics are discussed.
Keywords. Bamboo, Dendrocalamus, disturbance, * Gigantocalamus, Gigantochloa, hybrid
zone, hybridisation, Peninsular Malaysia
Introduction
A bamboo morphologically intermediate between Gigantochloa scortechinii Gamble
and Dendrocalamus pendulus Ridl., two common bamboo species in the foothills
of the Peninsular Malaysian Main Range, was discovered in two different localities
(Perak and Selangor states) in Peninsular Malaysia (Goh et al. 2011). The hybrid
nature of the morphologically intermediate taxon was demonstrated using partial
Granule-Bound Starch Synthase (GBSS) I sequences, and this natural intergeneric
hybrid was named = Gigantocalamus malpenensis K.M.Wong (Goh et al. 2011). Of
its two known localities, the second, in the Gombak valley in Hulu Gombak, Selangor,
Peninsular Malaysia, included a substantial number of clumps along a significant span
of the valley, and it was decided to attempt to understand the population characteristics
better, especially when its parental populations were also well represented in the same
valley.
The reproductive behaviour of this hybrid, especially in comparison to its
wild parental species, was of special interest. While entirely vegetative stands of
both Gigantochloa scortechinii and D. pendulus are frequently encountered (Wong
199Sa, b), they are known to flower gregariously, i.e., synchronised flowering of a
significant number of clumps within a localised population (Burgess 1975; Wong
199Sa, b), although ‘diffuse-sporadic flowering’ (isolated whole-clump flowering at
376 Gard. Bull. Singapore 63(1 & 2) 2011
irregular intervals in natural populations) is also known (Wong 1995a, b). Caryopses
(the one-seeded fruits of bamboos) and seedlings are easily encountered in the wild
amid gregarious flowerings of these two species. However, a certain degree of self-
incompatibility would be expected because caryopses and seedlings have not been
found with isolated flowering clumps of G. scortechinii; and only a very low level of
caryopsis formation was found with the flowering of an isolated clump of G. rostrata
K.M.Wong, a related species (Wong 1995b). Goh et al. (2011) reported that a solitary
cultivated clump of the hybrid from Perak had flowered quite soon after it grew into
mature size and died completely afterwards, with no caryopses formed. Flowering
and fruiting of the hybrid in the wild has not been recorded until the discovery of the
Selangor population at Hulu Gombak.
Study site, methods and materials
The study site was in the Hulu Gombak area (101°44°30” to 101°46’E, 3°19°15” to
3°20°25”N) of Peninsular Malaysia’s Selangor state. The valley was steep-sided but
the terrain beside the Gombak River was around 250-400 m elevation in this area.
The portion of the Gombak valley investigated resembled a southwest-to-northeast
corridor, flanked by the Karak Expressway to the north of the Gombak River, and by
the smaller, old Gombak Road to the south. The construction of the Karak Expressway
just before 1980 was impactful, involving many slope cuts, infills and bridging across
generally steep inclines. Wayside spaces abutting the Expressway and between that and
the Gombak River are frequently occupied by successional stands of low vegetation,
and there are patches of Acacia mangium Willd. cultivated for forestry or site-greening
purposes but occasionally escaping.
We used a GARMIN GPSMAP 60CSx handheld GPS unit to obtain satellite
readings of location coordinates for clump positions of the hybrid x Gigantocalamus
malpenensis and its parental species, Gigantochloa scortechinii and D. pendulus,
along a 4-km stretch of the Gombak valley, where these bamboos occurred in
September, 2009. Clumps were noted as ‘vegetative’ (complete lack of flowering
activity), ‘flowered’ (with signs of past/recent flowering but no fresh flowers seen), or
‘flowering’ (with fresh flowering persisting). An estimate of relative clump maturity
was obtained by classifying clumps as ‘mature’, when there were culms exceeding 3.5
cm diameter at 1.5 m height present, or “young”, when there were no culms exceeding
3.5 cm diameter. This is based on the observation that there would be larger culm
internode diameters produced with increasingly older clump age until a maximum that
was representative of the species, so that younger clumps typically produced culms of
smaller diameter (Holttum 1958, Wong 2004).
The coordinates were transferred from the GPS unit to a personal computer
using software provided by the manufacturer. The data was then viewed in a Google
Earth 5.2 scene of the Hulu Gombak area (on 1 Oct 2009) with the location coordinates
obtained represented by symbols plotted onto the satellite view. A tracing of this was
made (Fig. 1) to show the main Gombak River flanked by the Karak Expressway and
A x Gigantocalamus hybrid zone 377
the Gombak Road, as well as required contours to represent the physical landform,
together with plotted points showing the locations of bamboo clumps recorded.
Results
The results of the population census for the three bamboo taxa are summarised in
Table 1, with locations of < Gigantocalamus malpenensis and its parental species,
Gigantochloa scortechinii and D. pendulus, mapped in Fig. 1.
Distribution of taxa across the hybrid zone
The respective sites of the parental species were interesting: G. scortechinii occupied
flatter or gentler, streamside places along the Gombak River, adequately represented at
the southwestern and mid-southern parts of the valley, whereas D. pendulus typically
occurred on steeper hillsides along the same valley. Hybrid individuals occurred
together with D. pendulus clumps towards the northeastern part of the valley, and not
with the G. scortechinii population.
Bamboo flowering and seeding during the census
There were 55 individual clumps of D. pendulus (incl. | young clump as defined for
this study) and 67 clumps of G. scortechinii (no young clumps) recorded in the study
area. All were vegetative, showing no sign of flowering.
Among 48 clumps of the hybrid, 35 (c. 73%) were vegetative at the time of
the census and 13 were recently in flower or were still flowering (Table 1). These
Table 1. Observed clumps of the hybrid x Gigantocalamus malpenensis and its parental species
Dendrocalamus pendulus and Gigantochloa scortechinii along the Hulu Gombak valley,
Peninsular Malaysia: summary of some biological attributes (vegetative vs. reproductive states;
relative maturity). Clumps were noted as Vegetative, Flowered (signs of recent flowering but
no fresh flowers seen), or Flowering (with fresh flowering persisting). Clump relative maturity:
Mature (indicated by presence of culms exceeding 3.5 cm diameter at 1.5 m height) or Young
(no culms exceeding 3.5 cm diameter present).
Taxon Number of | Clumps in Live clumps that Dead
recorded vegetative state had flowered / were clumps
clumps flowering during that had
census flowered
Dendrocalamus 55 55 (incl. 1 young 0 0
pendulus clump)
Gigantochloa 67 67 (incl. 0 young 0 0
scortechinii clumps)
Hybrid 48 35 (incl. 7 young 7 had flowered + 3
(« Gigantocalamus clumps) 3 still flowering
malpenensis)
378 Gard. Bull. Singapore 63(1 & 2) 2011
N 3°20'13,92"
N3°19'22.08"
E 101°45'1.44" E 101°45'53.28"
Fig. 1. The Gombak River, Gombak Road and Karak Expressway (thick lines), and locations
of the hybrid = Gigantocalamus malpenensis (dots) and its parental species, Dendrocalamus
pendulus (squares) and Gigantochloa scortechinii (triangles), in Hulu Gombak, Selangor,
Peninsular Malaysia.
vegetative clumps included only seven young clumps (1.e., clumps without any
culms exceeding 3.5 cm diameter). It is reasonable to assume that this represented
an establishing hybrid population originating from possibly just 1—2 parental seeding
events, which must be rare, but when the flowering of both parental species on site
must have coincided. Flowering in the hybrid population appears to have involved
only a small proportion of hybrid individuals.
It stands to reason that the population of hybrid clumps found probably
represented Fl hybrid material. Seeding of the Fl must have been absent or very
limited, and seedling survival negligible, because caryopses were not found among
the flowering material, and young seedlings were absent on site among the flowering
hybrid clumps, in spite of a good number of these (10 out of 13) having completed
flowering and not having any more fresh flowering branches on them during the
census. Thus, the presence of an F2 generation was not detected although a cohort of
flowering F1 hybrids was present.
Discussion
Hybrid zone characteristics
The near segregation of the parental populations was compatible with what is known
about their ecological distribution: D. pendulus is more typical of steeper hillsides,
A = Gigantocalamus hybrid zone 379
whereas G. scortechinii prefers gentler slopes and frequently spreads onto the
disturbed and logged-over plains (Wong 1995b). The availability of sites suitable
for G. scortechinii thus appeared to constrain establishment of that species to the
southwestern and mid-southern parts of the study area. In particular, a protruding ridge
in the middle of the study area had an abundance of D. pendulus established on its
steep sides (Fig. 1).
The distribution of the hybrid relative to the two parental species (the hybrid
being associated with D. pendulus) suggested that D. pendulus was the maternal parent
for the hybrid cohort investigated. Caryopses are expected to have poorer dispersal
ability compared to pollen, and bamboo pseudospikelets and caryopses mostly fall
around the parent clump (Ridley 1930: Wong 1995a, b). A previous attempt at deducing
the direction of the cross between the parental species, using cpDNA that was likely to
have been maternally inherited and including a broader geographical sampling of the
species, was equivocal probably because of complex patterns of inheritance for which
past reciprocal crosses between taxa followed by introgression (sensu Anderson 1948)
could not be ruled out (Goh et al. 2011).
Ecological implications
Why were these hybrids not detected earlier, although this part of the Gombak valley
has been a routinely well-botanised locality? The hybrid bamboo was not observed
at all during the several visits annually to the study area from 1980, when the Karak
Expressway construction was being completed, to 1988, and 1996-2007 (K.M. Wong.
pers. obs.). Hybrid population characteristics, particularly its distribution. suggest
comparatively recent establishment. We suggest that the “corridor of change” brought
about by development (i.e. opening up) of the Karak Expressway (parallel to this part
of the Gombak valley) just before 1980 was a key factor that may have predisposed the
hybridisation event. This enormous cleared corridor caused greater openness, spread
and an increased abundance of both D. pendulus and G. scortechinii (and other pioneer
or early successional plant species), as well as the removal of forest tree cover that
probably served as a natural impediment to genetic exchange between the bamboo
species. In the dense vegetation of tropical rain forest, high humidity could dampen
pollen, increasing their difficulty in remaining airborne, and rain often removes
airborne pollen; also the dense tree canopy tends to be an effective filter of airborne
pollen (Turner 2001).
Greater exposure along this corridor had probably increased chances for
air currents carrying pollen from G. scortechinii clumps in flatter terrain at one
end (southwest) to find hillside patches occupied by D. pendulus near the other
end (northeast). The tendency for both D. pendulus and G. scortechinii to flower
gregariously or diffuse-sporadically from time to time probably increased the
possibility of coincident flowering and cross-fertilisation.
Will the hybrid persist? At the time of the census, it could be said that the
hybrid was beginning to establish. It was represented by 48 clumps, but over 70%
remained vegetative while 13 clumps had come into flower. This suggests there could
be some variability with respect to age at onset of flowering. The distribution of hybrid
380 Gard. Bull. Singapore 63(1 & 2) 2011
offspring clumps among so many maternal clumps (individuals) in this case also
implies a range of resulting heterozygosity in the hybrid offspring that would bring
increased variability even in flowering time (or its associated vegetative longevity).
There were also several clumps that had flowered but which were producing new culm
shoots. The survival of such new culm shoots was not certain, but again represented
a potential for regeneration following flowering, which has been noted for clumping
bamboos that exhibit whole-clump flowering (Wong 1995a, b). Thus, there appears to
be a good prospect for the survival of the hybrid.
Hybrids are probably more common among related species (e.g., Okada 1990,
Kiew et al. 2003, Gravendeel et al. 2004) than between genera, but intergeneric crosses
are also known for other plant groups (e.g., McKenzie et al. 2008). Plant taxa have also
been surmised to have originated from intergeneric hybridisation events (Tara 1977,
Wallace & Jansen 1995) and hybridisation between genetically divergent lines have
been known to give more vigorous progeny (Edmands 2002). The potential for viable
_ backcrosses and advanced-generation hybrids cannot be presently estimated.
Conclusion
Continuing research also examines generic boundaries among the group of
bamboos classified as the Bambusinae subtribe, which includes Dendrocalamus
and Gigantochloa. This includes the difficulties of classifying a mystifying complex
of bamboos that have defied simple lineage studies so far (e.g., Yang et al. 2008,
Stapleton et al. 2009, Yang et al. 2010, Goh et al. 2010), so that the use of more gene
regions, including from both nuclear and plastid domains, would be important. The
present case study of hybridisation between D. pendulus and G. scortechinii (Goh et
al. 2011, this study) confirms that natural hybrids do form in the tropical Southeast
Asian landscape, especially when ecological barriers between taxa change in nature.
Elsewhere, Clark et al. (1989) and Triplett et al. (2010) have documented interspecific
hybridisation among American bamboos. We have suggested (Wong 2004, Goh et
al. 2010, Goh et al. 2011; Wong et al., in prep.) that the role of hybridisation in the
evolution of Tropical Asian bamboos could have been underestimated, mostly because
such hybrids are difficult to detect and confirm. Hybridisation is a significant feature
of gene flow and evolutionary processes (Rieseberg 1995, Arnold et al. 2001) and
extensive adaptive radiation can occur after hybridisation (Givnish 2010). It should be
expected that among highly complex taxonomic groups such as the woody bamboos,
hybridisation and introgression studies would hold the key to a better understanding.
In conservation terms, we have hardly any documentation of the effects of
environmental change (especially through vegetation removal or degradation) on the
reproductive behaviour and gene flow among plants in Southeast Asia. The primary
concern Is perhaps correctly placed on the decimation or reduction of plant populations
in an affected area (Laurance et al. 1997), giving emphasis on potentially adverse
effects associated with breakdown of long-established and naturally viable pollinator,
breeding, dispersal and establishment biology (e.g., Washitani 2000). This study has
A x Gigantocalamus hybrid zone 38]
demonstrated how landscape-level changes equivalent to that seen with development
in the Hulu Gombak area may predispose the plant life to changes in reproductive
and ecological behaviour. In this case, the chance of hybridisation occurring between
D. pendulus and G. scortechinii, the two abundant species of bamboo in the Hulu
Gombak valley, was probably increased in the aftermath of landform and vegetation
changes, facilitated by removal of natural physical barriers that have helped maintain
reproductive isolation of the species, and the products of hybridisation began
establishing only nearly 30 years following disturbance. Changes in the environment
would be expected to alter relative species abundances, plant establishment, and
processes or patterns (including direction and symmetry) of gene flow and breeding.
Hybridisation would bring along fitness consequences influenced by the extent and
quality of the breeding interaction.
ACKNOWLEDGEMENTS. We thank Professor Nianhe Xia (South China Institute of Botany)
and Ms. Wei Lim Goh (University of Malaya) for their comments on earlier drafts.
References
Anderson, E. (1948) Hybridization of the habitat. Evolution 2: 1-9.
Arnold, M.L., Kentner, E.K., Johnston, J.A., Cornman, S. & Bouck, A.C. (2001)
Natural hybridization and fitness. Jaxon 50: 93-104.
Burgess, P.F. (1975) Silviculture in the Hill Forests of the Malay Peninsula. Peninsular
Malaysian Forestry Department Research Pamphlet No. 66. Kepong: Forest
Research Institute Malaysia.
Clark, L.G., Davidse, G. & Ellis, R.P. (1989) Natural hybridization in bamboos:
evidence from Chusquea sect. Swallenochloa (Poaceae: Bambusoideae). Nail.
Geogr. Res. 5: 459-476.
Edmands, S. (2002) Does parental divergence predict reproductive compatibility?
Trends Ecol. Evol. 17: 520-527.
Givnish, T.J. (2010) Ecology of plant speciation. Jaxon 59: 1326-1366.
Goh, W.L., Chandran, S., Lin, R.-S., Xia, N.-H. & Wong, K.M. (2010) Phylogenetic
relationships among Southeast Asian climbing bamboos (Poaceae: Bambusoideae)
and the Bambusa complex. Biochem. Syst. Ecol. 38: 764-773.
Goh, W.L., Chandran, S., Kamiya, K. & Wong, K.M. (2011) A natural hybrid
between Dendrocalamus pendulus and Gigantochloa scortechinii (Poaceae:
Bambusoideae: Bambuseae) in Peninsular Malaysia. Gard. Bull. Singapore 62:
223-238.
Gravendeel, B., Eurlings, M.C.M., Berg, C. van den & Cribb, P.J. (2004) Phylogeny
of Pleione (Orchidaceae) and parentage analysis of its wild hybrids based on
plastid and nuclear ribosomal ITS sequences and morphological data. Syst. Bot.
29: 50-63.
Holttum, R.E. (1958) The bamboos of the Malay Peninsula. Gard. Bull. Singapore 16:
1-135.
382 Gard. Bull. Singapore 63(1 & 2) 2011
Kiew, R., Teo, L.L. & Gan, Y.Y. (2003) Assessment of the hybrid status of some
Malesian plants using Amplified Fragment Length Polymorphism. Telopea 10:
225-233.
Laurance, W.F., Bierregaard, R.O., Gascon, C., Didham, R.K., Smith, A.P., Lynam,
A.J., Viana, V.M., Lovejoy, T.E., Sieving, K.E., Sites, J.W.J., Andersen, M.,
Tocher, M.D., Kramer, E.A., Restrepo, C. & Moritz, C. (1997) Tropical forest
fragmentation: Synthesis of a diverse and dynamic discipline. In: Laurance, W.F.
& Bierregaard, R.O.J. (eds) Tropical Forest Remnants: Ecology, Management,
and Conservation of Fragmented Communities, pp. 515-525. Chicago: University
of Chicago Press.
McKenzie, R.J., Ward, J.M. & Breitwieser (2008) Hybridization beyond the F1
generation between the New Zealand endemic everlastings Anaphalioides
bellidioides and Ewartia sinclairii (Asteraceae, Gnaphalieae). P/. Syst. Evol. 273:
13-24.
Okada, H. (1990) A natural hybrid of Monophyllaea (Gesneriaceae) in the tropical rain
forests of West Sumatra. P/. Syst. Evol. 169: 55-63.
Ridley, H.N. (1930) The Dispersal of Plants Throughout the World. Kent, Ashford:
Reeve & Co.
Rieseberg, L.H. (1995) The role of hybridization in evolution: old wine in new skins.
Amer. J. Bot. 82: 944-953.
Stapleton, C.M.A., Ni Chonghaile, G. & Hodkinson, T.R. (2009) Molecular phylogeny
of Asian woody bamboos: review for the Flora of China. J. Amer. Bamboo Soc.
22: 5—25.
Tara, M. (1977) Cytogenetic studies on natural intergeneric hybridization in Aster
alliances III]. Experimental confirmation of the hybrid origin of Aster ageratoides
subsp. ovatus. Bot. Mag. Tokvo 90: 253-258.
Triplett, J.K., K.A. Oltrogge, L.G. Clark. 2010. Phylogenetic relationships and
natural hybridization among the North American woody bamboos (Poaceae:
Bambusoideae: Arundinaria). Amer. J. Bot. 97: \—22.
Turner, I.M. (2001) The Ecology of Trees in the Tropical Rain Forest. U.K., Cambridge:
Cambridge University Press.
Wallace, R.S. & Jansen, R.K. (1995) DNA evidence for multiple origins of intergeneric
allopolyploids in annual Microseris (Asteraceae). P/. Syst. Evol. 198: 253-265.
Washitani, I. (2000) Creeping ‘fruitless falls’: Reproductive failure in heterostylous
plants in fragmented landscapes. In: Kato, M. (ed) The Biology of Biodiversity,
pp. 133-145. Tokyo: Springer-Verlag.
Wong, K.M. (1995a) The Morphology, Anatomy, Biology and Classification of
Peninsular Malaysian Bamboos. University of Malaya Botanical Monographs
No. |. Kuala Lumpur: University of Malaya.
Wong, K.M. (1995b) The Bamboos of Peninsular Malaysia. Malayan Forest Records
No. 41. Kepong: Forest Research Institute Malaysia.
Wong, K.M. (2004) Bamboo, The Amazing Grass. A Guide to the Diversity and
Study of Bamboos in Southeast Asia. Kuala Lumpur: International Plant Genetic
Resources Inst. (IPGRI) & University of Malaya.
_ A Gigantocalamus bybrid zone 383
Yang, H.-Q.., Yang, J.-B., Peng, Z.-H., Gao, J., Yang. Y.-M., Peng, S. & Li, D.-Z. (2008)
A molecular phylogenetic and fruit evolutionary analysis of the major groups of
the paleotropical woody bamboos (Gramineae: Bambusoideae) based on nuclear
ITS, GBSSI gene and plastid t7ml-F DNA sequences. Molec. Phylogenet. Evol.
48: 809-824.
Yang, J.-B., Yang, H.-Q., Li, D.-Z.. Wong, K.M. & Yang, Y.-M. (2010) Phylogeny of
Bambusa and its allies (Poaceae: Bambusoideae) inferred from nuclear GBSSI
gene and plastid psbA-trnH, rpl32-trnL, rps 16 intron DNA sequences. Taxon 59:
1102-1110.
'
>
1
( : I
‘
a o>
Ys -
a.
2 me
Fi -
2 all
it
Gardens’ Bulletin Singapore 63(1 & 2): 385-393. 2011 385
The ecology of ultramafic areas in Sabah:
threats and conservation needs
A. van der Ent
Centre for Mined Land Rehabilitation, Sustainable Minerals Institute,
The University of Queensland, Brisbane, QLD 4072, Australia
a.vanderent(@uq.edu.au
ABSTRACT. Ultramafics are characterised by high concentrations of magnesium and nickel,
low concentrations of calcium, low water retention capacity and low concentrations of essential
plant nutrients in soils derived from this substrate. These extreme chemical soil condition force
plants to adapt to survive. Sabah is one of the richest areas in the world for plant diversity on
ultramafic substrates. A range of species, including a number of pitcher plants (Nepenthaceae),
orchids (Orchidaceae) and trees and shrubs are endemic to ultramafic areas in Sabah, often
occurring on a few, or even just a single, site. Ultramafic vegetation types in Sabah are severely
threatened by land-clearing activities. Although only a small minority of the geological
substrates in Sabah are ultramafic, ecosystems on these substrates have a disproportionately
high number of endemic and rare plant species. Destruction of these types of ecosystems, in
particular, can potentially result in extinction of plant species.
Keywords. Endemism, Mount Kinabalu, rare species, Sabah, serpentine, ultramafics
Introduction
Preliminary research suggests that the Malaysian state of Sabah may be one of the
richest areas in the world for plant diversity on ultramafic substrates. Kinabalu Park,
covering approximately 1200 square kilometres, has in excess of 900 plant species
occurring on ultramafics. This extremely high diversity is a consequence of the immense
biodiversity of the Malesian region itself, as well as the locally diverse edaphic and
climatic conditions. Generally in the Malesian region, ultramafic vegetation types have
lower stature and different species composition compared to lowland dipterocarp-
forest (Proctor & Nagy 1992). The edaphic stresses that ultramafic substrates exert on
plant survival have resulted in high numbers of endemic species on these substrates.
Ultramafic rocks and substrates
Ultramafic (or ‘ultrabasic’) substrates are found all around the world. The often-
used geological term ‘serpentine’ refers to lizardite, antigorite and chrysotile, but
ecologists use it to describe the ecology of soils derived from ultramafic substrates
(Coleman & Jole 1992). Ultramafics are found in tectonic chunks (‘ophiolite suites’),
386 Gard. Bull. Singapore 63(1 & 2) 2011
which are fragments of the upper mantle obducted in continental margins on ocean
floors (Coleman & Jole 1992). In Sabah, ultramafics (predominantly in the form of
the mineral peridotite) that have been raised above sea level weather to iron- and
nickel-rich laterites (Baillie et al. 2000). Such soils are characterised by very high
concentrations of magnesium and nickel, both of which may be phytotoxic, have
low water retention capacity, low available phosphorus concentrations (partly due
to phosphorus-immobilising ferric sesquioxides) and low concentrations of other
essential plant nutrients such as nitrogen and potassium (Baillie et al. 2000).
The ultramafic effect (‘serpentine syndrome’)
The evolutionarily challenging selection forces that ultramafics exert on plants is a
result of the extraordinary geochemistry (Rajakaruna & Baker 2006). The excess of
_ magnesium, and specifically the very low calcium to magnesium ratio (<1) in soils
derived from ultramafic substrates, is an important factor in determining species
composition and structure of vegetation on ultramafic substrates (Brooks 1987). The
foliar concentration of magnesium is often significantly higher in plants growing
on ultramafics, while calcium concentrations are often significantly lower (Brady et
al. 2005). Extremely high nickel concentrations in such soils can also be limiting to
the growth of many plant species. The bioavailable and thus potentialiy phytotoxic
concentration of nickel in the soil depends on soil clay content, organic exudates in the
rooting zone, soil moisture pH and the presence of other divalent ions such as calcium
and magnesium.
Another important factor defining the structure and species composition of
ultramafics 1s water stress. Many plants found on ultramafics have morphological
adaptations to minimise water requirements and water loss in order to survive in
drought conditions (Brady et al. 2005). Such adaptations include a generally low
stature, small-crowned growth-form and other characteristics such as glaucous leaves
and sclerophyllous and microphyllous morphologies. Examples of vegetation types
in which many plants clearly display such adaptations are the graminoid vegetation
on higher elevations on Mt. Tambuyukon and the stunted vegetation at Mt. Silam
(Bruijnzeel et al. 1993). Morphological adaptations that evolved to reduce transpiration
are also likely to minimise nickel and magnesium uptake. The low water retention
capacity of most ultramafic soils might induce water stress, but may also increase the
toxicity of magnesium and nickel by the concentration effect in the soil solution (Proctor
& Nagy 1992). In addition, water stress and drought can increase the susceptibility of
vegetation to fire. Prolonged dry periods that may occur due to the El Nino effect often
result in fires affecting the ultramafic vegetation communities of Mt. Kinabalu. Aiba
& Kitayama (2002) showed that the growth rates of trees on ultramafics are lower
than those at equivalent altitudes on non-ultramafics, but they have also shown that
the mortality rate during droughts 1s distinctly lower on ultramafics. The recurrence
of fires delays succession and therefore prevents development of tall vegetation types.
Furthermore, the burning of leaf litter during fire could also have an effect on the soil
Sabah’s ultramafic areas 38
~)
nutrient status by increasing the availability of phosphorus and mineralising organic
matter.
The environmental and edaphic factors mentioned in the proceeding paragraph
have, taken together. been termed the ‘serpentine syndrome’ (Jenny 1980). a “syndrome”
expressed through the combined characteristic morphology, physiology and ecology
of these plant communities that likely results from a dynamic interplay of the above-
mentioned factors (Brady et al. 2005). Besides the geochemical anomalies, other
prevailing environmental and edaphic conditions may also be important in determining
the specific vegetation composition and structure of such plant communities. Soil depth
and the degree of exposure to wind, for example, may also be contributing factors that
restrict the successful survival of plants on ultramafics.
Hyperaccumulators of nickel in Sabah
Some plants have the remarkable physiological capacity to accumulate shoot nickel
at levels 100-10,000-fold greater than levels in non-accumulators (Lasat 2001). Such
*hyperaccumulators’ are confined to ultramafics because that substrate is the largest
and most widespread metal-enriched habitat on a global scale. It is an exceptionally
rare phenomenon with only 1—2% of plant species found on ultramafics being classed
as hyperaccumulators. As of 2010, some 400 nickel hyperaccumulator species have
been described (Baker 1981, Baker & Smith 2000). Nickel hyperaccumulators are
found in many different genera across at least 45 plant families (Baker et al. 1999). As
such, nickel hyperaccumulators have a great variety of growth forms, ecophysiological
characteristics and ecological requirements (Pollard et al. 2002). The plant families
that are most represented are Euphorbiaceae, Brassicacea, Asteraceae, Flacourtiaceae
(Salicaceae), Buxaceae and Rubiaceae (Reeves 2003), illustrating that this phenomenon
has independently developed multiple times on ultramafics. Nickel hyperaccumulators
are relatively easy to identify by chemical analysis of dried foliage in the laboratory
or in the field by pressing fresh leaves against white test paper impregnated with
dimethylglyoxime (‘DMG’), which is a nickel-specific colorimetric dye (Baker et al..
1992a).
A number of hyperaccumulators have been found in Sabah, including: Rinorea
bengalensis (Wall.) Kuntze, Phyllanthus balgooyi Petra Hoffm. & A.J.M.Baker,
Dichapetalum gelonioides (Roxb.) Engl. ssp. tuberculatum Leenh. and Shorea
tenuiramulosa P.S.Ashton. The small tree or scrambler, Dichapetalum gelonioides
ssp. tuberculatum, has up to 25 mg/g nickel in its dried foliage (Reeves 2003). While
hyperaccumulators are typically recognised on the basis of the concentrations of
metals in the foliage, other plant parts may have very different metal concentrations.
For instance, Reeves (2003) cites values for phloem tissues of the shrub Phyllanthus
balgooyi of up to 90 mg/g nickel on a dry weight basis. This is a widespread shrub
on ultramafics in Sabah (as well as in the Philippines) and it has only recently been
described formally (Hoffmann et al. 2003). Prior to this, it was provisionally recognized
as Phyllanthus ‘palawanesis’ (Baker et al. 1992b).
388 Gard. Bull. Singapore 63(1 & 2) 2011
gay
Epiphytes on ultramafics
Epiphytes are an ecologically important part of the plant diversity on ultramafics in
Sabah, particularly at higher altitudes. Epiphytes may be sensitive to the chemical
composition in the host tissues and may absorb elements, such as nickel, from their host.
However, the precise ecological relationship between epiphytes and their host is very
poorly understood. In New Caledonia, Boyd et al. (2009) found that epiphytes (mosses
and liverworts) growing on nickel hyperaccumulator hosts contained higher levels of
nickel than those growing on non-hyperaccumulator hosts. Even more so, the epiphyte
nickel concentrations often exceeded the threshold of nickel hyperaccumulation (Boyd
et al. 2009). This indicates that the capacity of epiphytes to grow on hyperaccumulators
depends on their capability to tolerate high nickel concentrations, which may in turn
define the spatial and ecological attributes of epiphyte community composition (Boyd
etal. 2009). There is limited evidence from Mt. Kinabalu were several epiphytic orchids
are known only from ultramafic substrates, suggesting important correlations between
the chemical composition of substrate and host. In addition to potential chemical
relationships, the open and low-statured structure (with, as a result, higher sunlight
intensity, higher surface temperatures, lower humidity and higher wind exposure) of
many ultramafic vegetation types is undoubtedly important for the distribution and
relative abundance of epiphytes on ultramafics.
Endemism on ultramafic substrates
Whittaker (1954) categorised three general characteristics common among vegetation
types on ultramafics: (a) low stature and biomass production; (b) high levels of
endemism; and (c) distinct differences from vegetation in surrounding areas. Based
on these characteristics, Whittaker (1954) attributed the ‘serpentine syndrome’ to
three causes: the edaphic or geochemical; the plant species-level (autecology); and
the plant community-level (synecology). The preponderance of endemism of species
occurring on ultramafic substrates in Sabah differs strongly between areas. High levels
of endemism are generally associated with especially isolated areas, either geographic
or altitudinal. Some species may have a high affinity for ultramafic substrates, being
more abundant or even solely confined to such areas, while others have wide ecological
amplitudes and occur on different substrates. Species endemic to ultramafics may
be paleo-endemics; species for which their population have become confined to
ultramafics due to competition elsewhere (Baker & Whiting 2002), or neo-endemics
which have evolved in-situ from closely-related species.
Ultramafics in Sabah
The geology of Sabah, Malaysia, is mainly composed of sedimentary rock, such as
sandstone and shale, but about 3500 square kilometres (4.6% of total landmass) of
Sabah’s ultramafic areas 389
Sabah, Malaysia
South
China Sea
ty Sulu Sea
M
e nt Kinabalu
jou
& © ")
~- A ~,
Meliau Range Re oe s=7G5
x Hills
{ Mount Tavai
ww
Se - ;
= Mount Silam
x C
= Celebes Sea
15 30 N : 4 = ts; A
Fig. 1. Delimitation of ultramafic outcrops based on the map ‘Igneous Rocks of Sabah,
Malaysia’ by Geological Survey, Borneo Region, Malaysia, 1965. In the map, the most
extensive ultramafic outcrops are featured, including Mount Kinabalu, Mount Silam, Mount
Tavai, Meliau Range and Bidu-Bidu Hills.
ultramafics occur (Proctor et al. 1988, Repin 1998). The most extensive ultramafic
outcrops are found in the Meliau Range, at Mt. Tawai, the Bidu-bidu Hills, Mt. Silam
and around Mt. Kinabalu (Collenette 1964) (Fig. 1). Ultramafics are found from
sea level (on islands in the Darvel Bay area) up to nearly 2900 metres above sea
level (on Mt. Kinabalu). Studies of the plant diversity on ultramafics are extremely
limited in Sabah. The Sabah Forestry Department (2005) has undertaken some general
site reconnaissance at Mt. Tawai and the Bidu-Bidu Hills and described three main
types of ultramafic vegetation: (1) lowland ultramafic forest, in which dipterocarps
dominate; (2) hill forest in which Casuarinaceae dominate but also some dipterocarps
are found: and (3) hill and lower montane mixed dipterocarp forest. Mt. Silam near
the town of Lahad Datu has extensive ultramafic vegetation and has been the subject
of more extensive research (Proctor et al. 1988, Proctor et al. 1989, Bruijnzeel et al.
1993). Although this mountain is only 884 m high, it ranges from lowland dipterocarp-
dominated forest to stunted Myrtaceae-dominated vegetation (Proctor & Nagy 1992).
Ultramafics on the islands in Darvel Bay are often composed of mono-specific stands of
Casuarina nobilis Whitmore (Fox & Hing 1971). The plant diversity of Mt. Kinabalu
390 Gard. Bull. Singapore 63(1 & 2) 2011
gap
has been the subject of extensive research (see Beaman & Beaman 1990; Beaman et
al. 2003), but the ecology of ultramafics has so far only been studied specifically by
Repin (1998).
Ultramafics around Mt. Kinabalu
Barthlott et al. (2007) lists the northern part of Sabah as one of the top five global
plant diversity centres, with more than 5,000 species per 10,000 square kilometres.
Extensive research at Mt. Kinabalu, has revealed that the plant diversity is as high as
5000-6000 species (excluding mosses and liverworts but including ferns), comprising
over 200 families and 1000 genera (Beaman et al. 2003). Given that this plant
diversity occurs in an area approximately only 1200 square kilometres, Mt. Kinabalu
probably represents the most floristically biodiverse place in the world (Beaman et
al. 2003). The high biodiversity at Mt. Kinabalu is mainly the result of the different
climatic zones, due to altitudinal differences, the large array of soils including
ultramafics (“geodiversity’), the distinctiveness of habitats and the proximity of older
mountain ranges (Mt. Kinabalu is geologically very young), which provide a ‘species
dispersion base’ (Beaman et al. 2003). At elevations of 1700 m and above, ultramafic
vegetation is characterised by the dominance of (1) 7ristaniopsis elliptica Stapf, (2)
Leptospermum javanica Sm.— Tristaniopsis elliptica Stapf, and (3) Leptospermum
recurvum Hook.f. - Dacrydium gibbsiae Stapf (Kitayama 1991). On lower elevations,
Gymnostoma sumatranum (Jungh. ex de Vriese) L.A.S.Johnson and Centhostoma
terminale L.A.S.Johnson are distinctive (Beaman & Beaman 1990). High on Mt.
Kinabalu (around 2900m) the vegetation on ultramafics is rather bare and consists
only of herbs such as Schoenus curvulus F. Muell., Euphrasia borneensis Stapf and
Machaerina falcata (Nees) Koyama. The graminoid vegetation types on the summit
of Mt. Tambuyukon and at the spur of Marai Parai are another characteristic of local
ultramafic vegetation types, with endemic species such as the pitcher plant Nepenthes
rajah Hook.f. and the herb Scaevola verticillata Leenh..
Mt. Kinabalu’s ultramafic vegetation is typified by endemics from the
Nepenthaceae family, which are confined to ultramafics such as the earlier mentioned
Nepenthes rajah Hook.f. , as well as N. burbidgeae Hook.f. ex Burb. and Nepenthes
macrovulgaris J.R.Turnbull & A.T.Middleton and a number of endemic orchids such as
Paphiopedilum rothschildianum (Rchb.f.) Stein, Paphiopedilum hookerae (Rchb.f. )
Stein var. volonteanum (Sander ex Rolfe) Kerch., Paphiopedilum dayanum (Lindl.)
Stein, Corvbas kinabaluensis Carr, Arachnis longisepala (J. J. Wood) Shim & A.
Lamb and Platanthera kinabaluensis Kraenzl. ex Rolfe in Gibbs.
In addition, a range of trees, shrubs and herbs are ultramafic endemics at Mt.
Kinabalu, including: Magnolia persuaveolens subsp. rigida Noot., Embelia cordata
Philipson, Syzygium dasyphyllum Merr. & L.M.Perry, Syzygium exiguifolium Merr. &
L.M.Perry, Syzygium myrtillus (Stapf) Merr. & L.M.Perry, Eriobotrya aff. bengalensis
(Roxb.) Hk.f., Hedvotis protrusa Stapf, Lasianthus membranaceus var. firmus Stapf,
Urophyllum subanurum Stapf, Wikstroemia indica (L.f.) C.A.Meyer and Elatostema
bulbothrix Stapf (Beaman & Anderson, 2004).
Only a few tree species are currently understood as true ultramafic endemics.
These include Borneodendron aenigmaticum (Meijer 1964), Dacrydium gibbsiae
Sabah’s ultramafic areas 39]
Stapf and Pittosporum silamense J.B.Sugau. Some other species, such as Buchanania
arborescens (Blume) Blume, and dipterocarps such as Dipterocarpus lowii Hook.f.,
Dipterocarpus ochraceus Meijer, Dipterocarpus geniculatus Vesque, Shorea
tenuiramulosa and Shorea kunstleri King are typical of ultramafics, but are not entirely
confined to this substrate (Ashton 1982: Proctor, 2003)
Threats to ultramafic vegetation in Sabah
Ultramafic vegetation types in Sabah are severely threatened by land clearing
activities, such as in the Meliau Range, Bidu-Bidu Hills, Mt. Silam and Mt. Tawai
and a range of smaller sites. Despite the status of a protected forest reserve of these
ultramafic localities, encroachment remains a problem, and many sites without legal
protection have disappeared in recent years, such as Morou Porou and the Lohan
Valley. Mt. Kinabalu and ultramafic sites within national park boundaries, such as Mt.
Tamboyukon are safeguarded, but land clearing has made the national park a virtual
‘island’. Although only a small minority of the geological substrates in Sabah are
ultramafic, these areas have a disproportionately high number of endemic and/or rare
plant species. Many plant species occur on only a few sites or even a single site and
habitat destruction, 1f it occurred, could therefore potentially result in extinction.
Conservation and research priorities
Only very limited research on the ecology of ultramafics in Sabah has been undertaken
to date. According to Baker & Brooks (1988), “... the ultramafic regions of Asia are
generally the least explored and poorly described of such regions in the world”. Due
to the lack of scientific information, it is difficult to formulate conservation strategies
and priorities, although given the history of, and potential risk for further, habitat
destruction, it is evident that many species endemic to ultramafics are threatened.
The need for conservation and research has been made strongly by Repin (1998) who
emphasised that scientific research and preservation of ultramafics in Sabah needs to
be given utmost priority, particularly in those areas without legal protection, before
there is any further destruction and loss.
ACKNOWLEDGEMENTS. The author wishes to thank Prof. David Mulligan and Dr.
Peter Erskine from the Centre for Mined Land Rehabilitation (CMLR) at the University of
Queensland for generous support and the opportunity to participate in the Flora Malesiana
Symposium in Singapore. Thanks also to Dr. Andrew Fletcher from the CMLR for critical
comments on the manuscript.
References
Aiba, S. & Kitayama, K. (2002) Effects of the 1997—98 El Nino drought on rain forests
of Mount Kinabalu, Borneo. J. Trop. Ecol. 18: 215-230.
392 Gard. Bull. Singapore 63(1 & 2) 2011
Ashton, P.S. (1982) Dipterocarpaceae. Flora Malesiana Ser. 1, 9(2): 237-600
Baillie, I., Evangelista, P. & Inciong, N. (2000) Differentiation of upland soils on the
Palawan ophiolitic complex, Philippines. Catena 39: 283-299.
Baker, A. (1981) Accumulators and excluders-strategies in the response of plants to
heavy metals. J. P/. Nutr. 3: 643-654.
Baker, A. & Brooks, R. (1988) Botanical exploration for minerals in the humid tropics.
J. Biogeogr. 15, No 1: 221-229.
Baker, A. & Smith, A. (2000) Metal hyperaccumulator plants: biological resources for
exploitation in the phytoextraction of metal contaminated soils, pp. 3—S. Proc.
InterCost Workshop for Bioremediation. Sorrento, Italy.
Baker, A. & Whiting, S. (2002) In Search of the Holy Grail: A Further Step in
Understanding Metal Hyperaccumulation? New Phytol. 155: 1-4.
Baker, A.J.M., Proctor, J.. Balgooy, M.M.J. van, & Reeves, R.D. (1992a)
Hyperaccumulation of nickel by the ultramafic flora of Palawan, Republic of the
Philippines. In: Proctor, J., Baker, A.J.M. & Reeves, R.D. (eds) The Vegetation of
Ultramafic (Serpentine) Soils. Andover, U.K.: Intercept Ltd.
Baker, A.J.M., Proctor, J., Balgooy, M.M.J. van & Reeves, R.D. (1992b)
Hyperaccumulation of nickel by the ultramafic flora of Palawan, Republic of the
Philippines. In: Proctor, J., Baker, A.J.M. & Reeves, R.D. (eds) The Vegetation of
Ultramafic (Serpentine) Soils. Andover, U.K.: Intercept Ltd.
Barthlott, W., Hostert, A., Kier, G., Ktiper, W., Kreft, H., Mutke, J., Rafiqpoor, M. &
Sommer, J. (2007) Geographic patterns of vascular plant diversity at continental
to global scales. Erdkunde 61: 305-315.
Beaman, J.H., Anderson, C. & Beaman, R.S. (2003) Biodiversity of Mount Kinabalu.
CBD Technical Series 8: 58.
Beaman, J.H. & Anderson, C. (2004) The Plants of Mount Kinabalu. 5: Dicotyledon
families Magnoliaceae to Winteraceae. Kota Kinabalu: Natural History
Publications (Borneo) Sdn. Bhd. & Kew: Royal Botanic Gardens, Kew.
Beaman, J.H. & Beaman, R.S. (1990) Diversity and distribution patterns in the flora
of Mount Kinabalu. In: Baas, P., Kalkman, K. & Geesink, R. (eds) The Plant
Diversity of Malesia. Dordrecht: Kluwer Academic Publishers.
Boyd, R., Wall, M. & Jaffre, T. (2009) Do tropical nickel hyperaccumulators mobilize
metals into epiphytes? A test using bryophytes from New Caledonia. Northeastern
Naturalist 16: 139-154.
Brady, K., Kruckeberg, A. & Bradshaw Jr, H. (2005) Evolutionary ecology of plant
adaptation to serpentine soils. Annu. Rev. Ecol. Evol. Syst. 36:243—66
Brooks, R.R. (1987) Serpentine and Its Vegetation: A Multidisciplinary Approach.
Dioscorides Press.
Bruijnzeel, L., Waterloo, M., Proctor, J., Kuiters, A. & Kotterink, B. (1993) Hydrological
observations in montane rain forests on Gunung Silam, Sabah, Malaysia with
special reference to the ‘Massenerhebung’ effect. J. Ecol. 81: 145-167.
Coleman, R.G. & Jole, C. (1992) Geological Origin of Serpentinites. The Vegetation
of Ultramafic (Serpentine) Soils. Andover: Intercept Limited.
Sabah’s ultramafic areas 393
Collenette, P. (1964) A short account of the geology and geological history of Mt.
Kinabalu. Proc. Roy. Soc. London, Ser. B, Biol. Sci. 161: 56—63.
Fox, J.E.D. & Hing, T.T. (1971) Soils and forest on an ultrabasic hill north east of
Ranau, Sabah. J. Trop. Geogr. 32: 38-48.
Hoffman, P., Baker, A.J.M., Proctor, J. & Madulid, D. (2003) Phyllanthus balgooyi
(Euphorbiaceae s.1.), a new nickel-hyperaccumulating species from Palawan and
Sabah. Blumea 48: 193-199.
Jenny, H. (1980) The Soil Resource: Origin and Behavior. New York: Springer-Verlag.
Kitayama, K. (1991) Vegetation of Mount Kinabalu Park, Sabah, Malaysia. Honolulu:
Environment and Policy Institute, East-West Center and Department of Botany,
University of Hawaii at Manoa.
Lasat, M.M. (2001) Phytoextraction of toxic metals: a review of biological mechanisms.
J. Environm. Qual. 31:109-120.
Meijer, W. (1964) Forest botany in North Borneo and its economic aspects. Econ. Bot.
18: 256-265.
Pollard, A., Powell, K., Harper, F., Andrew, J. & Smith, C. (2002) The genetic basis of
metal hyperaccumulation in plants. C.R.C. Crit. Rev. Pl. Sci. 21: 539-566.
Proctor, J. (2003) Vegetation and soil and plant chemistry on ultramafic rocks in the
tropical Far East. Perspect. Plant Ecol. 6: 105—124.
Proctor, J. & Nagy, L. (1992) Ultramafic rocks and their vegetation: an overview.
In: Baker, A., Proctor, J. & Reeves, R.D. (eds) The Vegetation of Ultramafic
(Serpentine) Soils: Proceedings of the First International Conference on
Serpentine Ecology. Andover, U.K.: Intercept Ltd.
Proctor, J., Lee, Y., Langley, A., Munro, W. & Nelson, T. (1988) Ecological studies on
Gunung Silam, a small ultrabasic mountain in Sabah, Malaysia. I. Environment,
forest structure and floristics. J. Ecol. 76: 320-340.
Proctor, J., Phillipps, C., Duff, G., Heaney, A. & Robertson, F. (1989) Ecological
studies on Gunung Silam, a small ultrabasic mountain in Sabah, Malaysia. II.
Some forest processes. J. Ecol. 77: 317-331.
Rajakaruna, N. & Baker, A.J.M. (2006) Serpentine: a model habitat for botanical
research in Sri Lanka. Ceylon J. Sci. 32: 1-19.
Reeves, R. (2003) Tropical hyperaccumulators of metals and their potential for
phytoextraction. P/. & Soil 249: 57-65.
Repin, R. (1998) Serpentine ecology in Sabah, Malaysia. Sabah Parks Nat. J. 1: 19-28.
Sabah Forestry Department (2005) Sabah Forestry Department [Online]. URL:http://
www.forest.sabah.gov.my/caims/Class%20I/A_FR1/bidu%20bidu.htm
[Accessed August 30 2010].
Whittaker, R.H. (1954) The ecology of serpentine soils. Ecology 35: 258-288.
‘
184 . ;
= _—
i} 7 -
¢ ¥ = Ye ‘
aay Ratt peering GAS ap
a ye toms 90, 6am ale
e 4] ; ;
>" feadiod’® rs | tee
evel) | hme ~
ia 3
om
aie, nein
7 ‘ ~
peyy = as cot
. .
4
re") Lew.
+ RAT
ang © er funn
hte get Da
(aed side munity
gts aT St eAeen
ee
hy nye i nl
Ve 7 raciac il yates
hte aia sh PEL
on | Aartiet Hg!
onan: pte
; | Cie eg vow» tht
| _ mi tie ia ia
.
om
\
ae na
ye
tie
ayy ia r
Gardens’ Bulletin Singapore 63(1 & 2): 395-407. 2011 395
Cultivation and conservation of
Lilium philippinense (Liliaceae),
the Philippine endemic Benguet Lily
Teodora D. Balangcod', Virginia C. Cuevas? and
Ashlyn Kim D. Balangcod*
‘Department of Biology,
College of Science, University of the Philippines Baguio,
2600 Baguio City, Philippines
balangcod@yahoo.com
"Institute of Biological Sciences, College of Arts and Sciences,
University of the Philippines, Los Bafios, Philippines
‘Department of Mathematics and Computer Science,
College of Science, University of the Philippines Baguio,
2600 Baguio City, Philippines
ABSTRACT. Lilium philippinense, endemic to the Cordillera Administrative Region (CCR) of
the Philippines, grows on steep mountain slopes of Benguet and the southwestern part of the
Mountain province. The flowers, fragrant and used as wedding decorations, occur from late
May to July. Recent observations indicate declining populations of this species, which is said
to be difficult to grow. Under greenhouse conditions, seed and bulb germination show only
27.63% and 16.67% success, respectively. The apparently acute sensitivity of this species to
environmental factors such as soil pH, light, humidity, air and soil temperature, and possibilities
for ex situ cultivation, are discussed.
Keywords. Benguet lily, conservation, Cordillera Administrative Region, cultivation, endemic,
Lilium, Philippines
Introduction
Lilium philippinense Baker (Liliaceae) is endemic to the Philippines (Elwes 1880)
and occurs in the southern part of the Cordillera Central Range (CCR). Its fragrant,
white, trumpet-shaped flowers sometimes have a reddish tinge at the corolla base, and
annual mass flowering events are spectacular. The species grows among grasses such
as Themeda triandra, Miscanthus sinensis and Imperata cylindrica. Each plant bears
one or two flowers per stem, rarely three or four. The fruit is a capsule with numerous
small seeds. At the end of the short flowering season, the floral parts wither and the
capsule matures and eventually releases its seeds. Following the end of the growing
season, only the subterranean bulbs remain.
This lily is known by local names such as kanyon or luplupak (Ilocano), us-
usdong (Mt. Province Kankanai), putputak (Benguet Kankanai), tuktukpao (Kayan,
396 Gard. Bull. Singapore 63(1 & 2) 2011
Tadian), swasoy (Ibalo1), and swvosoy (Ikalahan). Each local name illustrates a unique
characteristic of the plant. For instance, swyosoy means “flower of the mountain”;
us-usdong means “to bow” referring to the pendulous flowers; and kanyon means “a
bomb”, illustrating the somewhat explosive expulsion of seeds when mature fruits
dehisce. Descriptions of L. philippinense are provided by Chittenden (1956), Bailey
(1960), Steiner (1960) and Madulid (2001). Lilium philippinense is morphologically
similar to L. formosanum and L. longiflorum, which are endemic to Taiwan and Japan,
respectively, and which grow well in the CCR. Often, horticulturists regard these
three lilies as the same species and sometimes refer to them interchangeably. In the
Benguet and Mountain provinces, this lily is often a favorite adornment for wedding
ceremonies and special occasions because of its delicate white flowers and fragrance.
There is scanty research on L. philippinense. A monograph of the genus Lilium
is the only published account that gives some basic information of this species (Elwes
1880). In the Philippines, of two undergraduate reports on L. philippinense, one
. describes shoot and root initiation of bulb scales using low temperature stratification
and a rooting hormone. That study showed that root formation can be enhanced by
treating bulbs at 3°C for 60 days; and more and longer roots are initiated by soaking
the bulbs in Hormex, a rooting hormone, prior to planting (Alipio & Ladilad 2005).
In vitro propagation of L. philippinense was demonstrated by Ampaguey et al. (2002),
who showed that callus formation was enhanced with three different media used;
unfortunately, field evaluation of calli was not tried. There have been no previous
studies of seed germination in this species.
Recently, L. philippinense populations were reported to have declined
due to human activities such as over-collection (Madulid 2001) and destruction of
habitat including landuse conversion and road widening (Balangcod 2009). From a
conservation perspective, it is important to understand the reproductive behaviour of
a species. This includes seed and bulb germination in this case. As pointed out by
Schemste et al. (1994), detailed information on the different stages in the reproductive
cycle ofa species may contribute basic information helpful to conservation management
decisions. The aim of the present study was to determine the germination and survival
capacity of seeds and bulbs of this species, taking into account factors such as relative
humidity, light, air and soil temperature in the greenhouse, as well as elevation and
geographic location of the original material. Specifically, this study was conducted
to (1) evaluate the percentage germination of L. philippinense using seeds and bulbs;
(2) determine which of the two propagules, the seed or bulb, has a better performance
for ex-situ cultivation; and (3) compare the percentage germination of seeds and bulbs
from different L. philippinense populations.
Materials and methods
Seeds and bulbs of L. philippinense were collected from 28 populations from the
southern part of the CCR (Fig. 1). Elevation was determined using a Geographic
Positioning System (GPS). The collection sites ranged from 873 to 2091 m elevation,
Conservation of Lilium philippinense 397
121°0'0"E
18°0'0"N
Fig. 1. Location of the 28 Lilium philippinense collection sites.
extending from 16°28°44” to 17°10°45”N and from 120°63°55” to 121°83’23”E.
Collection of bulbs and seeds were made in August and September 2007 when mature
seeds were available. Experiments were performed to investigate the germination
capacity of seeds and bulbs and seedling survival of L. philippinense under greenhouse
conditions that simulated the natural environment of this species. The environmental
requirements of L. philippinense were noted in a separate study.
Seeds and bulbs were planted simultaneously after collection in standard
plastic pots (11.43 =< 10.16 cm) and maintained in a greenhouse. The germination
capacity of the two propagules was evaluated 150 days from sowing. Observation
was continued until the flowering phase to assess seedling survival over a 1-year
period for bulb-originated plants, and over a 3-year period for seed-originated plants.
Greenhouse parameters such as relative humidity, light, air and soil temperature were
recorded and monitored. The daily average readings of different environmental factors
were calculated and used for the monthly average within the first year.
The soil used in the experiment was collected from one of the population sites
to conform to the soil requirements of L. philippinense. Soil parameters such as soil
pH, organic matter content and phosphorus were analysed once before planting.
398 Gard. Bull. Singapore 63(1 & 2) 2011
Descriptive, Correlation, and Principal Component analyses using Statistical
Package for the Social Sciences (SPSS) were used to analyse the data. Graphs were
generated using Microsoft Excel.
Results
The greenhouse condition
The air and soil temperature measured in the greenhouse for the duration of the
experiment ranged from 22.57°C to 28.56°C and 19.07°C to 23.49°C, respectively.
The relative humidity showed a minimum of 61.61% and a maximum 80.72% while
light ranged from 581.53 footcandles (fc) to 1364.89 fe (Table 1). The Coefficients
of Variation indicate that relative humidity, air and soil temperature did not vary
significantly during the year, but that light varied slightly.
Relatively high air and soil temperatures were recorded from March to August
and falls slightly in the succeeding months (Fig. 2). Light peaked in June with a mean
value of 1363 fc. This corresponds with the time when L. philippinense starts to bloom
in its natural habitat. Relative humidity oscillated uniformly throughout the year.
Of the four parameters measured in the greenhouse, correlation analysis
shows that air temperature is highly and positively correlated with soil temperature (r
= 0.867, p = 0.000) and negatively correlated with relative humidity with a correlation
value of r = -0.657, p = 0.020 (Table 2). This suggests that there is positive relationship
between air and soil temperatures. On the other hand, there is a negative relationship
of air temperature and relative humidity.
The soil used in the experiment has a pH of 6.5, which is slightly acidic, a high
organic matter content of 4.68 per gram of soil and available phosphorus of 9.04 ppm.
The last is moderately low (PCARRD 1982). These quantities fall within the ranges of
parameters recorded in a previous study (Balangcod 2009). The ecological parameters
measured in the greenhouse were comparable with data obtained in the natural habitat
(Balangcod 2009).
Table 1. Environmental factors measured in the greenhouse, October 2007 to November 2008.
Relative Air Soil
ae? Light (foot-
Humidity temperature temperature savaies)
(%) (EC) Ge)
Minimum 61.61 YDS 19.07 581.53
Maximum 80.72 28.56 23.49 1364.89
Mean 70.50 Dil DADS 921.54
Standard Deviation 6.92 2.43 1.42 232.18
Coefficient of Variation 9.82 9.45 6.58 25.19
Conservation of Lilium philippinense 399
LIGHT (FC)
Nov-O7 Dec-O7 Jan-O8 Feb-O8 Mar-O8 Apr-O8 May-O8 Jun-OS JulO8 Aug-08 Sep-08 Oct-08
Monthly mean
RH (%)
Nov-07 Dec-07 Jan-08 Feb-O8 Mar-O08 AprO8 May-0O8 JunO8 JulO08 Aug-08 Sep-08 Oct-08
Monthly mean
AIR T (°C)
Degrees Centigrade
Nov-07 Oec-07 Jan-O8 Feb-08 Mar-O8 Apr-O8 May-08 Jun-O8 Jul-O8 Aug-08 Sep-08 Oct-08
Monthly mean
SOIL T (2C)
Degrees Centigrade
Nov-07 Dec-O7 Jan-O8 Feb-O8 Mar-O8 AprO8 May-0O8 Jun-08 Jul-O8 Aug-08 Sep-08 Oct-08
Monthly mean
Fig. 2. Monthly means of environmental factors (from top: light, relative humidity, air
temperature, soil temperature) measured in the greenhouse.
400 Gard. Bull. Singapore 63(1 & 2) 2011
Table 2. Correlation analysis of environmental factors measured in the greenhouse. RH =
Relative Humidity, temp = temperature.
RH Air temp Soil temp Light
Spearman’s Correlation
rho Re Coefficient 1-000
Sig. (2-tailed) =
N 12
Air Correlation 0.657% 1.000
temp Coefficient
Sig. (2-tailed) 0.020 _
N 12 12
eo Cua -0.538 0.867** 1.000
temp Coefficient
Sig. (2-tailed) 0.071 0.000 2
N 12 12 12
ici ee -0.434 0.434 0.420 1.000
g Coefficient , ec :
Sig. (2-tailed) 0.159 0.159 0.175 =
N 12 12 12 12
Germination from bulbs
Of the total 252 bulbs collected from the 28 population sites, only 16.67% germinated.
From the 16.67% that germinated, 57.14% seedlings developed till the flowering stage
(Table 3).
The number of bulbs collected from each population differed because of
availability. However, to solve the problem of sample size, the collection sites were
pooled by performing hierarchical cluster analysis in SPSS (Fig. 3). Consequently,
there were four clusters formed based on variables used, such as geographic location,
elevation, number of bulbs and seeds, and percentage germination of both propagules.
Characteristically, cluster | has an elevation range from 873 to 1103 m, cluster 2 has
an elevation range of 1232 to 1354 m, cluster 3 has an elevation range between 1481
and 1794 m and cluster 4 has an elevation range from 1959 to 2091 m.
To determine if there were differences of variation in the number of bulbs
collected and the germination capacity of the bulbs from different population sites
within and among the clusters, one way ANOVA showed that all the values obtained
were not significantly different, except for elevation (Table 4). This demonstrated that
the population sites and the number of bulbs have no significant differences in their
ability to germinate.
401
ilium phili
mservation of
Co
CO'LC
bOSC
LOSI
ff cv
8LL
9S'SS
009
00-0
£E'8S
OO9T
OO'OL
LYOI
0001
LO'IL
OLY
SL LI
ELEC
000
OO'C
8O'LI
LY'I9
Iv Lt
fe tv
fe Sl
BL LE
LOPS
bryce
tele
LS'8t
(sAup 0S |
JoyV) uo
“Buruniod
poas |
00 POO!
ts
9|
98
ol
Ot]
tc
vt
c8
vr
LV
[8
spoas
poywu
“TULIOL)
OO 096
OIC
OSI
OCI
06
06
OS]
Of
09
OSI
OL
OSI
09
OCI
OIC
OSI
O81
Ot
OS]
OPC
09
OLE
OOt
OS]
06
OST
O81
OSI
OLS
UMOS
spoos
JO'ON
L99|
oc vl
OLY
606
0009
tt et
000
69'L
000
000
SOLS
OO'SC
00°0
IL’St
00°09
00'0
OO'SC
OO'Ok
000
£t'8
LYYI
000
OI 0C
00:0
000
000
OO'OS
000
te et
(sAvp Qs |
Joye) von
“vurtuiod
Qing %
O00 Ch
OOC
OO |
OO |
OO't
OO |
00:0
OO |
000
00:0
OO’
OO0't
000
OOS
009
000
00 |
OO'¢
000
001
00]
00-0
009
000
000
00-0
OO'C
000
OOC
(4+)
powuruuod
sq[ng [BIOL
OOP
00'|
00:0
O00]
O0'C
00 |
000
00 |
000
00-0
OO't
OO’
00-0
OOP
OO’
00-0
OO]
00 |
000
00 |
OO |
00-0
000
000
000
000
00 |
000
00:0
(q]) OMOTy
O} SUIATAINS
Jou sqynq
powwurunar,
OO'8I
OO |
00 |
000
00°]
000
000
000
000
000
000
00-0
000
OO |
O0'¢
000
000
OOC
000
000
000
000
009
000
000
000
00 |
000
OO'T
(VY) Jomo]
0) SUIATAINS
sqynq
poyeuruior
00'TST
00'V I
OO'1
OO'TI
00'S
O0'£
000!
O0'ET
007
007
OO'EI
O0TI
00'|
00'FI
0001
00'|
O00'P
00°01
009
O00'TI
00'9
OO
00°67
O0OL
00°6
O0'T
00'P
OO'8I
00'9
pojueyd
SqIM el
[PIO],
pStl
6608
9SSI
S89
8CcLI
voll
1ocl
ctl
LOL
990 |
ctl
SSO]
tStl
116
1S6
COO
OLTI
Ictl
tL8
PIS
[8h
COCCI
PLOC
8S0C
Lo00C
COLI
OSOl
Chol
UONRA|A
LIPS OCI
LO86 OCI
OL16 01
€LL8 OCI
OS88 01
8768 071
87L9'0C1
97S9'0TI
O888 OT
SSt9 NTI
O6L8 OCI
O98 OCI
obb9 071
Peso OCI
ILt6 071
OSPO OCI
LSIOITI
698 OT
9O18 OCI
ISS8 OCI
PITS OCI
£1L9' 071
7760 ITI
C89L OCI
OP9L OCI
cLIL OCI
O8SL OCI
CCOL OCI
opnyisuo7]
bL8o 91
OF8O LI
SPOT LI
OSOO'LI
[PoO'L1
COCO 91
ceel Ol
S699'9|
OLSE OI
O8b9 9
OPl9'9I
CILS 91
bPsc Ol
17 a
6OSPO LI
LLLO'LI
SSIOLI
O90F 91
SSLP9l
cet 91
Tess 9]
8cLO 9I
LLCO LI
9LPL 9
41099]
Leys 91
£689 91
8CLP OI
apne]
UPIPRL,
‘py uvg-uog OgE Uy ‘Pyowes
oduopey ‘epedes
WwouRC, Joye ‘epedes
wourqd ‘epedes
urdurqes
Sunsdey ‘uedunqry
weddeg ‘qasoy ‘uedunqry
VAROZIA BAO ‘vdvdAry
jodunury ‘uvdurdey
uvoenq ‘uvAeqey
Aoyseg, “duayey ‘uvdrqey
ovondury ‘uodoy|
‘py Uvg-UOg gRE UY ‘O}UOg
uodouUoL ‘NOVO K]
Avjueq “oo uog
OAARE] SOOO
Old ‘Poyog
duo woypy ‘poyog
[[IUIMBS “Yoqog “poyxog
Bill Pog
sseq suvssog
INE] IOMO'T ‘ONE
unsurjseg ‘unyeg,
‘py Buosyey] “Yory
yepdoy ‘yory
duvqesuryeg ‘Yory
ubIpluy ‘oRpynqury
song uonepndog
‘SOS UOII]JOO JUdIAALJIP OY] WO’ Spoos puke sqyng aesuvauiddiyiyd wniptT JO VONBUTUIAS OdLUOII “E OIQEL
Gard. Bull. Singapore 63(1 & 2) 2011
Table 4. One way analysis of variance (ANOVA) of the different variables.
Sum of
df Mean Square
F
Squares ficance
latitude Between Groups 047 3 016 1911 901
Within Groups 1.956 24 082
Total 2.003 Pa
longitude Between Groups 037 3 0.012 .886 .462
Within Groups 336 24 0.014
Total se: PN
elevation Between Groups Sie) 1 T/ 3) 1245192.337 156.681 .000
Within Groups 1907 Ss5at 24 7947.296
Total S9263i2 27
number of Between Groups 152.730 3 50.910 e233 320
bulbs Within Groups 991.270 24 41.303
Total 1144.000 Di]
bulb Between Groups 1502.604 3 500.868 1.491 242
germination Within Groups 8063.130 24 335.964
Total 95651 34a ee),
number of Between Groups —-3212.500 3 1070.833 9S 900
seeds Within Groups 133330.4 24 5555.432
Total 136542.9 27
seed Between Groups 564.793 3 188.264 366 178
germination Within Groups 12353.843 24 514.743
Total
12918.636 27
Germination from seeds
Seed germination was observed and recorded after 150 days from sowing. Of the total
3960 seeds collected from the 28 population sites that were sown under greenhouse
conditions, total seed germination was 27.63%. Of the seeds collected from the 28
population sites, the highest percentage seed germination was observed from Ampucao,
Itogon with 71.67% germination. This is followed by seeds collected from Amunget,
Kapangan, which showed a germination of 70%. In contrast, seeds collected from
Bayyo, Bontoc and Napsung, Kibungan did not show signs of germination even after
150 days. Comparatively, the percentage seed germination in all the 28 populations
showed variable percentages, from a minimum of 0% to a maximum germination
of 71.67% (Table 3, Fig. 4). Nevertheless, despite the variable seed germination
percentages, one way ANOVA showed that the number of seeds and the capacity to
geminate did not significantly differ in the different population sites (Table 4).
Conservation of Lilium philippinense 403
50 0
fh
\
Se eeooononoge
2? INy o o oO Cy
0
¢
¢
¢
¢
¢
¢
¢
¢
¢
¢
¢
q
¢
¢
¢
€
¢
¢
¢
¢
¢
¢
¢
¢
¢
¢
¢
¢
¢
¢
¢
¢
©
¢
©
\
\
o 6
3
o o ¢
¢
+7
¢
¢
NS
t
emo 6
OO) im Wn
rs)
eX
g
tt
Bokod, Bi
By
%
QO ee Cees GS ee SS 6a
t
hm 9 0 Ow
im W W ® WO O Nh) if
el) Gal ied
Fig. 3. Dendrogram of the population sites using hierarchical cluster analysis.
Generally, seed germination had a higher value of 27.63% relative to bulbs,
which only showed a germination percentage of 16.67%. Across the 28 populations,
seed germination was higher than bulb germination with few exceptions (Fig. 4).
Additionally, in terms of seedling survival, despite the higher germination capacity
from seeds, the plants remained as seedlings up to the third year. These seedlings did
not reach maturity, unlike the sprouts from bulbs where 57.14% of the 16.67% that
germinated survived until the flowering and fruiting stage.
To determine if there is a relationship among the percentage germination of
seeds and bulbs across the 28 populations with respect to elevation and geographic
origin, principal component analysis (PCA) was performed. Results showed that the
28 population sites had high loadings on the principal axis (Factor 1) with an eigen
value of 99.89%. Furthermore, one way ANOVA shows that there were no significant
differences among the number of seeds, percentage germination and collection
sites. This demonstrates that elevation and geographic location of L. philippinense
populations did not show significant variation in terms of their germinating capacity,
both for the seeds and the bulbs.
Survival of plants from bulb and seed origin
The survival of plants originating from bulbs was observed over one year. Results
showed that 57.14% of the 16.67% plants that germinated from bulbs were able to
reach flowering stage. Almost half the seedlings from bulbs did not survive up to the
404 Gard. Bull. Singapore 63(1 & 2) 2011
80 B %Bulb Germination
@ %Seed Germination
Percentage Germination
F
iat
El
123 45 6 7 8 9 1041121314 15:16 I7 18 19°20) 21 22 23) 24)25 267026
Collection Sites
wo
lo)
——y
CFSE TE EY) EET IT ERD SEITE TSOP
oO
=
——
=
Fig. 4. Germination performance of Lilium philippinense bulbs and seeds from different
collection sites. Collection sites: 1 Ambuklao, Inidian; 2 Atok, Balangabang; 3 Atok, Topdak;
4 Atok, Halsema Rd.; 5 Bakun, Bagtangan; 6 Bauko, Lower Buga; 7 Bessang Pass; 8 Bokod,
Bila; 9 Bokod, Bobok, Sawmill; 10 Bokod, Moatong; 11 Bokod, Pito; 12 Bontoc, Bayyo;
13 Bontoc, Dantay; 14 Bontoc, Gonogon; 15 Bontoc, Km 388 Bon-Ban Rd.; 16 Itogon,
Ampucao; 17 Kabayan, Caleng, Bashoy; 18 Kabayan, Duacan; 19 Kapangan, Amunget; 20
Kayapa, Nueva Vizcaya; 21 Kibungan, Leseb, Sagpat; 22 Kibungan, Napsung; 23 Sabangan;
24 Sagada, Danom; 25 Sagada, after Danom; 26 Sagada, Madongo; 27 Samoki , Km 380 Bon-
Ban Rd.; 28 Tadian.
reproductive stage. This could perhaps be attributed to the presence of white aphids on
the bulb-originated material during the observation period.
The plants that germinated from seeds demonstrated a peculiar characteristic.
During the first year, these seedlings remained in their 2—4-leaf stage and wilted
after eight months, leaving tiny bulbs that became dormant in the soil. In the second
year, new sprouts developed from these tiny bulbs from the first year but these also
remained vegetative, after which they again wilted without reaching reproductive
stage, surviving once more as dormant bulbs in the soil. In the third year, sprouts
again grew from 2-year-old bulbs, but also remained in a vegetative stage. Even at
this particular stage, the bulbs had not reached their mature size (about 30-40 mm
diameter).
Discussion
One of the most basic ways of propagating plants is through seed. For plants that
develop both seed and bulbs such as lilies, seed germination has some advantages.
First, it allows growers to propagate lilies that are difficult to obtain as bulbs. Second,
seeds are usually disease and virus-free, even if they come from infected plants. Third,
using seeds for propagation allows genetic variation in the succeeding population.
Conservation of Lilium philippinense 405
Germination from seeds is influenced by different factors such as dormancy,
seed size, exposure to environmental factors and other factors that are inherent in the
plant. Lily seeds have different types of germination depending on the species. Lilium
philippinense exhibits an epigeal type of germination. In this study, seeds and bulbs
of L. philippinense were exposed to a uniform set of environmental conditions in a
greenhouse. There was low germination percentage for both seeds and bulbs. Lilies
display a unique characteristic in terms of germination capacity. According to Elwes
(1880), this is inherent to temperate species where germination from seeds in Lilium
would show alternating dormancy and seedling initiation for three or more years. This
period allows the bulblets to attain a functional size before it can finally grow into a
reproductively active plant. Inherent dormancy was also demonstrated by Silvertown
in 1999. He observed that Liliaceae species exhibit double seed dormancy. Two cold
seasons are required for the seeds to fully germinate; the first cold stratification releases
the radicle and the second releases the shoot so that the seeds require two years for
germination. Related to this, some plant species have inherently low germination
capacity even when exposed to suitable environmental conditions for germination.
This observation was described by Lanta et al. (2003) for Amaranthus cruentus and A.
retrofilexus, which exhibit poor seed germination.
Studies of some species have shown that seed size often have significant
effects on final germination percentage, seedling survival or seedling growth (Gross
1984, Navarro & Guitian 2003). Harper (1977) proposes that the poor performance
of smaller seeds is due to their lower endosperm content. Seed germination in L.
philippinense is possible to an extent of 27.63%; however, reproductive plants cannot
directly grow from seeds. The plants that germinated from seeds wilted before reaching
maturity but remained as tiny bulbs in the soil. This was also demonstrated in other
studies (Schaal 1980, Dolan 1984, Marshall 1987, Naylor 1993). The size of the bulbs
is an essential consideration owing to its capacity to store enough food materials to
supply the seedlings the needed food during its development. Accordingly, the size of
the bulb for commercial production of some lilies should range from 20 to 14 cm. In
addition, plants grown from bulbs that are below 10 cm bears small and fewer flowers
(Hermano 2000). In Benguet lilies, the size of the bulbs should reach an average of 30
to 40 mm.
In this study, bulb propagation in L. philippinense showed 16.67 % germination.
This low germination percentage seems rather normal for some temperate species,
even if the propagules were reared under optimal flowering conditions (Elwes 1880).
In the present case, out of this percentage, only 57.14% developed until the flowering
stage. Almost half of the germinated seedlings from bulbs died before reaching the
reproductive stage, but it is not possible to know if the observed presence of aphids
was a cause.
One important factor that determines the distribution and survival of a species
is its ability to exist in harsh environments. In the case of L. philippinense, the 28
populations are part of a narrow distribution in the southern part of the CCR. This
suggests that there is a suitable but restricted environmental condition for this species
in this part of the CCR. The observation of horticulturists in the region that this species
406 Gard. Bull. Singapore 63(1 & 2) 2011
is difficult to cultivate outside of its natural habitat can be attributed to its specific and
limited range of environmental requirements.
Conclusions
The fact that germination from seeds and bulbs can be obtained despite its limited
extent demonstrates that it is possible to propagate L. philippinense using bulbs and
seeds outside of their natural environment, provided that optimal environmental
conditions will be met.
However, in conservation terms, germination from bulbs and seeds are just
two techniques for propagating material. Other faster means of propagation, like
tissue culture, have yet to be explored for L. philippinense. With declining populations
of this species, it is essential and interesting that proper and efficient ways of rapid
- multiplication of material should be considered in future studies.
ACKNOWLEDGEMENTS. The authors are grateful to the International Tropical Timber
Organization (ITTO), the Commission on Higher Education (CHED), Idea Wild and the
University of the Philippines (UP) for the financial assistance. We are also grateful to the Soils
Department of Benguet State University for the use of their laboratory for our soil analyses. We
thank Julie, Ditas, Kryssa, Bino, Ben, and our driver, June, for invaluable assistance extended
during the field collection and experimental phase of the study.
References
Alipio, L.M & Ladilad, A.G. (2005) Shoot and root initiation of Benguet lily (Lilium
philippinense) bulb scales by low temperature stratification and kind of rooting
hormone. Benguet State Univ. Res. J. 45 & 46: 41-58.
Ampaguey, D.W., Cadelifia, E.P., Dimas, M.D., Mang-Oy, A.B. & Palaez, J.A. (2002)
In-vitro propagation of Benguet Lily (Lilium philippinense). Undergraduate
thesis. Benguet State University.
Bailey, L.H. (1960) Manual of Cultivated Plants. Revised Edition. New York:
Macmillan Co.
Balangcod, T.D. (2009) Autecclogy of Lilium philippinense Baker (Liliaceae), An
Endemic Species in the Cordillera Central Range. Ph.D. Dissertation, University
of the Philippines Los Banos.
Chittenden, F.J. (1956) Royal Horticultural Society — Dictionary of Gardening 4 vols.
Oxford: Clarendon Press.
Dolan, R.W. (1984) The effect of seed size and maternal source on individual size in
a population of Ludwigia lectocarpa (Onagraceae). Amer. J. Bot. 71: 1302-1307.
Elwes, J.H. (1880) A Monograph of the Genus Lilium. London: Taylor and Francis.
Conservation of Lilium philippinense 407
Gross, K.L. (1984) Effects of seed size and growth form on seedling establishment of
six monocarpic perennial plants. J. Ecol. 72: 369-387.
Harper, J.L. (1977) Population Biology of Plants. London: Academic Press.
Hermano, F.G.Sr. (2000) Lily Production: A Commercial Production Technoguide for
Highland Philippines. Unpublished.
Lanta, V., Havranek, P. & Ondrej, V. (2003) Morphometry analysis and seed germination
of Amaranthus cruentus, A. retroflexus and their hybrid (A. = turicensis). P/. Soil
Environm. 49(8): 364-369.
Madulid, D. (2001) A Pictorial Cyclopedia of Philippine Ornamental Plants, 2nd ed.
Philippines: Boolmark.
Marshall, D.L. (1987) Effects of seed size on seedling success in three species of
Sesbania (Fabaceae). Amer. J. Bot. 73: 457-464.
Navarro, L. & Guitian, J. (2003) Seed germination and seed survival of two threatened
endemic species of Northwestern Iberian Peninsula. Biol. Conservation 109:
313-320.
Naylor, R.E.L. (1993) The effect of parent plant nutrition on seed sizeability and
vigour, and on germination of wheat and triticale at different temperatures. Ann.
Appl. Biol. 123: 379-390.
PCARRD, Philippine Council for Agriculture, Forestry and Natural Resources
Research and Development (1982) Standard range of some chemical and physical
properties of soils.
Schaal, B.A. (1980) Reproductive capacity and seed size in Lupinus texensis. Amer. J.
Bot. 67: 703-709.
Schemeste, D.W., Husband, B.C., Ruckelshaus, M.H., Goodwillie, C., Parker, I.M.
& Bishop, J.G. (1994) Evaluating approaches to the conservation of rare and
endangered plants. Ecology 75: 584-606.
Silvertown, J. (1999) Seed ecology, dormancy, and germination: a modern synthesis
from Baskin and Baskin. Amer. J. Bot. 86(6): 903-905.
Steiner, M.L. (1960) Philippine Ornamental Plants. Manila: M & L. Licudine
Enterprises.
Gardens’ Bulletin Singapore 63(1 & 2) 409-423. 2011 409
Predicting distribution of Lilium philippinense (Liliaceae)
over Luzon’s Cordillera Central Range, Philippines,
using ArcGIS Geostatistical Analyst
Ashlyn Kim D. Balangcod
Department of Mathematics and Computer Science,
College of Science, University of the Philippines Baguio,
2600 Baguio City, Philippines
ashlynbalangcod@yahoo.com
ABSTRACT. The growing importance of geographical information system (GIS)-based output
in the analysis of biodiversity data is due to its convenient method of spatial analysis of data
and prediction of plant geographical distribution. The Interpolated Distance Weight Method
(IDW) of ArcMap ArcGIS 9 was used to determine possible areas of Lilium philippinense,
endemic to the Cordillera Central Range (CCR) and declining in population due to habitat
destruction, swidden activities and over-collecting. The variables considered in this study are
soil pH, soil phosphorus content, organic matter, elevation, latitude and longitude. All variables
were studied from actual L. philippinense sites and, using IDW, prediction maps were generated
that identified areas where L. philippinense are likely to thrive. The Geostatistical Analyst of
ArcGIS is a useful tool for predicting potential sites for introduction of L. philippinense as an
extended in-situ conservation strategy.
Keywords. ArcGIS, Cordillera Central Range, distribution prediction, Lilium, Luzon,
Philippines, potential geographic distribution
Introduction
The study of plant species distribution is an important aspect of biodiversity science.
Currently, modeling distribution is less taxing with the use of computer applications or
software. If modeled visually through maps, data taken from the field can be interpreted
and analysed more accurately. Geographic Information System (GIS) software,
specifically, ArcGIS 9, provides storage of quantitative data for generating visual
representation on a geographic reference, and retrieval and analysis of information
(Fischer 2009). According to Main et al. (2004), GIS help manage, analyse, and
present spatially related information combining multiple layers of environmental
and biological information related to a spatial location, to gain a better understanding
of a specific location (Main et al. 2004). Additionally, researchers can use GIS to
fully investigate data and develop spatially accurate graphical data displays. This
is very important, especially in geographic distribution where a more accurate and
graphical display of populations can be presented. This graphical display can help
in decision-making, such as in conservation. This paper focuses on the use of the
410 Gard. Bull. Singapore 63(1 & 2) 2011
Interpolated Distance Weight Method (IDW) in the interpolation of known data of
Lilium philippinense Baker (Liliaceae) population sites to predict potential areas for
cultivation.
In ArcMap, a data layer can be created for each variable of the observed
sites. ArcMap is a map-centric application that supports editing and viewing of maps
(Longley et al. 2005) and used in visualising different types of data providing an
interactive interface for data manipulation, information retrieval and spatial analysis
(Zeiders 2002). A data layer is generated from a database of the values of the variables,
with the longitude and latitude of the sites. Each data layer can be used by the IDW
method of ArcMap GIS to produce the prediction maps. The prediction map is another
layer which predicts the values of the predicted sites and compares it with the other
areas of unknown data.
The distribution of L. philippinense populations was addressed in this study
because, first, it is an endemic species in the southwestern part of the Cordillera Central
- Range (CCR); second, its populations are already declining due to anthropogenic
activities; third, few studies have documented this endemic species; and fourth,
a dataset consisting of variables such as latitude, longitude, elevation, and soil
parameters such as soil pH, phosphorus and organic matter content has been gathered
by Balangcod (2009) and allowed to be used in this study. These variables were used to
create data layers using ArcMap GIS 9. Specifically, the dataset was used to predict the
potential sites of distribution of L. philippinense in the CCR using the IDW method of
ArcMap GIS 9. With the population of L. philippinense dwindling, the use of a dataset
in extrapolating and predicting potential sites for the introduction of the species in an
extended in-situ conservation programme ts helpful.
In using the GIS software, a fair knowledge of computers is a necessity as
some parts of the software call for critical analysis, an element taught in computer
science.
The features and uses of ArcGIS
GIS is defined by Paul Longley et al (2005) as a computerised tool for solving
geographic problems, a mechanised inventory of geographically distributed features
and facilities, and a tool for performing operations on geographic data that are too
tedious or expensive or inaccurate 1f performed by hand. Childs (2004) describes GIS
as all about spatial data and the tools for managing, compiling and analysing that data.
GIS has numerous uses such as census, mapping, modelling and prediction.
ArcGIS is a leading software in the GIS market due to its extensive features and
global community of users (Information Management Editorial Staff 2004). Among its
features is the interpolation tool in the Spatial Analyst extension. Interpolation is a
process used to predict the values of cells at locations that have no information. Using
the principle of spatial autocorrelation which measures the degree of dependence
between near and distant objects, interpolation determines interrelation of values to
also determine the spatial pattern (Childs 2004). Another principle which is the basis
of spatial interpolation is Tobler’s Law, which states that “all places are related but
nearby places are more related than distant places” (Miller 2004). This means that
Lilium philippinense distribution studies with GIS 41]
the best guess for a point with no information is the value measured at the nearest
observation points (Longley et al. 2005).
Another interesting feature of ArcGIS is the use of interpolation methods
such as IDW, Kriging and Spline. The IDW method is a deterministic interpolation
technique. Deterministic interpolation uses mathematical formulas or measured
points to create surfaces (Childs 2004). Kriging is a popular statistical method based
on regionalised variables (Longley et al. 2005) and Spline is a method that uses a
mathematical function that minimises overall surface curvature (Childs 2004). IDW
is the method most often used by spatial analysts due to its simplicity. It estimates
unknown data by getting the average of known measurements of nearby points, with
the nearest neighbours getting greater weight in the computation of averages. The
formula used by IDW 1s as follows:
2(x) = »y w.z,/ yy W,
where x is the point of interest, the unknown value is denoted by z(x) and the known
measurements as z,. The weights (w,) are defined most often by the inverse square of
distances formula:
5
w. = 1/ do
I I
where d_ is the distance from x to x,, with x, as the points where measurements were
taken. The data points run from | to 7. (Longley et al. 2005).
Almost all information that requires mapping or modelling over the Earth’s
surface can now be effectively stored and retrieved using Information Systems.
Specifically, Geographic Information Systems or GIS are used for these tasks. In recent
years, modelling potential species distributions using GIS has become popular because
it is a powerful tool for researchers involved in vegetation mapping, biodiversity
mapping and population distributions (Moreno et al. 2007, Murray 2009, Hasmadi
2010). GIS is also a convenient tool in creating prediction maps (Pallaris 1998, Sergio
& Draper 2002, Vargas et al. 2004, Vogiatzakis & Griffiths 2006). The last-mentioned
study modeled the potential distribution of 36 endemic and 47 non-endemic species of
Anthurium (Araceae) in Ecuador based on mean annual temperature and humidity. GIS
was also used to identify and analyse the environmental tolerance limits of Cecropia
(Pallaris 1998).
Methodology
Study area
The Cordillera Central Range (CCR) is located in the northern part of the Philippines.
It is a mountainous region, with an estimated total area of 17,500 km (CPA Phil. 2006).
It has six provinces, viz., Apayao, Abra, Mt. Province, Ifugao, Kalinga and Benguet.
The CCR has a diverse flora and fauna, some of which are endemic to the area. Lilium
philippinense, a species described in 1880, is endemic there (Elwes 1880).
412 Gard. Bull. Singapore 63(1 & 2) 2011
Lilium philippinense 1s one of three species of Lilium L. found in the CCR
(Palima 1988). It is a bulb species with a strikingly white trumpet-like flower. The
flower has an aromatic fragrance and is produced singly per stem. Each plant bears
1-2, rarely to 4, stems. This species flowers only once per year and is visible during
the rainy season from late May to August (Balangcod 2009).
Between 2007 and 2009, a study indicated 118 population sites of L.
philippinense. These sites were georeferenced using a GPS receiver, Garmin’s GPSmap
60C, where latitude, longitude and elevation were recorded. Out of the 118 sites, soil
samples from 45 sites were collected for determining the soil pH, phosphorus and
organic matter content. These additional variables together with the longitude, latitude
and elevation were used with the IDW of ArcMap Gis 9 to predict potential sites where
L. philippinense would likely grow.
Two sets of data were used. The first set comprises data on latitude, longitude
and elevation taken from 118 sites, and the second data set comprises data on latitude,
_ longitude, elevation, soil pH, phosphorous and organic content taken from 45 sites that
were a subset of the 118 sites mentioned. All data gathered from the fieldwork was
initially saved in excel files (.xls) but were converted to database files (.dbf) since it is
the format needed by ArcMap to plot the data on the map.
Using the four characteristics: elevation, soil pH, phosphorous and organic
matter, four colour-filled contour prediction maps were produced. These four maps
were overlain on the CCR map to determine the areas where L. philippinense Baker
is predicted.
Preprocessing
A base map for plotting the data was needed for visualisation in ArcMap. A vector
and raster dataset of the Philippine map with provincial boundaries was taken from
PhilGIS, a website that provides free Philippine spatial data. Separate maps for each
province with municipality boundaries were also obtained from the same website. The
Philippine maps in shapefile format (.shp) were set in ArcMap using its default datum
World Geodetic System (WGS) of 1984 as its coordinate system which is also the
datum used in all layers of data in ArcMap.
Use of GIS in predicting sites for Lilium philippinense
Modelling the distribution of different sites across the CCR employed the two data
sets mentioned. Prediction maps were generated for each factor: elevation, soil pH,
phosphorous and organic matter content of the soil. The Spatial Analyst IDW was
used to generate prediction maps with settings of the default power of two. This power
value controls the significance of known points on the interpolated values, based on
the distance of the known and the output points. Higher values of power may cause
“non-smoothness” of values (Longley et al. 2005). The number of nearest neighbours
was set to 15, meaning computation of the unknown data depends on the nearest 15
known data.
Lilium philippinense distribution studies with GIS 413
Using IDW interpolator, the ArcMap calculates the value of each cell of the
map depending on the weight or attributes of its neighbouring data. Given that the
observed data were randomly distributed in the CCR, the Variable option was used
under the search option (Longley et al. 2005).
Verification
To verify the accuracy of the prediction, specifically the prediction maps for the
elevation, a Digital Elevation Map (DEM) was downloaded and was used to cross-
check the predicted values against the true values. The presence of the DEM allows
comparison of the predicted and the actual values.
For the prediction map for elevation using the 45 study sites, 60 sample areas
from the predicted sites were chosen randomly for verification. Ten areas were chosen
from each province (Fig. 1). Variables such as latitude, longitude, and elevation for
these areas were extracted from the DEM and then compared with the predicted
elevation.
To verify further, the GPS readings of the elevation of 33 population sites that
were included in the 118 study sites but excluded from the subset of 45 study sites
were also noted. These 33 sites or points were found within the coloured contours of
the prediction map. Note that the 118 sites and the subset of 45 sites have elevation
data gathered from the field survey using a GPS handset.
As for the prediction map for elevation using the 118 study sites, the same
60 sampling areas were used and the actual and predicted elevations noted. The
percentage errors for the predictions using only a subset of 45 study sites and using the
full complement of 118 study sites were computed and compared.
Results
Spatial analyst —IDW method
There were four data layers used in this study. The first type comprised the shape
files for the six provinces of the CCR Region; Abra, Apayao, Benguet, Kalinga, Mt.
Province and Ifugao which have data until the municipal level. The second type was a
Philippine Digital Elevation Map (DEM) which had the actual elevation for the entire
country. The third data layer was made up of the corresponding latitude and longitude
of the study sites. Finally, the prediction maps generated by the Spatial Analyst of
ArcMap constituted the fourth data layer.
The first data layer was taken from PhilGIS, a website that provides free
shape files for the Philippines. There are six shape files, all with boundaries until the
municipal level. The shape files have information including province, municipal or
town and barangay. The second data layer was taken from a GTOPO30 website. This
layer was a raster component which stored the elevation value for each latitude and
longitude coordinate of the Philippines. The third data layer comprised the two datasets
414 Gard. Bull. Singapore 63(1 & 2) 2011
Legend
a
Cieca
Elevation Prediction Map Pog rrovyce eee
from 45 Study Sites 2: Q
a
@ 45 Study Sites G —
© 118 Study Sites ;
= 60 Sample Areas
0 15,000 30,000 60,000 90,000 120,000
Meters
Fig. 1. Sixty sampling areas were randomly chosen to verify the predicted values of elevation
generated from the prediction maps that used data from just 45 sites, and all 118 sites. The 45
sites mentioned are a subset of the 118 sites.
taken from the 118 study sites and the subset of 45 study sites. Each dataset had its own
separate layer. These layers were superimposed on the others. These study site layers
were derived from the table initially saved in Excel format (.xls) that was later changed
into a database file (.dbf). The data needed for plotting were latitude, longitude and
elevation. The fourth layer, consisting of the prediction maps, was generated by the
ArcMap IDW Interpolator. There were five prediction maps produced. Two prediction
maps using the elevation were created from the subset of 45 sites and from the full
complement of 118 sites (Fig. 2 and 3). The other three factors, soil pH, phosphorous
content of soil and organic matter were the input used for the other three prediction
maps (Fig. 4 to 6).
Lilium philippinense distribution studies with GIS 415
18°0'0"N
N
\ a
S
Legend
Elevation Prediction Map (45 sites)
Filled Contours
121°0'0"E
Fig. 2. Prediction Map using elevation of 45 observed sites of Lilium philippinense.
However, the ArcMap Spatial Analyst IDW method of prediction uses only
values within the range of known data in predicting the values for areas that have
unknown data. For example, if the known data are 0, 5, 7 and 10, then unknown data
will only have a value within the range of 0 to 10. In this case, the predicted elevation
of the unknown areas would only have a predicted value within 754 to 2058 m, when
data from only 45 sites were used, and a range of 754 to 2155 m when data from all
118 sites were used. The same applies to prediction of the other factors, namely, soil
pH, phosphorous content and organic matter.
To delimit the area of prediction, the range of values where the most number
of sites were found was determined. For example, the elevation ranges of 754 to 2058
and 754 to 2155 were divided into ten classes. ArcMap automatically divides the
ranges with the specification of the number of classes as input. The top five ranges or
50% of the ranges with the highest number of sites were noted for each data set and
416
Elevation Prediction Map (118 sites
Filled Contours
754 -872 6541
272
E278 0126
aaa
1071 565
1176 924 -
1295578 -
Ce 1429.205-
1579 696 -
1749 177 -
1940 045 -
- 978 0126
- 1071.565
- 1176924
1295 578
1429 .205
1579 696
1749.177
1940045
2155
e 118 Study Sites
Gard. Bull. Singapore 63(1 & 2) 2011
90,000 120,000
Meters
Fig. 3. Prediction Map using elevation of 118 observed sites of Lilium philippinense.
were the only data visible in the prediction maps. Table | contains the values for the
ranges and the corresponding number of sites for each range. Separate prediction maps
were created for both sets.
For data on soil pH, phosphorous content and organic matter, their ranges
were also divided into ten classes each, and the top five ranges where the most number
of sites were found were noted and mapped.
After creating the prediction maps, merging of all maps using all factors
(elevation, soil pH, phosphorous content and organic matter) was done and the areas
where Lilium philippinense was predicted to thrive was determined. This is inferred
from the overlain map. The Identify Tool of ArcMap enables a point-and-click query
of the location using the shape files of the six provinces.
hn eden
- we. »—
Lilium philippinense distribution studies with GIS 417
0 15,000 30,000 60,000 90.000 120,000
Meters
Fig. 4. Prediction Map using soil pH generated from 45 observed sites of Lilium philippinense.
There were two output maps from the merging of all prediction maps. The
first used the three prediction maps of soil pH, phosphorous content of soil and organic
matter with the elevation prediction generated from the subset of 45 study sites. The
second map used the same three prediction maps but merging with the elevation map
from all 118 sites, instead of from just 45 sites. The areas where the four prediction
maps overlap were highlighted in black and isolated.
Percentage Error
Elevation is the factor that can be verified using actual measured GPS values and the
DEM. Prediction of elevation generated from 45 study sites, or 45 known points, was
computed against the actual values taken from the DEM and was compared with the
prediction of elevation generated from 118 study sites. From the 60 sample areas, the
elevation from both the predicted elevation generated from 45 study sites, and the
418 Gard. Bull. Singapore 63(1 & 2) 2011
APAYAO
Legend
Soil Phosphorous Content Predictior meg Promnce
Filled Contours
6.84 - 11.83
| se
ee
089-3891
38.91-4795
5-5 7
47.95-58 12
15.000 30,000 60,000 90.000 120,000
Meters
Fig. 5. Prediction Map using phosphorous content of soil generated from 45 observed sites of
Lilium philippinense.
predicted elevation generated from 118 study sites, were recorded and the percentage
error computed. Using the formula below, there is a percentage error of 68.57% for the
45 study sites while the 118 sites had only 47.16% error.
actual — predicted * 100
actual y
The prediction generated from the 45 study sites was further verified by
computing the percentage error with GPS readings from the 118 sites. The elevation
of 33 random sites not included in the 45 sites were used as the actual value. The
percentage error was 7.99%. However, if the DEM of the 33 sites were used as actual
values, the percentage error was only 4.31%.
Lilium philippinense distribution studies with GIS 419
Organic Matter Prediction Map
Filled Contours
9.23 - 12.93
12.93 - 18.21
45 Study Sites
nN
i=]
a
nN
o
s
0 15,000 30,000 60,000 90,000 120.000
Meters
Fig. 6. Prediction Map using organic matter generated from 45 observed sites of Lilium
Philippinense.
From the merged predicted maps of elevation, soil pH, and phosphorous
content of soil, a list of predicted municipalities were identified using the Municipal
maps taken from PhilGIS website. The areas are given in Table 2.
Discussion
The ArcGIS ArcMap creates a good visualisation of data geographically. It allows
modelling of the locations of where L. philippinense were observed for users to easily
identify the sites and the similarity of sites. This was shown in the different maps
produced.
mY)
420 Gard. Bull. Singapore 63(1 & 2) 2011
Table 1. Ranges for elevation, soil pH, soil phosphorous content, organic matter and the number
of study sites located within each range. * denotes ranges used in the final prediction mapping.
50% of the ranges with the highest number of study sites were taken into account.
2:0) —3.50
No. of No. of
Attribute Range Study Range Study
Sites Sites
Elevation (45 sites) 754.0000—-975.8405 i? 1483.230—-1550.718 a |
975.8405—1150.751 6* 1550.718-1636.448 2
1150.751—1288.658 5* 1636.448—1745.182 6*
1288.658-1397.392 5% 1745.182—1883.090 2
1397.392-1483.123 3 1883.090-2058.000 5
Elevation (118 Si) . 754.0000—872.6541 10 1295.578-1429.205 9
872.6541—978.0126 1S 1429.205—1579.696 14*
978.0126—1071.565 les 1579.696—1749.177 10
1071.565—1176.924 9 1749.177-1940.045 9
1176.924—-1295.578 ish 1940.045—2155.000 2s
i aE . 5.23-5.94 a o 683-687 3
5.94-6.36 4* 6.87-6.96 3
6.36—6.60 Oh 6.96—7.10 3
6.60—6.74 3 7.10—7.34 8*
6.74-6.83 4* 7.34-7.76 of
Soil phosphorous wana 6.84-11.83 ae : 38.91-47.95 55
11.83-17.44 4* 47.95-58.12 5
17.44-23.77 Sie 58.12-69.58 4
23.77-30.89 4 69.58-82.49 3
30.89-38.91 4 82.49-97.02 4
Orne anette : -0.87-1.31 = -3.56-4.83 "7 5H
1.31-1.61 6* 4.83—6.64 6*
1.61—2.05 4 6.64—9.23 l
2.05—2.67 4 9.23—12.93 4
8* 12.93-18.21 p)
2 421
Table 2. Identified municipalities under each province where Lilium philippinense was
predicted to grow.
Province Municipalities
Abra Bangued. Danglas, Langiden, La Paz, Luba. Penarrubia, Pilar. San Isidro,
San Quintin, Villaviciosa
Kalinga Balbalan. Lubuagan. Pinukpuk, Tabuk, Tanudan. Tinglayan
Apayao Conner
Mt. Province Bontoc, Sadanga. Sagada. Tadian
Ifugao Asipulo, Lagawe, Lamut. Tinoc
Benguet Atok, Bokod, Buguias, Itogon. Kabayan, Kapangan. La Kibungan. Trinidad
The Spatial Analyst IDW was able to predict the values of unknown data
given the actual points. However, the limitation of the IDW method, that did not
produce values other than the specified range of the known data, has a huge effect on
the accuracy of values. The output of the prediction cannot generate values higher or
lower than the observed or known values. A possible solution is to acquire more data
in sites, not necessarily where Lilium philippinense is observed, especially in places on
the northern part of the CCR Region. A wider spread of range of values and location
would perhaps enable better prediction.
In addition, the IDW method in interpolation to generate prediction maps
is based on proximity, and its accuracy in giving a good prediction depends on the
number of actual values surrounding that empty space or grid on the map. The nearer
and the more the actual values are to an empty space or grid, the better the prediction
for that space. This means that if the actual values were found in one specific area, the
correctness of the prediction grows less with farther distance (Longley et al. 2005). Due
to this limitation of the IDW, prediction is more accurate for areas nearer the observed
sites. In this study, the extent of the prediction maps were set to the boundaries of the
whole CCR, hence, predicted areas that were farther relative to the actual distribution
may not be accurate.
The presence of the majority of known points in the southern part of the CCR
Region also affects the prediction. The IDW relies greatly on distance, and therefore
the farther the distance of the unknown area from the area with known data, the lower
the chances for a good prediction. This is validated with the result of the percentage
error of the prediction from the 45 study sites.
422 Gard. Bull. Singapore 63(1 & 2) 2011
Conclusions and recommendations
GIS is a very useful tool in plant distribution. Specifically, the IDW feature of ArcGIS
is useful in predicting potential sites for cultivating endangered species like the L.
philippinense. The accuracy of prediction is dependent on available data and the
number of points (populations) that are plotted on the map. The more data and points,
the more accurate is the prediction.
The data provided for L. philippinense 1s still limited. Factors such as air
temperature and rainfall were not included in the prediction maps since there were no
complete measurements for the whole region. There are only three weather stations
located in the CCR and these stations are located in Baguio City and Benguet province.
References
Balangcod, T. (2009) Geographic Distribution, Ecology and Reproductive Biology of
Lilium philippinese Baker, an Endemic Species in the Cordillera Central Rang,
Luzon, Philippines. Ph.D. dissertation. Univ. of the Philippines Los Banos.
Childs, C. (2004) Interpolating Surfaces in ArcGIS Spatial Analyst. ArcUser, pp. 32—
35. www.esri.com
Elwes, J.H. (1880) A Monograph for the Genus Lilium. London: Taylor & Francis.
Fischer, F. (2009) Do we still need a desktop GIS? Geolnformatics: Magazine for
surveying, mapping & GIS professionals ESRI reprint. Vol. 12.
Hasmadi, M., Zaki, M., Ismail, H., Adnan, A.M., Pakhriazad, H.Z. & Fadlli, A.Y.M.
(2010) Determining and mapping of vegetation using GIS and phytosociological
approach in Mount Tahan. Malaysian J. Agric. Sci. 2(2): 80-89.
Information Management (2004) Group | Releases Geocoding Solution for ESRI’s
ArcGIS Software. http://www.information-management.com/news/1002282-1.
html. Accessed 13 June 2010.
Longley, P.A., Goodchild, M.F., Maguire, D.J. & Rhin, D.W. (2005) Geographic
Information Systems and Science, 2nd Ed. London: John Wiley & Sons.
Main, C.L., Robinson, D.K., McElroy, J.S., Mueller, T.C. & Wilkerson, J.B. (2004)
A guide to predicting spatial distribution of weed emergence using Geographic
Information Systems (GIS) online. App. Turfgrass Sci. doi:10.1094/ATS-2004-
1025-01-DG.
Miller, H.J. (2004) Tobler’s First Law and spatial analysis. Ann. Assoc. Amer. Geogr.
94(2): 284-289.
Moreno, E.C., Smith, J.K., Skidmore, A.K., Prins, H.H.T. & Toxopeus, A.G. (2007)
Modeling spatial patterns of plant distribution as a consequence of hydrological
dynamic processes in a Venezuelan flooding savanna. Ecotropicos 20(2): 55—73.
Murray, D.P. (2009) Spatial Distribution of Four Exotic Plants in Relation to Physical
Environmental Factors with Analysis using GIS. M.Sc. in Geography Thesis.
Virginia Polytechnic Institute & State Univ.
Lilium philippinense distribution studies with GIS 423
Palima, C.J. (1988) Ornamental vascular flora of Baguio City and vicinity, vol. 1, 2.
Unpublished.
Pallaris, K. (1998) Modelling the Distribution of Cecropia Species Using GIS-based
Techniques. M.Sc. in Geography Thesis. King’s College, London.
PhilGIS. (2009) Philippine GIS Data Clearing house. http://www.philgis.org/
freegisdata.htm
Sergio, C. & Draper, D. (2002) How to evaluate species when distribution is poorly
understood. The use of predictive studies for Iberian bryophytes. Portugaliae
Acta Biol. 20: 37-48.
Vargas, J.H., Consiglio, T., Jorgensen, P.M. & Croat, T.B. (2004) Modelling distribution
patterns in a species-rich plant genus, Anthurium (Araceae), in Ecuador. Diversity
Distrib. 10: 211-216.
Vogiatzakis, I.N. & Griffiths, G.H. (2006) A GIS-based empirical model for vegetation
prediction in Lefka Ori, Crete. Plant Ecol. 184(2): 311-323
Zeiders, M. (2002) Introduction to ArcMap. GIS Resource Document. http://www.
pop.psu.edu/gia-core/pdfs/gis_rd_02-13.pdf
54
Gardens’ Bulletin Singapore 63(1 & 2): 425-432. 2011 425
Towards the conservation
of Malaysian Johannesteijsmannia (Palmae)
Y.M. Chan', L.S.L. Chua and L.G. Saw
Forest Research Institute Malaysia, 52109 Kepong, Selangor, Malaysia
'vokemui@frim.gov.my
ABSTRACT. A total of 20 new localities were recorded for the genus Johannesteijsmannia
since 1972, demonstrating that the genus is less restricted in its distribution in Malaysia
than previously thought. Nevertheless, Johannesteijsmannia is regarded as threatened with
J. lanceolata, J. magnifica and J. perakensis assessed as endangered and J. altifrons as
vulnerable. Endangered status was given to endemic species with restricted occurrence and
small population size found in less than five localities. Recommended conservation measures
include the need to expand in situ protection for populations in vulnerable habitats, inclusion
of the species into forest management plans, and establishment of a sustainable seed harvesting
regime. We also suggest regular monitoring of populations situated along forest boundaries and
initiation of long-term conservation biology research. Habitats at risk in Jerantut-Benta (for
J. lanceolata), Serendah and Bukit Kinta Forest Reserves (for J. magnifica), and Perak, 1.e.,
Bintang Hijau, Kledang-Saiong and Bubu Forest Reserves (for J. perakensis) should be given
protected status and ex sifu conservation should be implemented.
Keywords. Conservation, extent of occurence, Johannesteijsmannia, red list, threat assessment
Introduction
Johannesteijsmannia 1s a small genus with only four species, 1.e., Johannesteijsmannia
altifrons (Reichb.f. et Zoll.) Moore, J. magnifica J.Dransf., J. lanceolata J.Dranst.
and J. perakensis J.Dransf. (Dransfield 1972). All species are endemic to Peninsular
Malaysia except J. altifrons, which is distributed from south Thailand to Peninsular
Malaysia, Sumatra and Borneo. The genus is threatened (Walter & Gillett 1998), and
seed harvesting for the ornamental plant trade contributes to its decline (Chan & Saw
2009).
Since Dransfield (1972), many new localities of Johannesteijsmannia have
been recorded based on herbarium collections and field observations. The known
extent of occurrence or distribution range in Malaysia for these species has thus
greatly increased. Johannesteijsmannia is found in tropical lowland moist forest and
lower montane forest, and the rapid change in land use patterns of such forests in the
peninsula in the last decades of the 20th century had caused further loss of habitat and
populations. This land use pattern had, however, slowed down significantly since the
early 1990s and the time is now ripe to re-assess the conservation status of its species
as more data on habitat status, population sizes, reproductive biology, and uses (Chan
& Saw 2009) are available.
426 Gard. Bull. Singapore 63(1 & 2) 2011
Here, we present the results of the threat assessment and discuss pertinent
issues regarding the conservation of Johannesteijsmannia. We also recommend
conservation and management measures for species which are threatened.
Materials and methods
Members of the genus were assessed following guidelines outlined in the Malaysia
Plant Red List Guide for Contributors (Chua & Saw 2006) using the IUCN Red List
Categories and Criteria version 3.1 (IUCN 2001). For each species, a map of its extent
of occurrence (EOO) and area of occupancy (AOO) was prepared based on specimens
lodged at the Kepong Herbarium (KEP) and the Sarawak Herbarium (SAR). This genus
is absent in Sabah, hence there was no attempt to collate records lodged at the Sandakan
Herbarium (SAN). Field observations and localities cited by Dransfield (1972) and
- Look (2007) were included. The EOO and AOO were calculated using the extensions
Crime Analysis Tool 2.E and Conservation Assessment Tools (CATS) version 1.2 for
ArcView GIS 3.2a, respectively. In assessing habitat decline for Peninsular Malaysia,
we referred to the forest cover data based on the National and State Forest Inventories
and land use data (MACRES & UTM 2008). Recent forest cover and land use data
from Sarawak, however, were unavailable for use.
Results
Since Dransfield (1972), a total of 20 new localities were recorded for
Johannesteijsmannia, with new records for Terengganu (J. altifrons) and Kedah (J.
perakensis) (Fig. 1). Johannesteijsmannia altifrons is far more commonly distributed
than once thought, especially in the east of the peninsula, with the majority of the
populations occurring in Terengganu and Johor. The collection from Temengor Forest
Reserve (FR) is anew record for Perak. With 16 additions of new localities since 1972,
the EOO and AOO have dramatically increased for J. altifrons (Table 1). This species
commonly inhabits valleys and hill slopes with lowland and hill dipterocarp forests on
well-drained soils. It also grows at elevations of 1000-1200 m in the lower montane
forests of Taman Negara, i.e., Gunung Tahan and Gunung Mandi Angin (Dransfield
1972). The population in Jerangau, Terengganu grows on waterlogged sandy soil in
low-lying areas. The habitat of J. a/tifrons in the peninsula now seems more general
and diverse than previously thought. In Sarawak, it is confined to the heath forests
(Dransfield 1972) in sheltered valleys.
Similarly, J. lanceolata, J. perakensis and J. magnifica which were thought
to be narrowly distributed endemics, now have wider distributions with new records
from Negeri Sembilan, Kedah and Perak, respectively. They are found in sheltered
hill slopes and valley bottoms, and are almost absent on ridges. One population of J.
magnifica in Bukit Kinta FR, Perak, was recorded from a limestone area.
Conservation of Johannesteijsmannia 427
Location based on: cif : Location based on ae ¥
e Herbarium specimen(s) : “| @ Herbarium specimen(s) = \
™ Forest cover as in National ‘ ™ Forest cover as in National Be
Forest Inventory III (1991-1993) > | Forest Inventory III (1991-1993) - \yws
50 0 50
—ae . 5 eee ie
Location based on: = Location based on
@ Herbarium specimen(s) FS . » | @ Herbarium specimen(s) ~~ ,
™ Forest cover as in National Ne: } Forest cover asin National ~ a By
Forest Inventory III (1991-1993) . Forest Inventory III (1991-1993) ~~ . ~}
J
Ny
So 0 SO
+ eek _
Location based on
@ Herbarium specimen(s)
50 0 50 100
=
wy 113 115 "7 us
Fig. 1. Distribution of Johannesteijsmannia in Malaysia. A. J. lanceolata. B. J. magnifica. C. J.
perakensis. D-E. J. altifrons. Polygons represent Extent Of Occurrence (EOO). See Appendix
A for locality details. National Forest Inventory, NFI III information in A-D courtesy of Forest
Department, Peninsular Malaysia.
428 Gard. Bull. Singapore 63(1 & 2) 2011
Table 1. Threat assessment of Johannesteijsmannia based on extent of occurrence (EOO) and
area of occupancy (AOO) following the IUCN Red List Category and Criteria ver 3.1 (2001).
Species EOO (km?) AOO (km?) eon Ge
Fonda snare 58,804 ; 176 _ VU acd, CI
altifrons
J. lanceolata 2,783 16 EN A4acd,B lab(1i,i11), C2a(i)
J. magnifica 1,306 16 EN Blab(ii,ili,iv), C2a(i)
J. perakensis 1,309 20 EN A4acd,B lab(i1,111,1v),C2a(1)
The genus is considered threatened, with three species endangered and one
vulnerable (Table 1). Populations of all species have declined or are declining due to
deforestation and dam construction, and the possibility that seed harvesting for the
ornamental trade could be a damaging factor requires to be better studied (Chan &
Saw 2009). Johannesteijsmannia lanceolata, J. magnifica and J. perakensis qualified
for the endangered category because they are endemics with restricted EOO and small
population sizes (with less than 250 mature individuals in the largest subpopulation).
Generally, Johannesteijsmannia spp. are gregarious but with patchy
distribution and probably limited dispersal ability. It is common to find populations
confined to a single hill or valley. Because the populations are so restricted, they are
extremely vulnerable to extinction. Clearly, any major catastrophe or destruction of a
single site is likely to wipe out the entire subpopulation. Although many new localities
have been added since 1972, the forest structure and quality in these sites and in sites
predating the Dransfield’s (1972) account have declined significantly. The species
favour pristine sites in the lowland forests and many of these sites, with the exception
of those in the National and State Parks, are no longer as pristine.
Populations located in the production forests of the Permanent Reserved
Forests (PRF) network are not spared from logging damage. The creation of large gaps
in the forest canopy and the disintegration of forest structure during logging could
harm Johannesteijsmannia, either by direct physical damage or physiologically. We
have examined a Johannesteijsmannia population in a logged-over forest of Berembun
FR and found that these palms are apparently failing to regenerate as juveniles are rare.
A few localities are within water catchment areas where no logging is allowed
and these populations are considered safe. Some of these populations, e.g., in Sungai
Lalang, Tembat and Linggiu, are now remnants of an originally larger population, after
dams were built. Only a handful of populations are in the totally protected areas of the
National and State Parks and Wildlife Reserve.
While most of the populations are located within the PRF, some occur in
state lands (Appendix A). A state land is land bank set aside by a state government
to accommodate future development. Populations that occur on state lands are thus
highly vulnerable to extermination. For example, the site of J. a/tifrons in Semariang
Road has probably been developed into a town. The population of J. /anceolata in
Conservation of Johannesteijsmannia 429
Jerantut-Benta Road is threatened by farm encroachment and future road expansion,
whereas the one along the Kota Tinggi — Mersing Road may have been destroyed by
conversion of the site to an oil palm plantation.
As the number of mature seeds is usually low, ranging from only 5 to 40 per
palm in each flowering episode (Dransfield 1970; pers. obs.), the long-term impact of
over-harvesting of seeds for the local ornamental trade needs to be examined.
Discussion
Although many new populations were recorded, the conservation status of the genus
remains threatened. Conservation measures are needed for Johannesteijsmannia
particularly for the endangered endemics J. magnifica, J. lanceolata and J. perakensis.
Key issues, challenges and suggestions pertaining to the conservation of the genus are
discussed below.
Many populations of Johannesteijsmannia altifrons are in the lowland forests
and there is a need to provide protection status to the habitats where they occur. The
initiative taken by a logging licencee to conserve a portion of the /. a/tifrons population
in the Temengor FR, Perak, through the High Conservation Value Forest approach
should be emulated. During logging, every effort should be taken to minimise niche
damage and leaving ample canopy cover.
The importance of keeping existing populations intact cannot be over-
emphasised. For Johannesteijsmannia lanceolata, J. magnifica and J. perakensis, all
localities should preferably be protected as they are endemics with small population
sizes found in less than five localities. Several localities were found to have populations
with high genetic diversity (Look 2007), i.e., Kledang-Saiong (for J. perakensis),
Temengor (for J. altifrons), Kinta (for J. magnifica) and Sungai Lalang FRs (for J.
lanceolata). These populations should be given utmost priority when proposals are
weighted. Among these, only the populations in the Sungai Lalang FR are protected.
The Sungai Lalang FR is a special area where three species, J. a/tifrons, J. lanceolata
and J. magnifica, grow sympatrically. Regular monitoring of populations sited within
several hundred metres from forest boundaries or fringes is also highly recommended
because past experiences have shown that these areas are easily encroached and
illegally converted into other land uses.
Palms and many understorey or herbaceous plants are traditionally ignored
in forest censuses and pre-felling inventory exercises, because only some targeted
economically valuable species and fruit trees are considered important. Ignorance,
which eventually leads to the lack of expertise of forest managers in the field to
recognise rare and threatened species, is a significant hindrance to conservation. In
this respect, Johannesteijsmannia should be listed in the forest management plan as a
species requiring conservation attention.
For some populations in logged-over forests, the after-logging effects on the
demography and population viability are not documented. A long-term demographic
monitoring on populations in logged-over forests, and preferably comparative studies
430 Gard. Bull. Singapore 63(1 & 2) 2011
with populations in undisturbed sites, can help to assess population health and
determine further conservation actions needed to conserve the affected populations.
Demographic studies have only been conducted on one undisturbed population of J.
lanceolata in the Angsi Forest Reserve (Rozainah & Sinniah 2005, Chan 2009). The
species is concentrated in valleys with densities ranging from 65 to 171 palms ha’.
In a 3.2 ha plot, the ratio of seedlings, juveniles and adult was 1:3:5 (Chan 2009),
indicating recruitment limitation. This could be because the population has reached the
maximum carrying capacity of the environment, or failed to regenerate because of low
seed production and significant seed loss. We also observed a low number of seedlings
in populations of J. perakensis and J. magnifica in the respective logged-over forests
of Kledang-Saiong and Berembun.
Although an export ban for all species of Johannesteijsmannia has been in
place under the Malaysia Customs (Prohibition of Export) Order since 1998, this
merely reduced demand from the international trade, but not the local trade (Chan &
Saw 2009). Seeds are still being harvested indiscriminately without regulatory limits.
A seed harvesting regime is needed to allow seed collection that does not jeopardise the
viability and regeneration of the populations. Permits are required for seed collection
and we suggest seed harvesting only at intervals of 3—5 years, during the mast flowering
years when flowering and fruiting are more intense. No seed collection from a site
should be allowed if the population shows poor regeneration. This can be indicated by
a demography census which can be easily carried out by visual enumeration or, better
still, with permanent tagging and proper count. We further recommend a certification
process for plants in trade. This would encourage nurseries to establish domestication
and propagation programmes and ultimately reduce harvesting pressures from the wild.
Palms in cultivation are known to flower and fruit more regularly. Ex situ conservation
should be the last resort because maintaining living collections in botanical gardens is
costly and often difficult. If this option is to be adopted, priority should be given to the
high-risk populations such as those on state lands.
ACKNOWLEDGEMENTS. We are indebted to the Forestry Department of Peninsular
Malaysia for providing forest cover data, and to the respective State Forestry Departments for
entry permission into forest reserves. We also thank Ms. Julia Sang from the Sarawak Forestry
Corporation for providing the SAR collection database.
References
Chan, Y.M. (2009) The Reproductive Biology and Ecology of Johannesteijsmannia
lanceolata J. Dransf. (Arecaceae). Unpublished M.Sc. thesis. University of
Malaya, Kuala Lumpur.
Chan, Y.M. & Saw, L.G. (2009) The uses of Johannesteijsmannia by indigenous
communities and the current ornamental trade in the genus. Pa/ms 53(3): 147-
S22
Conservation of Johannesteijsmannia 43]
Chua, L.S.L. & Saw, L.G. (2006) Malaysia Plant Red List. Guide for contributors,
p. 28. Forest Research Institute Malaysia, Kuala Lumpur. (http://www.tfbc.frim.
gov.my/MalaysiaPlantsRedBookGuide.html).
Dransfield, J. (1970) Studies in the Malayan Palms Eugeissona and Johannesteijsmannia.
Unpublished Ph.D. dissertation. Cambridge University, England.
Dransfield, J. (1972) The genus Johannesteijsmannia H.E. Moore, Jr. Gard. Bull.
Singapore 26: 63-83.
IUCN (2001) IUCN Red List Categories and Criteria ver 3.1. Gland, Switzerland:
IUCN Taxon Survival Commission.
Look, S.L. (2007) Population Genetics and Phylogeny of the Malesian Palm Genus
Johannesteijsmannia H.E. Moore (Palmae). Unpublished Ph.D. dissertation. The
National University of Singapore.
Malaysian Centre for Remote Sensing (MACRES) & Universiti Teknologi Malaysia
(UTM) (2008) Satellite Image Atlas of Malaysia: An Atlas of Satellite Imagery.
Kuala Lumpur: Malaysian Centre for Remote Sensing & Universiti Teknologi
Malaysia.
Rozainah, M.Z. & Sinniah, U.R. (2005) Population structure and spatial distribution
of umbrella leaf palms at Angsi Forest Reserve, Malaysia. Malaysian Forester
68(2): 98-104.
Walter, K.S. & Gillett, H.J. (eds) (1998) 1997 IUCN Red List of Threatened Plants.
Gland, Switzerland & Cambridge, UK: World Conservation Union.
Appendix A. Distribution of Johannesteijsmannia in Malaysia. # state land; * new localities
recorded since Dransfield (1972); FR — Forest Reserve.
Species Locality
State
Johannesteijsmannia lanceolata |. Jerantut-Benta Road # Pahang
2. Krau Wildlife Reserve Pahang
3. Angsi FR * Negeri Sembilan
4. Sungai Lalang FR Selangor
Johannesteijsmannia magnifica 1. Berembun FR Negeri Sembilan
2. Sungai Lalang FR Selangor
3. Serendah FR * Selangor
4. Bukit Kinta FR * Perak
Johannesteijsmannia perakensis 1. Gunung Bongsu FR * Kedah
2. Bintang Hijau FR Perak
3. Kledang Saiong FR Perak
4. Bubu FR Perak
5. Bubu FR Perak
432
Gard. Bull. Singapore 63(1 & 2) 2011
Johannesteijsmannia altifrons Sungai Lalang FR Selangor
Temengor FR * Perak
Sungai Durian FR * Kelantan
Bukit Bauk FR * Terengganu
Bukit Bandi FR * Terengganu
G. Arong FR Johor
Jemaluang FR * Johor
Panti FR Johor
Kluang FR * Johor
. Taman Negara Pahang &
Terengganu
11. Ulu Sedilv FR * Johor
12. Endau-Rompin State Park Johor
13. Kluang FR Johor
14. Lenggor FR Johor
15. Kota Tinggi— Mersing Road# Johor
16. Linggiu * Johor
17. Serasa FR Kelantan
18. Batu Papan *# Kelantan
19. Berkelah FR * Pahang
20% Lesong ER:+ Pahang
21. Tembat FR * Terengganu
22. Jerangau FR * Terengganu
23. Sungai Nipah FR * Terengganu
24. Rasau Kerteh FR * Terengganu
25. Pasir Raja Selatan FR * Terengganu
26. Bako National Park Sarawak
27. Kubah National Park Sarawak
28. Semariang Road # Sarawak
Gardens’ Bulletin Singapore 63(1 & 2): 433-450. 2011 433
Conservation status of Paraboea species (Gesneriaceae)
in Malaysia
R. Kiew', A.R. Ummul-Nazrah and L.S.L. Chua
Forest Research Institute Malaysia, 52109 Kepong, Selangor. Malaysia
‘ruth@frim.gov.my
ABSTRACT. Paraboea (including Trisepalum) is represented by 36 species in Malaysia
and displays a high level of endemism (80%) and 31 of its species are restricted to limestone
habitats. Two species are endemic in Sabah, of the 11 species in Sarawak 10 are endemic, and
in Peninsular Malaysia 16 of the 24 species are endemic. Paraboea culminicola K.G.Pearce
and P. obovata Ridl. are reinstated as distinct species and P. madaiensis Z.R.Xu & B.L.Burtt
is reduced to synonomy in P. sabahensis Z.R.Xu & B.L.Burtt. Based on information from the
Taxon Data Information Sheet, 15 species fall within the IUCN Category of Least Concern, four
as Near Threatened, three as Vulnerable, eight as Endangered, four as Critically Endangered,
and three as Data Deficient. None is Extinct. Most of the endangered species (94%) grow in
Peninsular Malaysia on limestone hills that do not lie within the network of Totally Protected
Areas and which are threatened by burning, quarrying and habitat destruction or disturbance,
from resort development or recreation or temple activities. Assignment of conservation status is
the first step in planning conservation management of endangered species, through advocating
legal protection of a network of limestone hills, particularly those where critically endangered
species grow (e.g.. Tambun Hot Springs, Perak and the Lambok hills, Kelantan), monitoring
populations of Critically Endangered species, taking steps towards resolving the status of the
poorly known Data Deficient species, and the establishment of endangered species ex situ.
Keywords. Conservation, Gesneriaceae, Malaysia, Paraboea
Introduction
In Malaysia, Paraboea (including Trisepalum) is a genus predominantly of obligate
limestone species with only five of the 36 species growing on other rock types, either
on soils derived from quartz (P. elegans), sandstone (P. obovata), igneous rocks
(P. leopoldii) and apparently from granite (the poorly known P. deterigibus and P.
paraprimulina from Sarawak). Trisepalum speciosum (Ridl.) B.L.Burtt is here included
because molecular work by Moeller et al. (2009) shows that it belongs in Paraboea.
Limestone in Malaysia is mostly of the tower karst type with a few raised coral
islands in Sabah (Lim & Kiew 1997). Although limestone occupies only a very small
fraction of the land area, it harbours disproportionate biodiversity, for example, in
Peninsular Malaysia almost 14% of the seed plant flora grows on limestone that covers
only 0.3% of the land surface (Chin 1977). There are more than 550 limestone karst
hills in Malaysia, the majority being found in Peninsular Malaysia. Sabah and Sarawak
434 Gard. Bull. Singapore 63(1 & 2) 2011
have about 60 each where in Sabah most are clustered along the Sungai Kinabatangan
(Lim & Kiew 1997). In Sarawak, the largest number is found in the Kuching Division
and the highest karst hills in Malaysia, the massive Gunung Api and Gunung Benarat
towering to 1700 m, are in the Gunung Mulu National Park (NP) (Kiew 2004). In
Peninsular Malaysia, the largest number of limestone hills occur in Kelantan (about 70
are named but there are many more small outcrops), followed by Perak with about 45
named hills. Of particular importance 1s the limestone in Langkawi, both on the main
island and about 17 islands that have been visited botanically. A few other scattered
hills occur in Kedah, Pahang, Perlis, Selangor and Terengganu.
In common with other obligate limestone species, such as balsams, begonias
and microchiritas, many of the species are narrow endemics that make them especially
vulnerable to habitat disturbance (Kiew 1991, 2001) and the flora of these isolated
limestone tower karsts are vulnerable to a variety of threats. Quarrying for cement,
road metal and marble is often considered the most severe threat but, in terms of
damage to the flora, the removal of the surrounding forest causes greater damage (Kiew
1997) because the vegetation of the limestone hills become vulnerable to burning
associated with agricultural practices in the surrounding areas. In Sabah, secondary
and logged-over forests are also prone to burning in the El Nifio years (Kiew 2001).
With increasing mobility, tourist and recreational activities associated with caving
or rock climbing are becoming a threat and in Peninsular Malaysia construction of
temples in caves and in Sabah the collecting of birds’ nests from caves (Kiew 1997)
all contribute to endanger the flora. Protection of the tower karst ecosystem is only
assured when they fall within national or state parks or geoparks. Forest reserves offer
protection provided their status does not change. However, many limestone hills are
on state land and are unprotected. For these reasons, the limestone ecosystem has long
been identified as one of the most endangered ecosystems in Malaysia (Davis et al.
1995a, 1995b; Saw et al. 2009; Chua et al. 2009).
The conservation status of Paraboea is a good indicator of the general state
of conservation of the limestone flora, because its species are found on almost every
limestone hill. Preliminary conservation assessments were provided in the monograph
on Paraboea by Xu et al. (2008), who noted that the assessments would need to
be revised once more accurate data were available. ‘Ground truthing’ is especially
important 1n assessing the threats to the flora of limestone hills because local destructive
activities have a disproportionate impact on these relatively small limestone hills and
the often narrow distributions of plant species. Local knowledge largely accounts for
differences between the categories assigned here compared with those of Xu et al.
(2008).
Endemism and distribution
Endemism is extremely high in Paraboea. There are no species in common among
Peninsular Malaysia, Sarawak and Sabah. The majority (23 species) of the 36
Conservation status of Malaysian Paraboea 435
Malaysian Paraboea species occur in Peninsular Malaysia, with 11 species in Sarawak
and two species in Sabah and within each of these areas, many species are restricted
to one hill or a group of adjacent hills. Of the 23 Peninsular Malaysian species, five
species extend into Peninsular Thailand that shares the same climate and topography
(the political boundary cutting through a floristic zone) and two are also recorded from
Sumatra (Xu et al. 2008).
Paraboea is poorly represented in Sabah. Paraboea leopoldii is known
only from Bodgaya Island (Wong et al., 1999) while P. sabahensis (Fig. 3C) is more
widespread. In Sarawak, paraboeas cluster in three limestone areas: Gunung Mulu
NP is the most biodiverse with six species (P. apiensis, P. bayvengiana, P. candissima,
P. clarkei, P. effusa and P. meiophylla), two (P. speluncarum and P. culminicola, Fig.
2C) on Gunung Subis with the latter species also known from Bukit Sarang with two
(P. clarkei and P. havilandii) on the many limestone hills in the Kuching Division
(Kiew et al. 2004). Only P. clarkei (Fig. 1C) occurs on limestone in both the Kuching
Division and the Gunung Mulu NP. The two non-limestone species (P. deterigihilis
and P. paraprimuloides) are poorly known and are as yet each known from a single
locality (See Appendix A).
In Peninsular Malaysia, the two non-limestone species are P. obovata that
grows on sandstone in Langkawi and P. elegans (Fig. 2B) that grows on quartzite
in Kedah, Kelantan and Selangor. The limestone species are grouped within three
main phytogeographical zones. The northern floristic zone is most biodiverse with 10
species. Most are found in Langkawi and associated islands (P. acutifolia, P. divaricata,
P. ferruginea, P. lanata, P. laxa, P. regularis and P. suffruticosa) with a few in Perlis
(P. gracillima) or on both Langkawi and in Perlis (P. bintangensis and Trisepalum
speciosum, Fig. 3B). Of these, four also occur in Peninsular Thailand.
The west coast zone (mainland Kedah, Perak and Selangor) is home to six
species (P. caerulescens (Fig. 2A), P. capitata (both varieties), P. paniculata, P.
parviflora, P. verticillata and P. vulpina), of which one also occurs in Sumatra. All six
occur in Perak with P. paniculata spreading to Selangor and P. verticillata distributed
from Selangor north to Kedah.
The central and northern zone (Kelantan, Terengganu and Pahang) is home to
six species and one variety (P. bakeri (Fig. 1A), P. brachycarpa (Fig. 1B), P. capitata
var. oblongifolia, P. lambokensis (Fig. 3A), P. nervosissima, P. treubii and P. vulpina),
of which one species is also found in Sumatra. Only P. capitata var. oblongifolia and
P. yulpina occur on both sides of the Main Range, 1.e., in both the west coast and in the
central and northern zones.
Three species, P. gracillima, P. obovata and P. regularis are known from
a single hill and a further seven from less than five hills (P bakeri (Fig. 1A), P.
bintangensis, P. divaricata, P. elegans (Fig. 1B), P. lambokensis (Fig. 3A), P. vulpina
and P. parviflora).
436 Gard. Bull. Singapore 63(1 & 2) 2011
A Paraboea bakeri B Paraboea brachycarpa
100°00 101°00" 102°00' 103°00 104°00' 100°00' 101°00' 102°00' 103°00' 104°00"
THAILAND THAILAND
6°00"
§"00'
4°00"
3°00"
Location based on
@ Herbarium specimen(s) ~~ 4
ears
=
Location based on
@ Herbarium specimen(s
D Forest cover as in National
Forest Inventory Ill (1991-1993) <
B Forest cover as in National
Forest Inventory Ill (1991-1993
2°00
NFI Ill Courtesy of Forestry Department Peninsular Malaysia ‘SINGAPORE
0 0 5D Kilometers
as
NFI Ill Courtesy of Forestry Department Peninsular Malaysia NU SINGAPORE
100°00' 101°00' 102°00' 103°00' 104°00' 100°00" 101°00' 102°00' 103*00' 104*00'
Selected Localities
Selected Localities (E00) na (EOO) : 15,184 sq km
; 1. Gua Pehah Kerbau 6. Bukit Serdam
1. Bukit Sagu (AOO): 8sq km 2. Gua Jaya 7. Gua Chermin (AOO) : 56 sq km
i 8. Taman Negara,
2. Bukit Tenggek Forest cover 3. Gua Gagak Batu ee Forest cover
within EOO - na : Gua Rusa 9. Taman Negara, | “thin EOO : 67 %
Gua Bama
C_ Paraboea clarkei
1: 5,500,000
6°00"
Location based on
400°
@ Herbarium specimen(s)
2°00"
INDONESIA
(Kalimantan)
110°00° 112°00°
Selected Localities
1. Gunung Mulu N-P. 4. Gunung Poing EOO): 13.490 sq km
Zs Gunung Bah 5. Braang (AO) : 76 sq km
3. Bukit Serapat 6. Seburan
Fig. 1. Distribution of Paraboea bakeri, P. brachycarpa and P. clarkei in Malaysia. EOO =
extent of occurrence and AOO = area of occupancy.
Conservation status of Malaysian Paraboea 437
100°00° 101ror 10200" 10307 10300 10000" 101°00" 102°00" 10s Oo 102°00"
002
00.9
00,9
00.
00.8
Forest cover as in National
Forest inventory Ifl (1991-1993)
00.2
SO Kacmeters 5 a 50 Kéometers
Las
NFI li Courtesy of Forestry Department Penmsuiar Malaysia NFI @ Courtesy of Forestry Oepartment Pennsule Malaysa
wo or 1c w200 1s Or 1007 blcta nd 10100" 20207 103°00" 10500"
i (E00) - 1480 sq km Selected Locateies: (E00) - 30,163 sqkm
eee je | beeen lee
3. Ipoh Forestcover | 3. Jerteh ee
C Paraboea culminicola
110°oo" 112°00 11507 115°0" 11500
6°00"
Location based on
4 00
@ Herbarium specimen(s)
2°00"
INDONESIA
(Kalimantan)
110°00° 112°00° 114°00° 116°00° 118°00°
Selected Localities |
5 EOO): 155sq km
1. NiahF.R
2 Gn Subis {AOO): 12sqkm
3. Bkt Sarang
/
———————
Fig. 2. Distribution of Paraboea caerulescens, P. elegans and P. culminicola in Malaysia. EOO
= extent of occurrence and AOO = area of occupancy.
438
Gard. Bull. Singapore 63(1 & 2) 2011
A Paraboea lambokensis
100°00 101°00 102°00" 103°00' 104°00'
THAILAND
Location based on
@ Herbarium specimen(s
DB Forest cover as in National
Forest Inventory Ill (1991-1993)
0 0 50 Kilometers
B_ Trisepalum speciosum
100°00' 101°00" 10200 103°00'
THAILAND
Herbarium specimen(s)
Forest cover as in National
Forest Inventory Ill (1991-1983)
x 0 SO Kilometers
NFL Ill Courtesy of Forestry Department Peninsular Malaysia NEI ll Courtesy of Forestry Department Peninsular Malaysia
100°00" 101700 102"00 103°00 104°00 100°00" 101°00" 102°00' 103°00' 104°00'
=
Selected Localities (EOO): na Selected Localities (E00): na
1. Bukit Senarip ee 1. Kaki Bukit 4. Pulau Anak Tikus :
2. Gua Renayang eee 2. Bukit Bintang 5. Tanjung Rhu niseeeon iii
Forest cover 3. Bukit Pinang 6. Kg. Batu Puteh
| within EOO : n.a.
Forest cover
7. Pulau Dayang Bunting | within E00 <n.a
C Paraboea sabahensis
110°00° 112°00
1: 5,500,000
Location based on
@ Herbarium specimen(s)
50 0 50 Kilometers
110°00° 112°00
114°00°
116°00° 118°00
INDONESIA
(Kalimantan)
114°00" 116°00" 118°00"
Selected Localities
EOO): 2,217 sqkm
1. Bilit, Sopiloring Hill 4. Madai Cave
2. Gamatong 5. Bukit Tengar. Sengarong F.R. | (AQO): 32sqkm
3. Bukit Batangan
Fig. 3. Distribution of Paraboea lambokensis, Trisepalum speciosum and Paraboea sabahensis
in Malaysia. EOO = extent of occurrence and AOO = area of occupancy.
Conservation status of Malaysian Paraboea 439
Conservation status
To assess the conservation status of each taxon, the Taxon Data Information Sheet
(TDIS) is completed to provide the baseline data for the assessment (Chua & Saw
2006). This includes taxonomic information, habitat and distribution, mapping of the
extent of occurrence (EOO) and area of occupancy (AOO), legal protection status,
threats and population decline, current conservation measures, utilisation, and finally
the Red List category and criteria based on the IUCN Red List Categories & Criteria,
Version 3.1, 2001 (hppt://www.iucn.org/) including the arguments for listing. For
these Paraboea species, data are based on the literature, herbarium specimens and
field work over a 30-year period. The information in the TDIS is then discussed with
one of the national assessors before it is accepted. The conservation status applies to
the situation in Malaysia, 1.e., it is a regional conservation status except for endemic
species, where it becomes the global status.
The most commonly used criteria are geographic distribution, limited number
of populations, protection status and population decline. For the last criterion, ground
truthing is essential. In a few cases an entire hill has been given over for quarrying,
e.g., Gunung Pondok (Perak) and Bukit Sagu (Pahang), are currently being quarried
and eventually will be razed to the ground. In most cases, only part of a hill is affected
by quarrying. In other cases, where the hills are on stateland, burning associated with
agricultural practices poses the major threat. This affects Paraboea species differently
depending on their niche requirements.
There are basically two groups of paraboeas in Malaysia, the ones with erect
stems, whorled decurrent leaves that are often covered in silvery indumentum, and
have purple flowers and twisted fruits; compared with the other group that are rosette
plants, with opposite petiolate leaves, white or pinkish flowers and straight fruits. The
former group grows in full sun on sheer vertical rock faces above the tree canopy or
on jagged summits. They are relatively resilient to disturbance because they may be
positioned above the burning of the surrounding vegetation and because their seeds
can invade bare rockfaces. In addition, they often occur in large numbers. The latter
group, in contrast, in most cases (P. /axa is an exception) are species that grow at the
base of the hill below the tree canopy on damp shaded rocks or rock faces, where they
are found as small local populations. These species are particularly vulnerable to all
types of disturbance. Because paraboeas occupy these different microhabitats, it is
common to find two species occurring on the same hill, for example, in the Kuching
limestone P. havilandii grows exposed on the summits and P. clarkei in damp shaded
habitats below the canopy.
No Malaysian species is extinct but 40% (VU, EN and CR species) are
endangered, with most (93%) of these occurring in Peninsular Malaysia (Table 1).
This is because most of the limestone areas in Sarawak (Gunung Mulu and Niah
NPs) are totally protected. In contrast in Peninsular Malaysia, the great majority fall
outside protected areas, namely the Langkawi World Heritage Geopark, Perlis State
Park, or on the six limestone hills in Taman Negara (Gunung Peningat, Batu Luas,
Batu Subis, Gua Telingga, Batu Biwa and Batu Ta’at), so that only eight species are
440 Gard. Bull. Singapore 63(1 & 2) 2011
Table 1. Conservation status of Malaysian taxa of Paraboea following IUCN Categories &
Criteria, ver. 3.1 (2001).
Number of taxa
Category Peninsular Malaysia Sarawak Sabah
“Extinct (EX) Agni 80> 0 0
Critically Endangered (CR) - 0 0
Endangered (EN) 7 0 1
Vulnerable 3 0 0
Near Threatened (NT) 3 l 0
Least Concern (LC) 6 8 1
Data Deficient (DD) l 2 0
totally protected. In addition, in Peninsular Malaysia relatively few hills fall within
Permanent Reserved Forests.
Species that require conservation action are those that fall within the CR
and DD categories. The three DD species (P. deterigibilis, P- paraprimuloides and
P. regularis) are all poorly known species, known from single specimens, some with
doubtful taxonomic standing (see Appendix A). Efforts need to be made to obtain
better material before their conservation status can be assessed.
Four species, all from Peninsular Malaysia, fall within the CR category. All,
P. bakeri (Fig. 1A), P. lambokensis (Fig. 3A), P. parviflora and P. vulpina, are known
from three or fewer localities where they have small populations that grow on the
damp shaded base of a hill or in mossy damp crevices or on hills that are particularly
vulnerable to disturbance. All the hills where they are found are on state land. The
most critically endangered is without doubt P. bakeri because the only two hills from
where it is known are both actively being quarried. For these CR species, a three-
pronged approach needs to be taken. The first is to monitor the populations to ensure
that further population decline does not occur, the second is to inform the relevant
stakeholders of their existence and to enlist their support in protecting the areas from
further disturbance, and thirdly to collect seed or leaf material for ex situ collections.
Already a plant-rescue project to grow and propagate P. bakeri ex situ has started.
Conservation of the limestone flora
The totally protected areas in Sarawak and in Langkawi and Taman Negara, Peninsular
Malaysia, offer protection to about half the species of Paraboea. Conservation of
paraboeas highlights the situation facing the limestone flora in general because, in
common with groups of plants that are restricted to limestone habitats, most species
Conservation status of Malaysian Paraboea 44]
are not widespread but have restricted distributions and lie within a particular floristic
zone and a few are confined to one or a few hills. Therefore, a network of hills, as is the
case in Sarawak, needs to be legally protected because with the high biodiversity of the
limestone flora, no single hill conserves more than a fraction of the flora. There is now
sufficient data to identify hills that either harbour narrow endemics or are particularly
rich in endangered species.
In protecting these tower karsts, it is important that a buffer zone of
forested vegetation surrounds them to protect against accidental burning as a result
of agricultural practices. This is crucial in preserving the damp, shaded habitat that
many paraboeas (and other limestone species) need. In addition, regular monitoring of
critically endangered species is necessary to prevent population decline.
In Malaysia, the limestone flora has repeatedly been recognised as one of the
critically endangered habitat types (Davis et al. 1995a, 1995b; Kiew 1994; Saw et al.
2009; Chua et al. 2009) and has been identified as one of the Important Plant Areas
(IPAs). One of the targets in the Malaysian National Strategy for Plant Conservation is
to conserve 50% of the IPAs (Saw et al. 2009). For example, Chua et al. (2009) have
already suggested that the limestone hills in the Meropoh/Gua Musang area of Kelantan
be included within Taman Negara. This would conserve at least one population of CR
species, P. vulpina, as well as P. nervosissima.
Mapping distribution patterns and assigning conservation status is the first step
in identifying endangered species and biodiversity hotspots that will enable a holistic
programme to be formulated that can balance the protection and conservation of the
limestone ecosystem with the various commercial, recreational and religious uses.
This is particularly critical in Perak where as yet no hill is protected and where many
hills are actively being quarried or the flora is being degraded by resort or recreational
development or activities associated with temples. Mapping shows that the most
important hill in Perak from the point of view of conservation of Paraboea species
is the Tambun Hot Springs (sometimes called Ayer Hangat), which is home to two
CR species P. parviflora and P. vulpina, as well as the Perak endemic P. capitata var.
capitata. Mapping also pinpoints those species that are known from a single or very
few hill(s), such as P. lambokensis from Gua Renayang and Gua Senarip, Kelantan.
This study on the conservation status of Paraboea, being based on sound
taxonomy and recent field surveys, serves as a model for the study of other groups of
plants that are either obligate limestone species or are endemic to Peninsular Malaysia,
Sabah and Sarawak. It aims at producing a map to show the distribution of endangered
species and those hills that require gazetting as Totally Protected Areas.
ACKNOWLEDGEMENTS. This study was carried out as part of the Flora of Peninsular
Malaysia Project (Project No. 01-04-01-0000 Khas) funded by the Ministry of Science,
Technology and Innovation. We are indebted to the Curators of the KLU, SAN, SAR and
SING herbaria for permission to examine specimens in their care; to D.J. Middleton for
supplying images of the type of Paraboea treubii, to various agencies that have funded surveys
of limestone hills, namely, the Malaysian Nature Society (Perak and Taman Negara); WWF-
442 Gard. Bull. Singapore 63(1 & 2) 2011
Malaysia (Perlis and Kelantan) and IPRA funding (Pahang and Sabah) and all those who have
accompanied us in the field, in particular S. Anthonysamy, G.W. Davison, C. Geri, B.H. Kiew,
S.P. Lim, A. Piee, A.R. Rafidah, J. Sang, J.H. Tan, Dennis G.C. Yong and many others, and to
BRAHMS (Botanical Research and Herbarium Management System) for enabling data access
from the KEP collection and mapping of localities.
References
Burtt, B.L. (1984) Studies in the Gesneriaceae of the Old World: XLVII. Revised
generic concepts for Boea and its allies. Notes Roy. Bot. Gard. Edinburgh 41:
401-452.
Chin, S.C. (1977) The limestone flora of Malaya. 1. Gard. Bull. Singapore 30: 165—
JANIS).
Chua, L.S.L. & Saw, L.G. (2006) Malaysia Plant Red List, p. 28. Malaysia, Selangor:
Forest Research Institute Malaysia.
Chua, L.S.L., Kiew, R. & Chan, Y.M. (2009) Assessing conservation status of
Peninsular Malaysian Begonias. Blumea 54: 94-98.
Curtis, C. (1896) Appendix B. Gardens Report for 1895. Singapore.
Davis, S.D., Heywood, V.H. & Hamilton, A.C. (eds) (1995a) Limestone flora of
Peninsular Malaysia. In: Centres of Plant Diversity, a Guide and Strategy for
Their Conservation, vol. 2, Asia, Australasia and the Pacific, pp. 303-307. U.K.,
Cambridge: IUCN Publications Unit.
Davis, S.D., Heywood, V.H. & Hamilton, A.C. (eds) (1995b) Limestone flora of Borneo.
In: Centres of Plant Diversity, a Guide and Strategy for Their Conservation,
vol. 2. Asia, Australasia and the Pacific, pp. 332-336. U.K., Cambridge: IUCN
Publications Unit.
Kiew, R. (1991) The Limestone Flora. In: Kiew, R. (ed) The State of Nature
Conservation in Malaysia, pp. 42-50. Kuala Lumpur: Malayan Nature Society.
Kiew, R. (1997) The Malaysian highlands and limestone hills: threatened ecosystems.
In: State of the Environment in Malaysia, pp. 66-73. Malaysia, Penang:
Consumers’ Association of Penang.
Kiew, R. (1998) The unique elements of the limestone flora of Batu Tengar Cave
(Segarong), Sabah, Malaysia. Gard. Bull. Singapore 50: 185-196.
Kiew, R. (2001) Towards a limestone flora of Sabah. In: Wong, K.M., Saari, G. & Lee,
S.S. (eds) Species, Landscapes and Islands, pp. 77-93. Kuala Lumpur: Malaysian
Nature Society.
Kiew, R. (2004) The limestone flora of Sarawak. Sarawak Mus. J., Special Issue 6:
79-89.
Lim, S.P. & Kiew, R. (1997) Gazetteer of limestone localities in Sabah, Borneo. Gard.
Bull. Singapore 49: 111-118.
Moeller, M., Pfosser, M., Jang, C.G., Mayer, V., Clark, A., Hollingsworth, M.L.,
Barfuss, M.H.J., Wang, Y.Z., Kiehn, M. & Weber, A. (2009) A preliminary
phylogeny of the ‘Didymocarpoid Gesneriaceae’ based on three molecular data
Conservation status of Malaysian Paraboea 443
sets: incongruence with available tribal classifications. Amer. J. Bot. 96(5): 989—
1010.
Saw, L.G., Chua, L.S.L. & Abdul Rahim, N. (2009) Malaysia National Strategy for
Plant Conservation, pp.1—63. Malaysia: Ministry of Natural Resources and the
Environment & Forest Research Institute Malaysia.
Wong, K.M., Pereira, J.T., Sugau. J.B. & Lim, S.P. (1999) A new species of Paraboea
(Gesneriaceae) from the volcanic islands off Semporna, Sabah. Sandakania 13:
23-30.
Xu, Z., Burtt, B.L.. Skog, L-E. & Middleton, D.J. (2008) A revision of Paraboea
(Gesneriaceae). Edinburgh J. Bot. 65: 161-347.
_ Appendix A. Conservation status of Paraboea and Trisepalum species (Gesneriaceae) in
Malaysia.
1. Paraboea acutifolia (Rid\.) B.L_Burtt
Distribution: Peninsular Malaysia and S Thailand
Habitat: On limestone rocks in forest
Conservation status: LC in Malaysia
Notes: Some populations lie within the Langkawi World Heritage Geopark. Xu et al. (2008)
note that it is considered threatened in Thailand.
2. Paraboea apiensis Z.R.Xu
Distribution: Endemic in Sarawak (Gunung Api)
Habitat: On limestone rocks.
Conservation status: LC
Notes: Gunung Api lies within the Gunung Mulu National Park.
3. Paraboea bakeri M.R.Hend. (Fig. 1A)
Distribution: Endemic in Peninsular Malaysia (Bukit Sagu, Pahang)
Habitat: Small populations on moss cushions in high shaded crevices in limestone hills where
water seeps down
Threats: Bukit Sagu and the adjacent small Bukit Tenggek are in the process of being totally
destroyed by quarrying.
Conservation status: CR B2b(iii) + c(iv)
Notes: Xu et al. (2008) gave this species a conservation status of EN Blab(ii. 111, 1v) based on
one collection from a second locality, Gua Charas. However, this species has not been
recollected from this hill suggesting that this specimen is wrongly localised. In 2011. both
these hills were surveyed and the small population of plants discovered was collected for
ex situ propagation in the Forest Research Institute Malaysia Nursery. Eventually this
species will certainly become extinct in the wild.
4. Paraboea bayengiana BL Burtt
Distribution: Endemic in Sarawak (Gunung Mulu and Gunung Benarat)
Habitat: On limestone rocks
Conservation status: LC
Notes: All its populations fall within the Gunung Mulu National Park.
444 Gard. Bull. Singapore 63(1 & 2) 2011
5. Paraboea bintangensis B.L.Burtt
Distribution: Endemic in Peninsular Malaysia (Langkawi, Perlis)
Habitat: On limestone rockfaces in light shade
Threats: Its three localities fall outside totally protected areas.
Conservation status: EN Blab(ii1)
Notes: Xu et al. (2008) gave this species a conservation status of VU D2.
6. Paraboea brachycarpa (Ridl.) B.L.Burtt (Fig. 1B)
Distribution: Endemic in Peninsular Malaysia (Kelantan, Pahang and Trengganu)
Habitat: On summits and exposed limestone cliff faces, where it occurs in sizeable populations.
Conservation status: LC
Notes: Known from many hills, including two within Taman Negara.
7. Paraboea caerulescens (Ridl.) B.L.Burtt (Fig. 2A)
Distribution: Endemic in Peninsular Malaysia (Perak)
Habitat: On summits and exposed vertical limestone rock faces
Threats: It is recorded from 6 hills, none of which fall within the network of Totally Protected
Areas and many are actively being quarried or house temples or are on state land that is
disturbed by development and/or agriculture.
Conservation status: EN B2ab(i1i)
Notes: The Batu Kurau locality is part of the Gunung Pondok massif that is currently being
razed to the ground for cement. Xu et al. (2008) gave this species a conservation status of
LC on the assumption that it is known from ‘several sites over a wide area and there are
no major threats’. The specimen from Pahang (Gua Charas) listed in Xu et al. (2008) is
based on a misidentification.
8. Paraboea candidissima B.L.Burtt
Distribution: Endemic in Sarawak (Gunung Buda and Gunung Benarat)
Habitat: On limestone rocks
Conservation status: LC
Notes: All localities lie within the Gunung Mulu National Park.
9a. Paraboea capitata Ridl. var. capitata
Distribution: Endemic in Peninsular Malaysia (Perak).
Habitat: Near base of limestone hills, on damp shaded rocks faces and in gullies
Threats: Although it is recorded from about 5 hills, none falls within the network of Totally
Protected Areas and many are actively being quarried or house temples or are on state
land that is disturbed by development and/or agriculture.
Conservation status: EN B2ab(i1)
Notes: The type specimen of P. capitata (together with the type specimens of its synonyms P.
curtisii Ridl. and P. polita Ridl.) were all collected by Curtis on 28" December 1895 from
Gunung Bujang Melaka, Perak. However, this is a granite mountain and apart from these
collections P. capitata has never been collected on anything other than limestone and,
although Gunung Bujang Melaka has been visited by other botanists on several occasions,
this species has not been recollected from there. Prior to visiting Gunung Bujang Melaka
on the way from Ipoh to Kuala Dipang, Curtis had ‘examined the limestone hills at three
or four places’ (Curtis 1896) so it is probable that a mistake was made in labelling these
specimens.
Conservation status of Malaysian Paraboea 445
9b. Paraboea capitata Ridl. var. oblongifolia Ridl.
Distribution: Endemic in Peninsular Malaysia (Perak and Kelantan)
Habitat: Near base of limestone hills, on damp shaded rocks faces and in gullies
Threats: Although it is recorded from two states and from about four hills, none falls within the
network of Totally Protected Areas and many are actively being quarried or house temples
or are on state land that is disturbed by agriculture and/or development.
Conservation status: EN B2ab(ii1)
Notes: The two varieties are distinct in the lamina size and shape. Collections from limestone
labelled Bukit Kurau and Padang Rengas both refer to different parts of the Gunung
Pondok massif that is currently being razed to the ground for cement.
10. Paraboea clarkei B.L.Burtt (Fig. 1C)
Distribution: Endemic in Sarawak (Kuching Division and Gunung Mulu National Park)
Habitat: On shaded limestone rocks below the tree canopy, common where it occurs.
Conservation status: LC
Notes: It is protected within the G. Mulu National Park. It is the only species in Sarawak to be
collected from more than one phytogeographical area. The Mulu population has much
larger leaves (24—26.5 = 11.5—12 cm) as compared with the Kuching Division specimens
(11.5—16 x 6—8.5) and perhaps deserves to be recognised as a distinct variety.
11. Paraboea culminicola K.G.Pearce (Fig. 2C)
Distribution: Endemic in Sarawak (Gunung Subis, Bukit Sarang)
Habitat: On summits and exposed vertical limestone cliffs
Conservation status: NT
Notes: It is protected within the Niah National Park. Xu et al. (2008) considered this species
synonymous with P. treubii. Paraboea culminicola is clearly different in its oblanceolate
leaves in whorls of 4 with 35—40 pairs of veins and cinnamon-brown undersides, and in its
inflorescences with large floral leaves and the peduncles of the side branches that are c. 8
times longer that the ultimate branches. So it is here treated as a distinct species.
12. Paraboea detergibilis (C.B.Clarke) B.L.Burtt
Distribution: Billiton, Bangka and W Sumatra, Indonesia, and Sarawak (Gunung Gaharu).
Habitat: Not known but Gunung Gaharu, Kuching Division, is not a limestone mountain.
Conservation status: DD in Malaysia
Notes: Xu et al. (2008) reported this species in Sarawak from a single specimen. The population
on Gunung Gaharu needs to be relocated before the status of this species can be assessed.
13. Paraboea divaricata (Ridl.) B.L.Burtt
Distribution: Endemic in Peninsular Malaysia (Langkaw1).
Habitat: On limestone rocks
Threats: Known from 2 or 3 localities, it nowhere occurs within a Totally Protected Area.
Conservation status: EN B2ab(ii1)
Notes: Xu et al. (2008) gave this species a conservation status of VU D2 on the grounds that
the Ayer Hangat site falls within a Forest Reserve.
446 Gard. Bull. Singapore 63(1 & 2) 2011
14. Paraboea effusa B.L.Burtt
Distribution: Endemic in Sarawak (Gunung Mulu National Park)
Habitat: On limestone rocks in forest and also on the summit of limestone karsts.
Conservation status: LC
Notes: Fully protected within the Gunung Mulu National Park.
15. Paraboea elegans (Ridl.) B.L.Burtt (Fig. 2B)
Distribution: Peninsular Malaysia (Kedah, Kelantan and Selangor) and S Thailand.
Habitat: In light shade on quartzite outcrops, on Gunung Jerai, Kedah, at 1000 m and in
Kelantan and Selangor at c. 300 m.
Threats: The rocky habitats where it grows are vulnerable to disturbance, for example, on
Gunung Jerai by tourist pressure, at the Kelantan site by maintenance work to the nearby
hydroelectric dam and in Selangor by logging that has caused landslips at the base of the
outcrop.
Conservation status: VU B2ab(ii1) in Malaysia
Notes: Xu et al. (2008) included P. obovata as a synonym of this species but it is quite different
(see 25. P. obovata below).
16. Paraboea ferruginea (Ridl.) Ridl.
Distribution: Endemic in Peninsular Malaysia (Langkaw1)
Habitat: On limestone rocks in damp, shaded places
Conservation status: NT
Notes: Known from several sites on the main and smaller islands of Langkawi, some of which
lie within the Langkawi World Heritage Geopark.
17. Paraboea gracillima Kiew
Distribution: Peninsular Malaysia and S Thailand
Habitat: It occurs in small populations on shaded limestone cliffs from the base under forest
canopy to near the summit
Conservation status: LC in Malaysia
Notes: It grows within the Perlis State Park.
18. Paraboea havilandii (Ridl.) B.L.Burtt
Distribution: Endemic in Sarawak (Kuching Division)
Habitat: On the summit and on exposed limestone cliffs where it is quite common
Conservation status: LC
Notes: This species is common and collected from many hills. The specimen cited by Xu et al.
(2008) from Pahang (Henderson SFN 25250) is in fact the type specimen of Emarhendia
bettiana (M.R.Hend.) Kiew et al.
19. Paraboea lambokensis Kiew (Fig. 3A)
Distribution: Endemic in Peninsular Malaysia (Kelantan)
Habitat: Small populations grow at the base of limestone cliffs or around cave mouths in light
shade
Threats: The two hills (Gua Senarip, Gua Renayang) are on state land surrounded by agriculture
that has removed the surrounding forest cover leaving the population vulnerable to burning.
Conservation status: CR BI b(i1i) + c(iv)
Conservation status of Malaysian Paraboea 447
20. Paraboea lanata (Rid|.) B.L.Burtt
Distribution: Endemic in Peninsular Malaysia (Langkawi).
Habitat: On limestone rocks by the seashore or in rocky hillsides on limestone islands.
Conservation status: VU B2ab(iii)
Notes: Known from several localities on Langkawi and several minor islands, some of which
fall within the Langkawi World Heritage Geopark.
21. Paraboea laxa Ridl.
Distribution: Endemic in Peninsular Malaysia (Langkaw1).
Habitat: On limestone rock faces and summits
Conservation status: VU Blab(iii)
Notes: Known from several localities on Langkawi and several minor islands, some of which
fall within the Langkawi World Heritage Geopark.
22. Paraboea leopoldii K.M.Wong, J.T.Pereira, Sugau & S.P.Lim
Distribution: Endemic in Sabah (Bodgaya Island)
Habitat: Its localised populations grow on exposed igneous rocks from 4 m above the shoreline
to high up on vertical cliffs.
Threats: Its population is difficult to access since it grows on rocky headlands and along the
coast.
Conservation status: LC
Notes: Xu et al. (2008) erroneously recorded this species from limestone.
23. Paraboea meiophylla B.L.Burtt
Distribution: Endemic in Sarawak (Gunung Benarat)
Habitat: On limestone rocks
Conservation status: LC
Notes: Totally protected within the Gunung Mulu National Park.
24. Paraboea nervosissima Z.R.Xu & B.L.Burtt
Distribution: Endemic in Peninsular Malaysia (Kelantan, Pahang)
Habitat: Exposed summit and vertical limestone cliffs
Conservation status: LC
Notes: Quite widespread with some localities lying within Taman Negara.
25. Paraboea obovata Ridl.
Distribution: Endemic in Peninsular Malaysia (Langkawi).
Habitat: In heath forest, in light shade on sandstone rocks.
Conservation status: LC
Notes: This species is known only from Gunung Machinchang, a sandstone hill that lies
within the Langkawi World Heritage Geopark. Paraboea obovata has been considered as
a synonym of P. elegans (Burtt 1984, Xu et al. 2008), perhaps because both species do not
grow on limestone. However, they are clearly different. Paraboea obovata has opposite,
obovate leaves that measure 6—7.5 x 2.24 cm and has a distinct petiole 1.24.5 cm long
and peduncles 9-11 cm long. In contrast, P. elegans has whorled, narrowly lanceolate
leaves 4 x 1.7 cm and either lacks a petiole or has a short petiole c. | cm long, and it has
long peduncles 15—18.5 cm long. For these reasons, P. obovata is therefore here reinstated
as a distinct species.
448 Gard. Bull. Singapore 63(1 & 2) 2011
26. Paraboea paniculata (Ridl.) B.L.Burtt
Distribution: Peninsular Malaysia (Perak, Selangor) and Sumatra.
Habitat: On the summit and vertical limestone cliffs
Threats: None of the hills where it is found lies within the network of Totally Protected Areas.
Only Bukit Takun lies within a Forest Reserve but vegetation of this hill is disturbed by the
activities of rock climbers.
Conservation status: EN B2ab(ii1) in Malaysia
Notes: Xu et al. (2008) gave this species a conservation status of LC because it ‘has been
collected at several sites over a wide area and there are no major threats’.
27. Paraboea paraprimuloides Z.R.Xu
Distribution: Endemic in Sarawak (Hose Mountains)
Habitat: Growing on cliffs, but not on limestone
Threats: Unknown.
Conservation status: DD
Notes: This species is known only from the type collection made in 1967 and is still
incompletely known. Until the population has been relocated, its conservation status cannot
the presumption that species known only from the type are confined to the area where they
were collected and that their extent of occurrence is below 100 km’.
28. Paraboea parviflora (Ridl.) B.L.Burtt
Distribution: Endemic in Peninsular Malaysia (Perak)
Habitat: Very uncommon and where it occurs it is found in small populations in shaded gullies
in limestone cliffs below the tree canopy.
Threats: None of the four limestone hills where it occurs lies within the network of Totally
Protected Areas and all are no longer surrounded by forest but are heavily disturbed by
tourism (Gunung Tempurung), resort development (Ayer Hangat as the Tambun Hot
Springs is sometimes known) or quarrying (Kinta).
Conservation status: CR B2ab(111)
Notes: Xu et al. (2008) gave this species a conservation status of LC stating that ‘the relatively
few collections ... are almost all from protected areas and there are no major threats’.
29. Paraboea regularis (Rid\.) Ridl.
Distribution: Endemic in Peninsular Malaysia (Langkawi) and S Thailand.
Habitat: Not known
Conservation status: DD in Malaysia
Notes: For Peninsular Malaysia, it is known only from the type specimen which comprises
detached leaves and inflorescences taken from a plant grown in the Singapore Botanic
Gardens in 1893. In spite of Langkawi being well-collected, this species has not re-found,
which raises the possibility that it is a Thai species.
30. Paraboea sabahensis Z.R.Xu & B.L.Burtt, Edinb. J. Bot. 48 (1991) 12. Type: Sabah,
Kinabatang District, Bilit, Sopiloring Hill, Ampuria SAN 35269 18 April 1963 (holo E; iso
K, L, SAN, SAR). (Fig. 3C)
Synonym nova: Paraboea madaiensis Z.R.Xu & B.L.Burtt, Edinb. J. Bot. 48 (1991) 4. Type:
Sabah, Semporna District, Madai Caves. Tamura & Hotta 722 (holo E; iso KYO).
Conservation status of Malaysian Paraboea 449
Distribution: Endemic in Sabah (Kinabatangan and Sempoma Districts)
Habitat: On exposed summits and vertical limestone hills. quite common locally
Threats: Batu Tengah Cave. Bilit. and Bukit Batangan limestone hills lie within very disturbed
forest or secondary forest that in Sabah is prone to buming in El Nifio years, while the
limestone vegetation is somewhat disturbed on Gunung Madai and Bukit Dulong Lambu
(Gomantong) because they are the most important caves in Sabah for the collecting of
bird nests. However. a summit species can withstand some disturbance. for example, the
population on Bukit Dulong Lambu has spread onto summit areas laid bare by the Great
Burn in 1982/83.
Conservation status: EN B2ab(i1i)
Notes: With the collection of specimens from more populations, it has become clear that the
differences in lamina shape and size, including shape of the leaf base, and petiole length that
were used to distinguish between P. sabahensis and P. madaiensis are not discrete. Paraboea
sabahensis is preferred for the name of this species because of its wider distribution with
the consequence that P. madaiensis is reduced to synonomy.
31. Paraboea speluncarum (B.L.Burtt) B.L_Burtt
Distribution: Endemic in Sarawak (Gunung Subis).
Habitat: On limestone growing in light shade on stalactites from the arch of cave mouths.
Conservation status: LC
Notes: It lies within the Niah National Park.
32. Paraboea suffruticosa (Rid!.) B-L.Burt
Distribution: Endemic in Peninsular Malaysia (Langkawi).
Habitat: On limestone karsts
Conservation status: NT
Notes: It is known from several localities on the main island and from smaller islands, some of
which lie within the Langkawi World Heritage Geopark.
33. Paraboea treubii (H.O Forbes) B-L_Burtt var. rreubii
Distribution: Peninsular Malaysia (Kelantan, Pahang) and Sumatra.
Habitat: Not common, it grows exposed on summits and on vertical limestone cliffs.
Conservation status: LC in Malaysia
Notes: Some of the hills on which it occurs lie within Taman Negara.
34. Paraboea verticillata (Rid|.) B-L-Burt
Distribution: Endemic in Peninsular Malaysia (Kedah. Perak, Selangor)
Habitat: Common, growing exposed on summits and vertical limestone cliffs
Threats: None of the hills where it is found lie within the network of Totally Protected Areas.
Only Bukit Takun lies within a Forest Reserve but vegetation of this hill is disturbed by the
activities of rock climbers.
Conservation status: EN B2ab(111)
Notes: Xu et al. (2008) gave this species a conservation status of LC because it “has been
collected at several sites over a wide area and there are no major threats’.
450 Gard. Bull. Singapore 63(1 & 2) 2011
35. Paraboea vulpina Ridl.
Distribution: Peninsular Malaysia (Kelantan, Perak) and S Thailand.
Habitat: Small, local populations on the shaded base of vertical limestone cliff faces.
Threats: None of the three hills from where it is known is protected within the network of
Totally Protected Areas or is surrounded by forest. All are in heavily disturbed areas close
to agriculture or tourist developments.
Conservation status: CR B2ab(i11) in Malaysia
Notes: Xu et al. (2008) gave this species a conservation status of LC because ‘although this
species is only known from a few collections it is found over a large area and several of the
known sites are in protected areas and there are no major threats’.
36. Trisepalum speciosum (Ridl.) B.L.Burtt (Fig. 3B)
Distribution: Peninsular Malaysia (Langkawi, Perlis) and S Thailand.
Habitat: On exposed summits of limestone hills
Conservation status: NT in Malaysia
Notes: It is known from several localities on the main island of Langkawi and from smaller
islands, some of which lie within the Langkawi World Heritage Geopark. It also grows on
the mainland.
Gardens’ Bulletin Singapore 63(1 & 2): 451-464. 2011 451
Utricularia (Lentibulariaceae) habitat diversity in
Peninsular Malaysia and its implications for conservation
M.Y. Chew! and N.W. Haron?
'Forest Research Institute Malaysia, 52109 Kepong, Selangor, Malaysia
chew@frim.gov.my (corresponding author)
"Institute of Biological Sciences, Faculty of Science,
University of Malaya, 50603 Kuala Lumpur, Malaysia
ABSTRACT. Utricularia is a cosmopolitan carnivorous genus with more than 30 species
in Malesia, of which 14 occur in Peninsular Malaysia. Utricularia species exhibit a range
of habits including free-floating or affixed aquatic, semi-aquatic, terrestrial, lithophytic
or epiphytic. In terms of habitat preference, three arbitrary groups are recognised, namely,
habitat specialists, habitat generalists, and open and wayside pioneers. This grouping allows
information on niches to be interpreted into conservation management measures. One third
of the Peninsular Malaysian species are habitat specialists, found either in single localities
or in one microhabitat. Among them, U. furcellata and U. scandens are listed as ‘Critically
Endangered’ for the Red List for Peninsular Malaysia, whereas U. involvens and U. punctata
are ‘Vulnerable’ and U. vitellina is ‘Rare’. Four species, U. caerulea, U. gibba, U. striatula and
U. uliginosa, are found in many sites and microhabitats and are thus considered generalists,
with their conservation status varying from ‘Vulnerable’ to ‘Least Concern’. Utricularia aurea,
U. bifida and U. minutissima are adaptable pioneers able to co-exist with weeds and they may
also be indicators of past disturbance. Two rare species, U. limosa and U. subulata, have not
been relocated recently and their local habitat preferences are uncertain.
Keywords. Conservation, habitat diversity, Peninsular Malaysia, Utricularia
Introduction
Utricularia L. (Lentibulariaceae) is a large genus of carnivorous plants, with c. 220
species worldwide and c. 30 species in Malesia (Taylor 1989). It is cosmopolitan,
found in all continents from subarctic landscapes to tropical rain forest, at oases in
deserts and on oceanic islands (Brummit 2007). To date, 14 species are recorded from
Peninsular Malaysia.
The genus has a unique body plan among flowering plants as highlighted by
Rutishauser & Isler (2001). Utricularia has no true root; the rudimentary anchoring
rhizoids lack a root cap. Its stolons have randomly arranged phloem and xylem instead
of the collateral vascular bundles typical of angiosperm stems. The foliar organs or
leaves arise at the bases of peduncles or along stolons. Leaf laminas of the terrestrial
species are often minute, while in the aquatic species they are much dissected. The
inflorescences are racemes with indefinite growth, sometimes branched or twining.
The minute traps are highly modified, glandular organs arising from the leaves, stolons
452 Gard. Bull. Singapore 63(1 & 2) 2011
or rarely from other parts. These traps function to supplement their nutrient intake by
trapping microfauna, microflora and microbes (Richards 2001, Sirova et al. 2009).
In Peninsular Malaysia, Utricularia is found almost exclusively in nutrient-poor
environments with low pH.
Three species of Utricularia in Peninsular Malaysia are free-floating aquatics
from the section Utricularia, that includes U. aurea Lour., U. gibba L. and U. punctata
Wall. ex A.DC., of which the latter two also exist as affixed aquatics. Nine species
are terrestrials that are sometimes semi-aquatic. They are U. minutissima Vahl from
section Meionula, U. caerulea L. from section Nigrescentes, U. limosa R.Br. from
section Nelipus, U. subulata L. from section Setiscapella and U. bifida L., U. involvens
Ridl., U. scandens Benj., U. uliginosa Vahl and U. vitellina Ridl. from section
Oligocista. There are only two species of minute and rosette lithophytic herbs, namely
U. furcellata Oliv. and U. striatula Sm. from section Phyllaria, which at times are
terrestrial or epiphytic within the cloud forest or waterfall splash zones.
Authors prior to Ridley gave little mention of the ecology and distribution
of Utricularia species found in the region. Ridley (1893, 1895, 1901, 1908, 1923),
Henderson (1928) and Spare (1941) provided background knowledge on the habitats
and commonness of local Utricularia but held slightly different opinions on taxonomic
delimitation. Taylor’s (1977) treatment of the genus in Malesia detailed the distribution
and ecology for every taxon, although this was not specific to Peninsular Malaysia.
Turner (1995) summarised habitat information provided by Ridley and Taylor in his
plant checklist for the Peninsula but did not add new information. Parnell’s (2005)
account on Thai Ufricularia provided ecological details on habitat and substrate type,
altitudinal range and flowering period, of which eleven of the Peninsular Thailand
species also extend into Peninsular Malaysia.
This study was carried out as part of the revision of Utricularia for the Flora of
Peninsular Malaysia. It aimed to document the range of habitats and niches occupied
by various Utricularia species, to understand the implications of their distribution
ranges and habitat preferences, and to formulate conservation measures based on
the assessments of such information. For non-endemic species, the Red List status
obtained 1s only applicable to Peninsular Malaysia.
Materials and methods
General collection was carried out at various wet habitats to obtain materials of the
common and widespread species, and specialised fieldtrips were organised to relocate
rare species at specific sites. Habitat information including GPS readings, elevation,
exposure, substrate type, water depth, pH, associated plants, range of niches occupied
and known history of disturbance were recorded for specimens collected in the field.
Relevant label information for existing herbarium specimens was databased. In the
absence of any habitat records, information was inferred from the Kepong BRAHMS
gazetteer database.
Utricularia habitats and conservation 453
Table 1. Criteria for arbitrary habitat-preference groupings.
Number of .
Microhabitat
Habitat-preference group collection tee Tolerance to disturbance
specificity
localities
; ye = ee Pristi / jj di
Habitat specialists 12 <5 niche subtypes PSHE “emi eismags
(by trails etc.)
‘ ; & we Mostly 1 2
Habitat generalists E25) > 5 niche subtypes stly in or near natural
vegetations
Mostly disturbed
(abandoned land)
Open / wayside pioneers Many (=>30) > 8 niche subtypes
The lowest and highest spatial range occupied by a species was shown by the
elevation data. Latitude and longitude of collection sites were plotted with ArcView
to generate the Extent of Occurrence (EOO) and Area of Occupancy (AOO) maps.
Three arbitrary habitat-preference groups were established according to criteria listed
in Table 1.
Conservation assessment was carried out following the IUCN Red List
Categories and Criteria version 3.1 (2001). The assessment was largely specimen
based; verified or published records were included when specimen information was
lacking. The Taxon Data Information Sheets (TDIS) modified from the IUCN Red List
Assessment Questionnaire to suit the requirements for local plants, as recommended
by the Malaysia Plant Red List guidebook (Chua & Saw 2006), were then completed
for each species. TDIS comprises five parts, i.e., Taxon Attributes, Geographical Range
and Demographic Details on Population, Red List Category and Criteria Assessment,
Current Conservation Measures for the Taxon and Utilisation.
Habitat preferences and the associated population size data provided the basis
for the IUCN Red List criterion A, scoring on population reduction. EOO, AOO and
distribution patterns were used to evaluate the Criterion B, scoring on geographical
range, which has been designed to identify populations with restricted, declining or
fluctuating distributions in the present or near future (IUCN 2010).
Results and discussion
Microhabitat types, biotic and abiotic properties and altitudinal range
In Peninsular Malaysia, Utricularia occurs in a range of perpetually or seasonally wet
microhabitats—from pristine lowland and montane swamps to fairly disturbed, meso-
eutrophic ditches, as detailed in Appendix A. The aquatics usually occupy the open
shallow waters; the terrestrials or semi-aquatics grow on multifarious waterlogged or
shallowly inundated soils; lithophytic species often grow on rock-faces with dripping
water. Some of these habitats are subject to periodic drought, where the annual or
ephemeral Utricularia populations die out in the dry months. In line with the cost-
454 Gard. Bull. Singapore 63(1 & 2) 2011
benefit model for carnivorous plants (Givnish et al. 1984), the genus generally prefers
sunny, moist and low-nutrient habitats with low pH (3-7). Some species are able to
tolerate deep shade and eutrophic waters but do not flower under such conditions.
The altitudinal range of the genus extends from the coastal Typha reed beds in
Pulau Langkawi to the montane sandstone plateau of Gunung Tahan, the highest peak
in Peninsular Malaysia at 2187 m. Edaphic conditions and light availability are more
crucial in determining the presence or absence of Utricularia, rather than altitude. It
is not found in mangroves, coastal lagoons, inland salt-licks and hot springs, due to
its intolerance for high pH and salinity; nor reservoirs with steeply shelving shores,
large swift rivers and concrete waterways which do not provide stable substrate for
establishment. The genus is generally absent from tall-forested areas with closed
canopies while being fairly common along streams, heaths, swamps, along forest
fringes, in forest gaps, trails or other edaphic and biotic vegetation types nestled within
the climactic forests, wherever there is adequate sunlight and moisture.
Habitat preferences and conservation status
The habitat specialists are species that are found only in a single locality or one type
of microhabitat. One third of species from the Peninsula falls within this group. The
habitat generalists consist of plants that are found in many sites and suited to live in
many types of wet microhabitat but rarely found in heavily disturbed sites. Common
pioneers of open and wet-habitat are adaptable plants that are able to co-exist with
weeds and may indicate past disturbance.
The conservation status ofa particular species is related to its habitat-preference
because this affects its overall distribution and commonness. If the particular locality
where a species is found is within a protected area, the conservation status then falls
into a much lower category, as listed in Table 2. However, although population decline
is perceived as minimal in protected areas for most plants, the type of fringe habitats
Utricularia favours is often sacrificed in the process of amenity or trail development,
or depleted by high-impact or uncontrolled recreational activities. Thus, their survival
is not fully guaranteed, especially for the habitat specialist and habitat generalist.
Habitat specialists
Utricularia furcellata (Fig. 1A) is a new record for Peninsular Malaysia (Chew
et al. 2011). It is differentiated from the more common U. striatula as detailed in
Appendix B and Fig. 2A—B. Utricularia furcellata grows in a localised population
on the montane heath of Gunung Ayam, within Gunung Stong State Park, Kelantan.
The species is previously known to grow on moist rocks (1500-2700 m) in North-
eastern India and mossy wet tree boles and rocks in lower montane forest (>1700 m) in
Northern Thailand. In Kelantan, U. furcellata exists as a terrestrial herb on a patch of
white-sandy heath within the cloud forest zone inundated with a thin film of water. The
site is traversed by a major hiking trail. Up to 2003, the trail was reported to receive an
average of more than 2500 climbers (out of 5000 visitors to the area) annually with an
increasing trend (Maser et al. 2006).
Utricularia habitats and conservation 455
Table 2. Conservation status of Utricularia species in Peninsular Malaysia in relation to habitat
preferences. CR = Critically Endangered: VU = Vulnerable; RA = Rare; NT = Near Threatened:
LC = Least Concern; DD = Data Deficient.
Habitat Species Conserva- Rationale used in assessment
preference tion status
Habitat U. furcellata CR Single microhabitat in State Park but with heavy
specialist trekker traffic
U. scandens CR Single microhabitat in State Park with natural
catastrophe (flash flood) risk
U. involvens VU Single protected locality but affected by amenity
development
U. punctata VU Last remaining protected locality with invasive
waterweed risk
U. vitellina RA Sensitive montane species in 2 protected localities
Habitat U. caerulea VU Highland pink form in 2 protected localities;
generalist lowland white form experiencing population
decreased
U. striatula NT Fairly common in mountainous waterways but
sensitive to drought
U. uliginosa NT Fairly common in acidic swamps but always
within forested areas
U. gibba LC Fairly common in acidic open waters
Open / U. aurea LE Most common aquatic Utricularia in Peninsular
wayside Malaysia
ioneer Sree : ;
P U. bifida LC Common pioneer in wayside nutrient-poor wet
habitats
U. minutissima LC Fairly common in natural and wayside nutrient-
poor wet habitats
Uncertain U. subulata DD Last collected in 1925, not rediscovered, sites
developed
U. limosa DD Single collection in 1937, site developed
Utricularia scandens (Fig. 1B) used to be recorded from a number of rocky
heaths or ridges that were constantly inundated on the small isolated hill range of the
Gunung Ledang State Park, Johor (Ridley 1901). Gunung Ledang is popular among
local and foreign tourists and received more than 11,000 visitors annually (Suksuwan
& Ong 2005). According to the description by Ridley and nature guides who trekked
the hill since the 1980s (Gan & Kueh, pers. comm.), the rocky heaths originally had
Gard. Bull. Singapore 63(1 & 2) 2011
Fig. 1. Habit and habitat of four Utricularia species. A. Utricularia furcellata from the Gunung
Ayam heath, Gunung Stong State Park. B. Utricularia scandens at Padang Batu, Gunung
Mering, Gunung Ledang State Park. C. Utricularia involvens at Gunung Jerai Forest Reserve,
Kedah. D. Utricularia punctata at the Tasik Bera RAMSAR site, Pahang.
Utricularia habitats and conservation 457
2/5/2010 |mag| W spot| det HV
10:39:26 AM |500 ead mm] 5.0 |ETD|15.00 kV U. furcellata-seed
2/5/2010 |mag| WD _ |spot} det pressure
2:37:27 PM|700 x|22.6 mm} 5.0 |LFD|15.00 kV| 60Pa
Fig. 2. Scanning electron microscopy images of Utricularia seeds. A. Utricularia
furcellata seed showing periclinal testa cell-walls densely covered with globose or shortly
clavate verrucae and the long processes that end in knobbly, clavate tips; size c. 455 um.
B. Utricularia striatula seed showing relatively smooth periclinal testa cell-walls with
verrucae occurring only along sinuate boundaries, and short processes with glochidiate,
stellate tips, size c. 251 um.
458 Gard. Bull. Singapore 63(1 & 2) 2011
sparse forested cover on thin soils and were constantly inundated by a thin film of
seepage that flowed from the peaty, wooded ridges higher up. Much of the tracks
along the rocky heath and the narrow summit ridge lost part of their vegetation cover
either from the firewood cutting, campsite clearing or trees toppling over on heavily
trampled soils, exposing the bedrock. Further soil erosion and loss of ground cover
followed in these areas, which eventually caused the microhabitats to dry up. Parts
of the trails and campsites along the waterways were littered and polluted by human
waste that resulted in siltation and algal blooms in slow-flowing or stagnant waters.
By the time the State Park was established in 1997 under the Johor National Parks
Cooperation and stricter rules were enforced against destructive camping practices,
U. scandens populations were depleted from all previously known collecting spots
that were impacted by heavy tourist traffic. It was recently relocated from Padang
Batu on Gunung Mering, a remote rocky heath above the Lampung Jatuh waterfall.
The particular site was later destroyed by a rare flash flood event, further reducing the
population.
Both U. furcellata and U. scandens are found in single microhabitats at one
restricted site, making them highly susceptible to localised disturbance and microclimate
changes. The case of U. scandens 1s a typical instance in which a sensitive species
occupying an open microhabitat fringing a waterway frequented by human trekkers
would inevitably disappear from the intensively utilised zones, persisting only in the
more inaccessible sites. The current management practices in the two State Parks do
not restrict current or future use of the sites where these sensitive species occur, hence
qualifying both species ‘Critically Endangered’ status.
Although occurring 1n a variety of microhabitats, U. involvens (Fig. 1C) has
never been found outside the Gunung Jerai Forest Reserve, Kedah. There it occurs in
various open to shaded wet microhabitats. Although protected, the site is a popular
tourist spot with an army camp at the summit. The species is under much pressure from
amenity development and this has already led to some previously wet sites drying out.
Being susceptible to local microclimate change, the conservation status of this species
is scored as ‘Vulnerable’.
Utricularia punctata (Fig. 1D) grows in slow-flowing, shallow open water.
Habitat conversion affecting the backwaters of Sungai Pahang and Kota Tinggi, and
the introduction of the noxious weed from South America, Cabomba furcata Schult.
& Schult.f. in Roem. & Schult., that had taken over the niche of U. punctata in
Tasik Chini, has probably wiped out the original population of the species in these
places. Currently, it is surviving in Tasik Bera, Pahang—the largest freshwater lake
in Peninsular Malaysia that is protected as a RAMSAR wetland site. Although locally
abundant, it warrants a ‘Vulnerable’ status from the risk of being displaced by invasive
aquatic weeds.
The sole endemic species for Peninsular Malaysia, U. vite/lina, is found only on
the two highest summits in Peninsular Malaysia, on montane peaty bryophyte mounds
usually along stream banks. The highest peak, Gunung Tahan, is located within Taman
Negara Pahang, while Gunung Korbu, Perak, lies within a forest reserve. Although
locally abundant in the two protected sites, this species occurs only in one type of
Utricularia habitats and conservation 459
microhabitat and is currently absent from the campsites and most parts along the trails.
It is susceptible to disturbance from human activities and microclimate change, and is
given a ‘Rare’ status.
Wet habitat generalists
Utricularia caerulea has two flower colour forms in Peninsular Malaysia. White-
flowered populations are mostly lowland terrestrial plants, found in wet, sandy
or muddy sites. Only two collections of plants with white flowers have been made
after 1980, the others were collected from 1890 to 1964. The pink-flowered form is
a terrestrial or semi-aquatic plant that grows along stream banks, and has only been
recorded from Gunung Jerai and Gunung Ledang. Although the species was previously
found in many sites and occupies a variety of habitats, the white-flowered form of the
lowlands is becoming rare while the pink-flowered form is restricted to two sites. It is
therefore given a ‘Vulnerable’ status.
Found on major mountain peaks and a few large waterfalls, U. striatu/a is
either a minute terrestrial, lithophytic or is an epiphytic herb. Utricularia uliginosa 1s
usually found in the lowlands and sometimes in the highlands. It is a stout terrestrial
or semi-aquatic Utricularia of peat swamps, heaths and forest edges. Despite being
found in many sites and able to adapt to many different microhabitats, these two ‘Near
Threatened’ species appear to be sensitive to human disturbance and are not found
outside naturally vegetated areas.
Another Utricularia that is rarely found away from natural habitats is U. gibba.
The species grows in slow-flowing to stagnant waters usually bordering dryland forest
or peat swamps. It sometimes colonises old man-made water bodies but has never
been seen in an eutrophic lake. This species is thus given a ‘Least Concern’ status.
Wet habitat pioneers
Utricularia bifida and U. minutissima are two common terrestrial or semi-aquatic
species found in open wayside wetlands and grasslands. Naturally, both are restricted
to open stream banks or heath but have managed to colonise many man-made areas
with fairly established vegetation and relatively stabilised soil and are often found in
association with each other.
Utricularia aurea is the most common aquatic free-floating Utricularia found
locally, usually in slow-flowing or stagnant water. Although commonly found in
ditches around agriculture areas, it is not known to be weedy. Like all other Utricularia
species, it is sensitive to all forms of fertiliser, and is noticeably absent from intensively
worked paddy fields and eutrophicated old mining lakes, although it had been reported
as common in these habitats by Ridley (1923) and Spare (1941).
Rare species with uncertain habitat preferences
Utricularia limosa was last recorded from low-altitude swamps in 1925 by Holttum,
while U. subulata was last collected from Teluk Merbau, Selangor, in 1937. Repeated
attempts to relocate these species from previously known sites has yielded nothing.
460 Gard. Bull. Singapore 63(1 & 2) 2011
All the historical collection sites have been converted to built-up areas, agricultural or
degraded lands. Due to a lack of information, both species are given a ‘Data Deficient’
status. Both species are possibly already exterminated in Peninsula Malaysia.
However, because the plants of these species are so minute they could still be extant
but overlooked in other wet habitats. Existing records are not specific in regard to
habitat preferences and could not be verified on-site recently, leaving them with
uncertain habitat preferences.
Conclusion
Utricularia is a carnivorous plant genus that grows in various wet habitats subjected
to seasonal water fluxes. From field observations and growth experiments, the
species are generally sensitive to changes in water nutrient content, pH, humidity,
the micro-organism community and microclimate. Therefore, where they might occur,
Utricularia species (especially those that fall within the habitat specialist group) could
serve as convenient indicator species for quick assessments of the health of a habitat.
The balance between habitat generalists and common pioneer species can help indicate
the level of disturbance or recovery of a habitat. Conservation monitoring efforts can
therefore be broadened from the species to the habitat perspective.
The initiatives of the state and federal governments of Peninsular Malaysia
in setting up and managing the respective parks and reserves deserve commendation.
In order to maximise the species conservation role of these protected areas while
allowing recreational use, the current management practices could be further refined
to cater for naturally occurring fringe species. As it stands, zonation is mostly pre-
determined by accessibility or remoteness. Strict conservation zones should be re-
aligned to capture different microhabitat types instead of conveniently using rivers
and ridges as boundaries and assigning all the open-vegetated microhabitats as
campsites. Periodic closure, especially in the wet seasons, is pivotal in preventing
irreversible transformation of inundated sites and erosion in steep areas. Continuous
documentation of the local flora is also essential in providing baseline information
so that any degradation to the plant community that is associated with substandard
amenity development or overuse can be pin-pointed.
ACKNOWLEDGEMENTS. The authors would like to thank the following institutions for
partial financial support for the project: Forest Research Institute Malaysia (FRIM) Research
and Pre-commercialisation Grant (GPP-TFBC-1208-001) and Flora of Peninsular Malaysia
project (01-04-01-0000 Khas 2); University of Malaya (UM), iPPP Postgraduate Research Grant
(PS169/2008B / PS235/2009C) and a Ministry of Natural Resource and Environment, RMK-9
Masters Scholarship. We are grateful to the curators and keepers of the herbaria of BKF, C, K,
KEP, KLU, L and SING, and the SEM and the anatomy laboratories of FRIM and the Institute
of Biological Sciences, UM. Last but not least, thanks to Dr. R. Kiew for co-supervision and
editing help, Dr. L.S.L. Chua for reviewing the conservation status assessments, and Dr. L.G.
Utricularia habitats and conservation 46]
Saw, Dr. R.C.K. Chung and Dr. E. Soepadmo for their advice and to the staff of the Forest
Biodiversity Division, FRIM, for assistance and support, especially in the field.
References
Brummit, R.K. (2007) Lentibulariaceae. In: Heywood, V.H., Brummit, R.K., Culham,
A. & Seberg, O. (eds) Flowering Plant Families of the World, pp. 190-191. Kew:
Royal Botanic Gardens.
Chew, M.Y., Rafidah, A.R., Haron, N.W. & Kiew, R. (2011) New findings from the
study of seed testa surface characters using SEM on Microchirita, Damrongia and
Utricularia section Phyllaria of Peninsular Malaysia. Malaysian J. Microscopy
7, In press.
Chua, L.S.L. & Saw, L.G. (2006) Malaysia Plant Red List—Guide for Contributors.
Kepong: Forest Research Institute Malaysia.
Givnish, T.J., Burkhardt, E.L., Happel, R.E., & Weintraub, J.D. (1984) Carnivory
in the Bromeliad Brocchinia reducta, with a cost/benefit model for the general
restriction of carnivorous plants to sunny, moist, nutrient-poor habitats. Amer.
Naturalist 124: 479-497.
Henderson, M.R. (1928) The flowering plants of Kuala Lumpur, in the Malay
Peninsula. Gard. Bull. Straits Settlem. 4: 295.
IUCN (2001) JUCN Red List Categories and Criteria: Version 3.1. Gland, Switzerland
& Cambridge, UK: IUCN Species Survival Commission.
Maseri, N.M., Mohd-Ros, A.H. & Suksuwan, S. (2006) Gunung Stong State Park
(GSSP), Kelantan Darul Naim, Preliminary Management Plan—Final version
for approval. Unpublished report. Malaysia: WWF-Malaysia, State Government
of Kelantan & Kelantan State Forestry Department.
Parnell, J.A.N. (2005) An account of the Lentibulariaceae of Thailand. Thai Forest
Bull., Bot. 33: 101-144.
Richards, J.H. (2001) Bladder function in Utricularia purpurea (Lentibulariaceae): Is
carnivory important? Amer. J. Bot. 88: 170-176.
Ridley, H.N. (1893) On the flora of the eastern coast of the Malay Peninsular. Trans.
Linn. Soc. London, Bot. 3(9): 327.
Ridley, H.N. (1895) Utriculariae of the Malay Peninsula. J. Bot. 33: 10-11.
Ridley, H.N. (1901) The flora of Mt. Ophir. J. Straits Branch Roy. Asiat. Soc. 35: 28.
Ridley, H.N. (1908) On the collection of plants made by H.C. Robinson and L. Wray
from Gunong Tahan, Pahang. J. Linn. Soc. Bot. 38(266): 317-318.
Ridley, H.N. (1923). Lentibulariaceae. The Flora of the Malay Peninsula 2: 490-495.
London: L. Reeve & Co. Ltd.
Rutishauser, R. & Isler, B. (2001) Developmental genetics and morphological evolution
of flowering plants, especially bladderworts (Utricularia): Fuzzy Arberian
morphology complements classical morphology. Ann. Bot. 88: 1173-1202.
Sirova, D., Borovec, J., Cerna, B., Rejmankova, E. & Adamec, L. (2009) Microbial
community development in the traps of aquatic Utricularia species. Aquatic Bot.
90: 129-136.
462 Gard. Bull. Singapore 63(1 & 2) 2011
Spare, G.H. (1941) Malayan bladderworts. Malayan Nat. J. 1: 86-90.
Suksathan, P. & Parnell, J.A.N. (2010) Three new species and two new records of
Utricularia L. (Lentibulariaceae) from Northern Thailand. Thai Forest Bull., Bot.
38: 23-32.
Suksuwan, S., & Ong, D.J. (2005). Management Effectiveness Assessment of National
and State Parks in Malaysia. Putrajaya: Ministry of Natural Resources and the
Environment, Malaysia.
Taylor, P. (1977) Lentibulariaceae. In: Steenis, C.G.G.J. van (ed) Flora Malesiana 8:
275-300.
Taylor, P. (1989) The Genus Utricularia—a Taxonomic Monograph. Kew Bull. Addit.
Ser. 14. Kew: Royal Botanic Gardens.
Turner, I.M. (1995). A catalogue of the vascular plants of Malaya. Gard. Bull.
Singapore 47: 1-757.
~ Appendix A. Habitat information on Utricularia species in Peninsular Malaysia. Notes: *No
Section Meionula
U. minutissima 0-2100 Wet, damp or rarely Exposed damp ground in lowlands
dry mud, silt, sand, and highlands; peat swamp edges;
laterite and shallow lowland heaths; highland sphagnum
soil overlying rocks bog; rockfaces of waterfalls or rapids;
[4-6(—7.5)] lowland river or stream banks;
lowland and highland mammal trails;
wayside turfs and fields; roadside
seepages; constructed wetlands;
recorded previously in rice fields and
old mining areas
Section Nigrescentes
U. caerulea 0—1400 Damp or wet Lowland and highland — heaths;
sand, laterite, silt highland stream banks; swamps;
and shallow soil recorded previously in wet
overlying rocks grasslands
[4.5—6 (-7)]
Section Oligocista
U. bifida 0—1100 Wet, damp or rarely Exposed damp ground in lowlands
dry mud, silt, sand, | and highlands; peat swamp edges;
laterite and shallow lowland heaths; lowland river or
soil overlying rocks stream banks; lowland mammal
[4-6(-7.5)] trails; wayside turfs and _ fields;
roadside seepages; —_ constructed
wetlands; recorded previously in rice
fields
Utricularia habitats and conservation
463
U. involvens 700—1200 Damp or wet mud Highland exposed to shaded stream
and shallow soil banks; highland exposed to shaded
overlying rocks wet rockfaces; wet grasslands;
[3.56] exposed to shaded springs and wells;
wet montane road-cut outcrops
U. scandens 300-1000 Damp or wet sand, Highland heath or wet grasslands on
mud and shallow rocky heaths
soil overlying rocks
[e. 5(-6)]
U. uliginosa 0—1000 Seasonally flooded, Exposed to shaded forest edges;
damp or wet mud, highland and lowland rocky stream
peat and sand [(3—) banks; peat swamps; highland
3.5—6.5 (-7)] sphagnum bog; highland mammal
trails; shallow pools; exposed to
shaded springs and wells
U. vitellina 1500-2100 Damp or wet Exposed to shaded montane mossy
bryophyte peat banks
mound [3.5—5]
Section Phyllaria
U. furcellata c. 1500 Gently sloping, Exposed heaths nestled within lower
damp (white) sand montane forest
and mud [c. 5]
U. striatula 150-2100 Damp, wet or Highland exposed to shaded, mossy
Section Setiscapella
dripping rock and
tree trunk [3.5—5.5]
earth banks; highland exposed to
shaded wet rock faces; waterfall or
rapid splash zones; tree trunks or
branches within splash zones; wet
montane road-cut outcrops
U. subulata* coastal Wet or damp sand Low open marshes; ditches; stream
plain and shallow soil and pool sides’
overlying rocks’
[n.a.]
Section Nelipus
U. limosa* coastal Wet sand and mud! = Open country wet spots’; swamps;
plain [n.a.] pool margins; lowland — shallow
waters
Section Utricularia
464
U. aurea
U. gibba 0—1300
U. punctata 0-250
Gard. Bull. Singapore 63(1 & 2) 2011
Often edges of still
or slow flowing
water, rarely wet or
damp mud and silt
[3—7(-7.5)]}
Often shallow and
sometimes deep,
still or slow flowing
water with low pH,
sometimes wet or
damp mud and silt
[3-5.5(-7)]
Often edges of still
or slow flowing
black water with low
Peat swamp edges; river backwaters;
open wetlands; puddles in open
fields; ge/am swamp forests; tidal
typha reed beds; natural or man-
made lakes, ponds, reservoirs,
dams and depressions; old oil palm
estate, village or wayside ditches
and canals; abandoned rice-field
patches or organic rice-fields; ex-
mining ponds; constructed wetlands;
recorded previously in highland
catchment pond
Peat swamp edges; river backwaters;
open wetlands; tidal typha reed
beds; natural or man-made lakes and
ponds; ditches and canals bordering
peat swamps or forests; abandoned
rice-field patches or organic rice-
fields; constructed wetlands; tanks;
recorded previously in highland
catchment pond
Various niches at the edge of natural
lakes; recorded previously in river
backwaters
pH, rarely wet or
damp mud and silt
[3-5]
Appendix B. Taxonomic notes on the Utricularia furcellata specimen of Peninsular Malaysia.
Vegetatively, Utricularia furcellata is very similar to U. striatula from the same section
U. striatula
More-or-less 4-lobed, lateral lobes
much smaller than the apical pair
Lower corolla lip More-or-less regularly 5-lobed
Average seed length 455 tum (N = 15) 251 um (N = 15)
Relatively smooth with verrucae
only along the sinuate boundaries
Periclinal testa cell- Densely covered with globose or
walls shortly clavate verrucae
Sparsely papillate, processes short
with glochidiate, stellate tips
Papillae Densely papillate, processes long
with knobbly, clavate tips
Note: the lower corolla lip of the Kelantan specimen (Chew et al. FRI53603) is only shallowly
lobed compared to the plants recently recorded from Northern Thailand (Suksathan & Parnell
2010), but the flower and inflorescence dimensions and descriptions match that of Taylor
(1989).
Gardens’ Bulletin Singapore 63(1 & 2): 465-470. 2011 465
Establishment of Enrekang Botanic Garden,
South Sulawesi: an effort to conserve plant diversity
in the Wallacea region
Julisasi T. Hadiah
Pusat Konservasi Tumbuhan, Kebun Raya Bogor
Jl. Ir. H. Juanda 13, Bogor 16122, Indonesia
jhadiah@yahoo.com
ABSTRACT. The Enrekang Botanic Garden is newly established in Kabupaten Enrekang,
South Sulawesi. Indonesia, to document and conserve the diversity of plants from the Wallacea
Region. The new botanic garden is a collaborative venture since 2005, in which Kebun Raya
Bogor is assisting the local authorities of Sulawesi Selatan to form a development plan and
establish the garden’s living collections. The garden has developed nursery facilities, an
irrigation system, and road access. A total of 4601 living plants have been planted, representing
36 families, 156 genera and 232 species.
Keywords. Enrekang Botanic Garden, Indonesia, plant diversity, Sulawesi, Wallacea
Establishment of new botanic gardens in Indonesia
The establishment of new botanic gardens in Indonesia is in line with Article 9 of the
Convention on Biological Diversity, on Ex Situ Conservation (CBD 1993): “establish
and maintain facilities for ex sitw conservation of and research on plants, animals
and micro-organism, preferably in the country of origin of genetic resources”; and
Target 8 of the Global Strategy for Plant Conservation (GSPC 2002): “sixty per cent
of threatened plant species in accessible ex situ collections preferably in the country
of origin, and ten per cent of them included in recovery and restoration programmes”.
Indonesia has committed to international conventions such as the CBD (1993),
Agenda 21 in 1992 (United Nations 2009) and GSPC (2002), and as part of this
commitment has been establishing new botanic gardens as ex sifu conservation sites.
This activity was also included in Agenda 21 Indonesia (United Nations Development
Programme & Indonesia Kantor Menteri Negara Lingkungan Hidup 1997). The
establishment of new botanic gardens in every province in Indonesia will benefit not
only the local community, but also represent national and international needs; for
example, the conservation of plant species, soil and water by a botanic garden will
not only benefit the local community but also a much wider area (Sutrisno 2010). A
botanic garden also serves to inspire a wider community to respect plants and their
wise use.
The establishment of new botanic gardens in Indonesia was initiated by
the then Indonesian President, Megawati Sukarno Putri, during “Hari Kebangkitan
466 Gard. Bull. Singapore 63(1 & 2) 2011
Teknologi Nasional” (the National Technology Awareness Day) on 11 August 2004,
when she emphasised the importance of establishing a botanic garden in every province
in Indonesia. The Minister of Research and Technology followed up on the President’s
speech by issuing a letter to all Governors in Indonesia (Surat Edaran Menteri Riset dan
Teknologi kepada seluruh Gubernur di Indonesia No. 77/M/VIII/2004). The Minister
suggested the Governors establish at least one botanic garden in each province, and to
collaborate with the Indonesian Institute of Sciences (LIPI) to undertake the task. As
a result, many proposals were sent to LIPI to establish botanic gardens. Consequently,
from 2004 until the end of 2009, 16 Local Authorities had started to establish new
botanic gardens. One of them is Enrekang (Fig. 1).
Kuntngan Lombok Timur
fe comp 2067 by Wor Trae rows Ad Aight Banervod -
Fig. 1. Location of new botanic gardens (white circles) and the four long-established botanic
gardens of Kebun Raya Indonesia (black squares) in Indonesia. (After Sutrisno 2010)
Wallacea
The Wallacea region is one of the world’s biodiversity hotspots situated east of the
Wallace line—an imaginary line dividing Indonesia into two floral regions: an Asian
floristic region to the west and an Australian floristic region to the east. Wallacea is
somewhat transitional between the two flora regions, and contains floristic elements
from both regions. The area covers the islands of Sulawesi, Lombok, Lesser Sunda
(including Timor Leste) and the Moluccas (see Fig. 2).
Wallacea has a high extent of floral and faunal endemism, although in general,
information on the flora from the region is still lacking. However, it is estimated that c.
10,000 species of vascular plant grow in the region, of which 1500 species (15%) are
endemic (Conservation International 2007).
Sulawesi is the largest island in the region, covering about 53% of the
area, with c. 500 endemic species of plants. Although many new species have been
discovered from the island since several decades past, the island’s plant diversity is
still poorly known as the number of specimens collected per 100 km/? is still quite
low, i.e., 23, whereas, ideally, an adequate representative of collections would be 100
Enrekang Botanic Garden in South Sulawesi 467
Fig. 2. The Wallacea region: a transition zone from Australia to Asia.
specimens per 100 km?’ (Whitten et al. 2002). Moreover, this plant diversity needs to
be represented in ex situ living collections in botanic gardens.
Sulawesi’s plant diversity is facing serious threats, as also happens in the region
generally, due to habitat loss, deforestation, plantation development and farming. The
establishment of a new botanic in Kabupaten Enrekang, South Sulawesi, is therefore
timely. The new garden is named Kebun Raya Enrekang.
Establishment of Kebun Raya Enrekang
In August-September 2005, the Local Auhorities of Enrekang in collaboration with
Kebun Raya Bogor (KRB) and The National Survey and Mapping Coordination
Agency (Bakosurtanal) surveyed and mapped some prospective localities as potential
sites for the new Kebun Raya Enrekang (KRE). Following the survey, an MoU was
signed by LIPI and the Enrekang Local Authorty for the establishment of KRE. At
the same time, a master plan of the new garden was completed (PT. Tata Guna Patria
2006). In April 2006 a team from KRB worked together with some local staff to clear
the site and started a nursery. A year later, some of the plant collections were officially
planted in the garden on 14 March 2007 by VIPs from both Enrekang and LIPI. This
first planting is regarded as the establishment date of the new garden.
KRE is located on the main Trans Sulawesi road between Makassar and
Tana Toraja. It is situated at the Batu Mila Village, Kecamatan Meiwa, Kabupaten
468 Gard. Bull. Singapore 63(1 & 2) 2011
Enrekang, Province of Sulawesi Selatan (3°33°47.58”S 119°45°40.56”E). Covering
300 ha in area with an elevation range of 70-155 m asl, the new garden is administered
by the local authority Dinas Kehutanan dan Perkebunan. Since the development of
KRE in 2006, the number of plant collections (excluding the nursery’s collection and
orchids) is 4601 accessions, comprising 36 families, 156 genera and 232 species; some
of which are listed in Table 1. The new garden has 18 employees, six of whom have
attended a training course on managing botanic gardens at the Kebun Raya Bogor.
There 1s a building that functions as a temporary management office as well as a short-
term accommodation for employees, a 3-km road access, nursery facilities including
awning and shading, and an irrigation system, fencing around the garden that also
keeps free-ranging cattle out, a motorbike, and two gazebos built by teams from KRB.
Table 1. Some plants already established in the KRE living collections.
Plant name Family
Aleurites moluccana (L.) Willd. Euphorbiaceae
Borassus flabellifer L. Arecaceae
Cinnamomum celebicum Miq. Lauraceae
Cyrtostachys microcarpa Burret Arecaceae
Diospyros blancoi A.DC. Ebenaceae
Diospyros celebica Bakh. Ebenaceae
Eucalyptus deglupta Blume Myrtaceae
Mimusops elengi L. Sapotaceae
Myristica lancifolia Potr. Myristicaceae
Neonauclea celebica Mert. Rubiaceae
Pangium edule Reinw. Achariaceae
Parkia timoriana Merr. Fabaceae
Pigafetta elata (Mart.) H.Wendl. Arecaceae
Pterospermum celebicum Miq. Malvaceae
Sandoricum borneense Miq. Meliaceae
Sapindus rarak DC. Sapindaceae
Schleichera oleosa (Lour.) Oken Sapindaceae
Syzygium malaccense (L.) Merr. & L.M.Perry Myrtaceae
Syzygium zeylanicum (L.) DC. Myrtaceae
Timonius stipulosus Valeton Rubiaceae
Vatica pauciflora Blume Dipterocarpaceae
Vitex cofassus Reinw. ex Blume Lamiaceae
Enrekang Botanic Garden in South Sulawesi 469
KRE focuses on collecting plant species native to Wallacea. Some examples
include the palms Pigafetta and Borassus flabelliber, the tree genus Agathis, the
gum tree Eucalyptus deglupta, the rosewood Pterocarpus indicus and the fragrant
sandalwood Santalum album. The collecting sites have so far included the forests
surrounding Enrekang to start with. Also, seeds have been brought from Kebun Raya
Bogor.
There are some themed displays and other features to be built at KRE (see
Master Plan in Fig. 3), namely, Taman Wangi (aromatic plant collections), Formal
Garden, Amphitheatre, Landscaped Corridor, Herb Garden, Aquatic Plant Collections,
Taman Mexico (succulent collections), Orchidarium, Pandan Garden, Palm Garden
and Flower Park.
LANDSCAPE PLAN
LEGEND
ial BUILDING PLAN
A. TOSERBA FLORA (Botanic Shop)
F. INFORMATION CENTRE
G, ADMINISTRATION OFFICES
H. RESORT 1
1. RESORT 2
J, EXERCISE AREA
K. OFFICIAL RESIDENTIAL HOUSE
L GREEN HOUSE
M. LABORATORIUM
AROMATIC GARDEN
FORMAL GARDEN
AMPHITEATER
LANDSCAPE CORRIDOR (Borossus
flabeliifer)
HERB GARDEN
AQUATIC PLANT COLLECTION
'WALLACEA CORRIDOR
WALLACEA PARK
ORCHIDARIUM
EDUCATION PARK & PANDAN
GARDEN
ORNAMENTAL PARK
1 PALM GARDEN
FORMAL PLAZA (FLOWER PARK)
SITE ENGINEER
Fig. 3. Landscape plan of Kebun Raya Enrekang (taken from the Master Plan of Kebun Raya
Enrekang, PT. Tata Guna Patria 2006).
Closing remarks
Establishment of a new botanic garden in each province of the vast Indonesian
Archipelago is crucial as forests become more degraded and pressures to natural
habitats increase due to the country’s high population. The establishment of Kebun
Raya Enrekang plays an important role in the conservation of plant diversity,
particularly in Wallacea.
470 Gard. Bull. Singapore 63(1 & 2) 2011
ACKNOWLEDGEMENTS. I would like to thank Helen Stevenson and Barry Conn (both NSW)
for preparing the poster for presentation at the 8th Flora Malesiana Symposium; Catherine
Wardrop and Julia Siderus (both NSW) for preparing figures | and 2. I also acknowledge
Pak Mursalim, Pak Zainal, Bu Harni, Bu Hasna, Bu Cica and colleagues of KRE for their
considerable efforts to establish Kebun Raya Enrekang.
References
CBD, Convention on Biological Diversity (1993) Article 9 Ex Situ Conservation.
http://www.cbd.int/convention/articles.shtml?a=cbd-09 (accessed 4 Sep 2010).
Conservation International (2007) Biodiversity Hotspots: Wallacea. http://www.
biodiversityhotspots.org/xp/hotspots/wallacea/pages/default.aspx (accessed 2
Sep 2010).
GSPC (2002) Global Strategy for Plant Conservation. Convention on Biological
Diversity & Botanic Gardens Conservation International. http://www.cbd.int/
gspc/ (accessed 2 Sep 2010).
PT. Tata Guna Patria (2006) Master Plan Pembangunan Kebun Raya Enrekang.
Unpublished.
Sutrisno (2010) Membangun Kebun Raya Baru di Indonesia. Warta Kebun Raya
10(1): 3-12
United Nations (2009) Agenda 21. UN Department of Economic and Social Affairs
Division for Sustainable Development. http://www.un.org/esa/dsd/agenda?2 1/
(accessed 4 Sep 2010).
United Nations Development Programme & Indonesia Kantor Menteri Negara
Lingkungan Hidup (1997) Agenda 21 — Indonesia: A National Strategy for
Sustainable Development. Jakarta: State Ministry for Environment, Republic of
Indonesia: United Nations Development Programme.
Whitten, T., Henderson, G.S. & Mustafa, M. (2002) The Ecology of Indonesia, vol. IV,
The Ecology of Sulawesi. Singapore: Periplus Editions (HK) Ltd.
ae
Gardens’ Bulletin Singapore 63(1 & 2): 471-483. 2011 471
Comparative pollen morphology
of three Alternanthera species (Amaranthaceae)
E.H.S. Chin! and A.L. Lim
Institute of Biological Sciences,
Faculty of Science, University of Malaya,
50603 Kuala Lumpur, Malaysia
'evansee_82(@yahoo.co.uk
ABSTRACT. The pollen morphology of three Alternanthera species, A. sessilis (L.) R.Br. ex
DC., A. bettzickiana (Regel) G.Nicholson and A. paronychioides A.St.-Hil., is reported. Pollen
grains of Alternanthera are dodecahedric, isopolar and small (12.56—23.57um). The pollen
morphology of the green and red varieties of A. sessilis shows no significant difference in the
polar length [t (98) =-1.35, p > 0.05] and equatorial diameter [t (98) = 1.32, p > 0.05]. Apertures
of A. sessilis and A. bettzickiana are pantoporate with twelve round pores, whereas the pollen
grains of A. paronychioides distinctly differ from the other two species in having eighteen oval
pores. The pores of all the species are covered by rectangular, sinuous, or elongated ektexinous
bodies. The sexine is metareticulate and tectum perforate with unevenly distributed perforations
at the top and base of the mesoporia, except in the pollen grains of A. bettzickiana, in which the
perforations are distributed unevenly at the top of the mesoporia only.
Keywords. Alternanthera, Amaranthaceae, pollen morphology
Introduction
Alternanthera Forssk. (Amaranthaceae) consists of 80 species distributed in the tropics
and subtropics (Mabberley 2008). Of these, palynological data of 20 species have
been reported (Table 1). The majority of these species originated from the New World
(Borsch 1998, Eliasson 1988, Miiller & Borsch 2005b, Nowicke & Skvarla 1979),
India (Kajale 1940, Mittre 1962, Rao & Shukla 1975), Pakistan (Bashir & Khan 2003,
Perveen & Qaiser 2002) and China (Li et al. 1993, Liang et al. 1978). Therefore, it
would be interesting to study the pollen morphology of the A/ternanthera species in
Malesia particularly in Peninsular Malaysia, and compare these observations with the
earlier data.
This is especially meaningful as in recent years, detailed study of pollen
morphology has helped to resolve the relationship among members of the
Amaranthaceae (Borsch 1998, Miiller & Borsch 2005b). Palynological data has also
supported the findings in phylogenetic studies of the subfamily Gomphrenoideae
(Sanchez del-Pino et al. 2009, Miiller & Borsch 2005a).
Though Erdtman (1966) has reported that the pollen grains of A/ternanthera
species possess deeply recessed pores, the apertures are not well studied. The pollen
morphology of A/ternanthera is further defined by Borsch (1998) and described as the
472 Gard. Bull. Singapore 63(1 & 2) 2011
Table 1. References to published palynological studies on Alternanthera species.
Species References
4. albida Griseb. Borsch 1998
4. bettzickiana (Regel) G.Nicholson Li et al: 1993
A. caracasana Kunth Miller & Borsch 2005b
4. costaricensis Kuntze Borsch 1998
A. filifolia (Hook.f.) Howell J.T. Howell Borsch 1998
A. flavescens Kunth Borsch 1998,
Nowicke & Skvarla 1979
A. galapagensis (Stewart) J.T. Howell Eliasson 1988
A. geniculata Urb. Eliasson 1988
A. gracilis Loes. Erdtman 1966
A. maritima (Mart.) A.St.-Hil. Borsch 1998
A. nesiotes 1.M.Johnst. Eliasson 1988
A. nodiflora R.Br Liet al. 1993
A. olivacea Urb. Borsch 1998
A. paronychioides A.St.-Hil. Perveen & Qaiser 2002
A. peruviana (Moq.) Suess. Borsch 1998, Eliasson 1988
A. philoxeroides Griseb. Li et al. 1993
A. pungens Kunth Bashir & Khan 2003
A. reineckii Brig. Eliasson 1988
A. repens Steud. Mittre 1962
A. sessilis (L.) R.Br. ex DC. Kajale 1940, Mittre 1962,
Rao & Shukla 1975,
Li et al. 1993,
Liang et al. 1978,
Perveen & Qaiser 2002
Pfaffia type. The pollen grains 1n this category are dodecahedric, metareticulate, tectate
perforate with microspines distally arranged in a line. The apertures are pantoporate
with pore structure of Type I (Borsch 1998). The diameter of the pores is 3—6 um and
is covered by 20-60 ektexinous bodies. The length of these ektexinous bodies is 1.5—
3.0 times its width. These ektexinous bodies are rectangular, sinuous or elongated and
are arranged in a mosaic-like pattern, closely adjoined but separated from each other.
Each of the ektexinous bodies has one to four distinct microspines attached onto it.
Studies of A/ternanthera in Malaysia date back to the 18th century when
Ridley (1924) reported A. sessilis in the Malay Peninsula. Alternanthera sessilis
(L.) R.Br. ex DC., A. sessilis var tenuissima (Suess.) Backer, A. repens Kuntze, A.
bettzickiana (Regel) G.Nicholson, A. philoxeroides Griseb., A. brasiliana (L.) Kuntze
Pollen morphology in A/ternanthera 473
and A. porrigens Kuntze from Malesia were subsequently described (Backer 1949).
Alternanthera triandra Lamk. has been recorded as a common weed in Malaya.
However, in Turner’s catalogue of Malayan Plants (1995), only three species (A.
sessilis, A. philoxeroides and A. bettzickiana) were recognised. Of these, the pollen
grains of three species were investigated in the present study, i.e., 4. sessilis (both the
red variety as well as the green, the later sometimes considered as a distinct species, A.
triandra Lamk.), A. bettzickiana and A. paronychioides.
In Peninsular Malaysia, A. sessilis occurs in two leaf colours; green and red.
Both the green and red varieties of A. sessilis are perennial creeping herbs. The red
variety is commonly cultivated, whereas the green variety is a common weed found
growing in various types of habitat from moist areas near drains, ditches and ponds to
dry open wastelands. Similarly, 4. bettzickiana is usually found in dry open wastelands
or moist areas near ditches. However, these two species can be easily identified by the
morphology of stem, leaf and flower (Table 2). Lastly, A. paronychioides could be
found on dry sandy ground such as along canals or roads. This species forms a dense
mat with numerous prostrate branches with roots at the nodes.
Alternanthera sessilis is regarded by some communities as a useful medicinal
plant indigenous to Malesia. The red variety is usually used as a herb to treat
heart disease and hypertension. Scientifically, A. sessilis has been reported to have
antibacterial activities (Kumaresan et al. 2001, Jalalpure et al. 2008). Free radical
scavenging properties have also been found in A. sessilis indicating the presence of
antioxidant properties in the plant (Balasuriya & Dharmaratne 2007, Bhaskar et al.
2007, Shyamala et al. 2004). Other reported pharmacological activities of A. sessilis
include reduction of hypertension (Goh et al. 1995), having an hepatoprotective effect
against liver injuries induced by hepatotoxins (Song et al. 2006) as well as functioning
as a diuretic (Goh et al. 1995). As the two varieties of A. sessilis are similar except
in leaf colour, this study also aims to elucidate their pollen morphology to determine
whether they differ.
Materials and methods
Pollen samples for the green and red varieties of A. sessilis were collected from Varsity
Lake and the Rimba IImu Botanic Garden, respectively, in the University of Malaya,
Kuala Lumpur. The pollen grains of A. bettzickiana were collected from the Forest
Research Institute Malaysia (FRIM) and 4. paronychioides from the Kuala Selangor
Nature Park, Selangor, Peninsular Malaysia. Voucher specimens were deposited in the
herbarium of the Institute of Biological Sciences, University of Malaya (KLU).
Anthers of 15 flowers of each species were collected after anthesis. The
pollen samples were then acetolysed according to the standard procedure of Erdtman
(1960). Slides for light microscopy (LM) study were prepared with glycerine jelly and
sealed with wax. The measurements of equatorial diameter (E) and polar length (P)
were taken with a Leica DM100 microscope. Then, the pollen shape and size were
determined. Acetolysed pollen grains for scanning electron microscopy (SEM) study
474 Gard. Bull. Singapore 63(1 & 2) 2011
Table 2. Morphological characters of Alternathera sessilis (green and red varieties), A.
paronychioides and A. bettzickiana.
Species
Colour, habit and selected morphological traits
Stem Foliage Flowers
colour _ petals stamens ovary
, reen;
green tinged e
simple;
with purple i
7 opposite;
A. sessilis at the node; creamy 5, subequal
: narrowly , < 5
(green variety) ascending, AL eto white in size compressed
creeping or P
narrowly
decumbent
oblong
red; simple;
red; opposite;
Me decumbent _ variable,
A. sessilis 5, subequal
or creeping mainly red et 3
(red variety) in size compressed
narrowly
elliptic or
oblanceolate
green;
reen; simple; bee
e De 5, distinctly
es prostrate and opposite; creamy
A. paronychioides (. _. unequal 2)
forming a thick cuticle; white are compressed
mat elliptic or
obovate
green tinged
green; ee,
with purple canis vate 5, distinctly
we ; ¢c
A. bettzickiana at the node; ee oie unequal in 5) subconical
opposite; white
erect or ahs size
elliptic
ascending
were coated with gold, examined and photographed using a JOEL JSM-6400 scanning
electron microscope.
Terminology used generally follows that of Fegri & Iversen (1950), Erdtman
(1952), Borsch (1998) and Borsch & Barthlott (1998). An independent samples t-test
was conducted to compare the polar length, equatorial diameter, diameter of the pores
and number of ektexinous bodies between the following: 1) the green and red varieties
of A. sessilis. 2) the green variety of A. sessilis and A. bettzickiana. 3) the red variety
of A. sessilis and A. bettzickiana. The analyses were performed using the SPSS version
11.5 for Windows.
Pollen morphology in Alternanthera 475
Results
The SEM photomicrographs of pollen in polar view, equatorial view, and pore structure
are shown in Fig. 1. The pollen grains of Alternanthera species are dodecahedric,
isopolar and radially symmetrical (Table 3). The pollen grains are small (12.56—23.57
um in length) with the average polar axis varying between 15.19 and 21.36 um and the
average equatorial axis 17.18—21.85 um. The t-test (Table 4) indicated there was no
significant difference in polar length [t (98) =-1.35, p > 0.05] and equatorial diameter [t
(98) = 1.32, p > 0.05] of the green and red varieties of A. sessilis. For the green variety
of A. sessilis and A. bettzickiana, the t-test indicated significant difference in the polar
length [t (86.58) = 3.12, p< 0.05], but no significant difference in equatorial diameter
[t (98) = 1.76, p> 0.05]. Similarly, the red variety of A. sessilis and A. bettzickiana
are significantly different in polar length [t (98) = 5.24, p< 0.05] but not in equatorial
diameter [t (98) = 0.46, p> 0.05].
Under the scanning electron microscope, the apertures of A. sessilis and A.
bettzickiana are pantoporate with twelve round pores, whereas pollen grains of A.
paronychioides have approximately eighteen oval pores. Each of the pores is situated
in a pentagonal face which contributes to the dodecahedric body of the pollen grain.
The green variety of A. sessilis possesses the smallest pore (4.17 um), followed by the
red variety of A. sessilis (4.48 um), A. bettzickiana (4.83 um) and A. paronychioides
(6.06 um). The result of the t-test shows that there is a significant difference in the pore
diameter of both green and red varieties of A. sessilis [t (58) = -3.86, p < 0.05].
The average number of ektexinous bodies in the red and green varieties of
A. sessilis is the lowest, 28 and 29 respectively, followed by A. bettzickiana (33) and
A. paronychioides (36). The number seems to be related to the size of the pores. For
instance, the pollen grains of A. paronychioides have the biggest pores (6.06 um) and
highest average number of ektexinous bodies (36). The t-test shows that there is no
significant difference in the number of ektexinous bodies of the green and red varieties
of A. sessilis [t (17.82) = 1.67, p > 0.05]. On the other hand, both varieties of 4. sessilis
and A. bettzickiana are significantly different in the number of ektexinous bodies, 1.e.,
the green variety of A. sessilis and A. bettzickiana [t (26) = -2.42, p< 0.05], and the red
variety of A. sessilis and A. bettzickiana [t (16.86) = -4.53, p< 0.05]. The length of the
ektexinous bodies is 4—6 times its width. They are rectangular, sinuous or elongated
and each of the ektexinous bodies has 3-4 distinct microspines attached onto it. These
ektexinous bodies are arranged in a mosaic-like pattern, closely adjoined but separated
from each other.
The sexine of the pollen grain is metareticulate with a row of microspines
that are regularly and distally arranged. These microspines are cylindrically elongated
and blunt. The average height of these microspines ranges from 0.22—0.45 um. The
pollen grains of all the species studied are tectum perforate with unevenly distributed
perforations at the top and base of the mesoporia, except in the pollen grains of A.
bettzickiana, in which the perforations are distributed unevenly at the top of the
mesoporia only.
476 Gard. Bull. Singapore 63(1 & 2) 2011
im
*x6,.866
8382 16KY
Fig. 1. SEM photomicrographs of pollen in polar view, equatorial view, and pore structure.
A-C. A. sessilis (green variety); D-F. A. sessilis (red variety); G—-I. A. paronychioides and J—L.
A. bettzickiana.
477
Pollen morphology in Alternanthera
euodosau doy
AyuoAoun
Ip
20
(Sh'0-) 60'0 ¥ 670 COLO)
UA[NOoIByOUU
(9¢°0-) €O'0 F 1£°0 (970)
(Lv'l-) 00'T F 97 T CZ)
(6S'€-) 970 ¥ bl'€ (-b8'7)
(00'0b-) OLE F 1L'7E (-00°L7)
su]
(Z1'S-) O70 €8'b CIS'b)
cl
ojesodojued
|yetus
an)
(L6'17-) OF I FEULI (C 8L'b1)
(POS I>) P80 F 6L'S1 (-90'PL)
UdIpPSyRoapop
eououruds Ayperpes
Ivjodos]
DUDIYOIZIOG “WV
m
al
vLodosour ajoyM
Ajuoaoun
Ip
00
u (6S'0-) b0'0 ¥ SP’0 ETO)
wuNpNoHoreyou
(SE'0-) b0'0 ¥ 1€°0 L770)
(PS 1-) 610 F PLT (-€0'7)
(Zo'b") $90 FRE'b CLT’)
WU (00'6E-) OS'T F OR'SE (-00'7E)
Lie |
u (9L'9-) €8°0 ¥ 90°9 GPP Pb)
SI
oyerodojyurd
[pews
860
U(€S°ET-) SOT FE8'IT (- OF'07)
UU (LS'EZ-) 00' 1 FOE 17 (- Z1'07)
UOApayRoopop
[POLoWUAS jou
Injodos]
saplolyoAuodnd *y
x
¢=u
gau
os =u
u
Blodosout ojoym
Ajuoaoun
Ip
09
(PE'0-) €0'0 F 770 (S10)
LUN|NOMOIBPOU
(6¢'0-) €0'0 * €€'0 (-67'0)
(6r'I-) Z7V0 F 871-0
(OL'€-) LUO F €7'E (187)
(00°0€-) 8h 1 + 6L'LZ (-00'SZ)
Shay
(bI'S-) 87:0 # 8b'b CII b)
a
aesodojued
jjeuus
88'0
(99°61-) 78:0 F 69 LI 6091)
(QUS8I-) S60 F8P SI CLO EL)
Uorpoyeoopop
Jeonjouuds Ayyperpea
Ivjodos]
PRI
Of
Ot
OS
OS
BLodosout sjoyM
uoAaoun
Ip
99
u (Ob'0-) £0'0 £ €7'0 CEL)
wunypNonoreyou
(8€'0-) SO’; * 97'0 (-07'0)
(SEI O10 F STI CEO
(8L°€-) PEO F €Z7'€ (-ZS$'7)
U (00'S E-) SE F OF'67 (-00' EZ)
U (C8'b-) SEO FLI'b (-8S'€)
a
oerodojurd
[jets
S80
UU (L6'61-) COL F POLI (- LISD)
UU (FO 81>) OT 1 FGT'ST (- 99'71)
UOIpoyRoapop
[eoLnouruwAs Aypeipes
Injodosy
uodddry
SIJISSAS “V
UONNqLSIp
JuoWOsUR.LIe
odeys
Og =u (wun )ysioy
(wud pian
(wun)y suo]
O'S
pl =u Joquinu
odeys
qo
Og =U (WN )JO}OUIRIP
Joquinu
0S =u
0g=u
UONLIOI.AD
souds
OUIXOS
10d
ainjiody
sseo
d
(wi)
(win)
aZIS
odeys
AnjouwudAg
AVRO
s19joRsey)
saioadg
‘SaIpog SNOUTXd}J9 dy} JO YoRa UO poyorye sourdsosorUU
“o's ssnonuls “s caR[NSuLIOOI “I “JOJOUIVIP [eLOJeNbs pure yIsuo] Avjod usdajoq ones “A/q SyPSud] svjod ‘gq Susoyed oresow ‘WI foyeBUOIA ‘| Ssarpoq
SNOUTXO}YO “qo “Ja}oUvIP [eLOJwNba “y sour] ev UL posure AjJejNdo1 pur A]peysIp “Ip <poyesuoja AjjeouputpAd ‘99 ‘poyesedos nq pourolpe A]asoyo
‘OCWAS pur WT Jopun puppjoizyog “y pur saplolyrAuodbd ‘y ‘(SONORA Pot PUL UdDIB) si/ISSas “p JO SIdJeALYO [eoIBojoydiow! Udo *¢ IGRI.
Gard. Bull. Singapore 63(1 & 2) 2011
478
SaIpog snourxa}ya
DUDIYIIZJAG “VW
pue
SIJISSAS “PV
pue
SIJISSAS “PV
SIJISSAS “PV
he) Ge COiL- 60'1 €60- 00°0 OS es Va 00°0 89°01 gO JoquInN,
6 0 vc 0- 910 LO'0 S90 00°86 86940 c8'0 S00 (um) q (AJOLIBA po.)
8c | 80 810 £60 00°0 0086 cS LV0 €s'0 (wim) d
SOIPOG SNOUIX9}49 = PUDIYIIZJAq ‘¥
OS 0- L0'9- Ure OCits c0'0 NONE ORG vs 0 6t 0 gO Sq uN
89°0 70'0- 810 ce 0 80:0 00°86 = =9L'I tc 0 sv (um) q — (Ajarea uals)
bO'l ec 0 0c 0 p90 00°0 8598 “Cle cO'0 vss (wim) d
SdIpog snourIxayyo
Wins cr O- 860 y9'l ITO cs LI LOT cO'0 v9 JO J9quInN|
S05 87 0- 80:0 CEOs 00°0 OO'8S Y98'E- 0c 0 69] (win) JajauIeIp 10g
19°0 cl 0- 610 AY) 610 00°86 =e ce 0 S60 (win) q —- SANOLIBA OG
vI0O cL 0- CC 0 6C 0- 810 0086 St I- 10 ENG (wil) d
Joddq) 1aMmo7
-aouasaffip . (papi -Z) ia aa - 7 " i
ay) JO ]DAsaJUI aUaaffip —_ aouadaffip AN]DA anda
aUapluo) % C6 AOAAD ‘PIS uvoul AOUDIYIUBIS JP ] JIUDIYIUSIS A
soIpog snourxazya Jo Joquinu pue
suvaw Jo Ayypenba 10J 4$9}-}
soouvlieA Jo Ayiyenba
JOJ 1S9] $,9UdAIT]
a]qeuea juspuadaq
saroeds
‘DUDIYIIZ]]JAqG “PY PUR SIJISSAS “f UIIMJOQ
Jajoureip oiod ‘(q) Jojoueip jeoyenbs ‘(q) ysuea] Jejod ay} Ssulsedwu0d jsaq-} sajdwies yuspuadoapuy] “pf aqeuy
Pollen morphology in Alternanthera 479
The pollen morphology of the green and red varieties of A. sessilis is remarkably
similar despite significant differences in their pore diameter. This difference alone is
not sufficient to delimit them into two species. Furthermore, the pollen grains of these
two varieties show no significant difference in polar length, equatorial diameter and
number of ektexinous bodies.
Interestingly, the present study has shown that pollen grains of 4. bettzickiana
are almost identical to those of A. sessilis. Still, the pollen grains of these two species
could be differentiated by significant difference in their polar length and number of
ektexinous bodies. The pollen of 4. bettzickiana is longer (A. bettzickiana: 15.79 um;
green and red varieties of A. sessilis: 15.19 um and 15.48 um, respectively) and the
pore is covered by more ektexinous bodies (A. bettzickiana: 33; green and red varieties
of A. sessilis: 29 and 28, respectively). Another character that differentiates these two
species is the distribution of perforations. The perforations in 4. bettzickiana are at
the top of the mesoporia only whereas those from A. sessilis are around the whole
mesoporia.
Clearly, the pollen of 4. paronychioides is very different from the pollen of A.
sessilis and A. bettzickiana in having around 18 oval pores while the other two species
have only 12 round pores. The t-test could not be carried out to compare the polar
length, pore diameter and number of ektexinous bodies between 4. paronychioides
and the other two species due to the difficulty in obtaining sufficient number of pollen
grains from A. paronychioides.
Discussion
On the whole, pollen grains of the three species studied conform well to the Gomphrena-
type of Erdtman (1966) which corresponds to the Pfaffia type of Borsch (1998).
The pore structure is similar to the Type I of Borsch (1998). Similar observations
have been reported in the pollen grains from the New World (Borsch 1998, Eliasson
1988, Miiller & Borsch 2005b, Nowicke & Skvarla 1979) and China (Li et al. 1993,
Liang et al. 1978) (Table 1). However, a few exceptions have been reported, 1.e., A.
philoxeroides (Li et al. 1993) and A. costaricensis Kuntze (Borsch 1998) are reported
to have spheroidal pollen grains.
The present palynological results support existing data which have indicated
that A/ternanthera is stenopalynous in terms of pore number. Most of the species
examined (Borsch 1998, Eliasson 1988, Li et al. 1993, Liang et al. 1978, Miiller &
Borsch 2005b, Nowicke & Skvarla 1979), including A. sessilis and A. bettzickiana
in the present study, have 12-14 pores. Only a few exceptions were identified, such
as 20-24 pores in A. philoxeroides (Li et al. 1993), 25-30 in A. costaricensis (Borsch
1998) and 18 in A. paronychioides (present study).
Compared with previous studies, the pollen grains of A. sessilis in the present
study are distinctly different from those in Pakistan and India. For instance, the number
of pores was reported as six in the grains of A. sessilis from the Upper Gangetic plain
480 Gard. Bull. Singapore 63(1 & 2) 2011
(Rao & Shukla 1975) and 3—3.2 from Pakistan (Perveen & Qaiser 2002). In fact, the
pollen of A. sessilis from India is reported to have granulated sexine ornamentation
(Rao & Shukla 1975) and without spinules (Mittre 1962). However, this kind of
apparent contradiction, especially the study from India, is difficult to resolve without
further confirmatory work because the methodology and voucher specimens of A.
sessilis were not mentioned by these authors and therefore taxonomic verification
could not be carried out.
Further, the pollen grains of 4. beftzickiana in the present study are different
from those reported in China (Li et al. 1993) in having bigger pollen and pore (present
study: polar length = 15.79 um, pore: 4.83 um) while those from China are smaller
(polar length = 10.90 um, pore = 3.60 um). Moreover, only a single row of spinules
are observed in the present study, whereas 1—2 rows of spinules are observed in the
pollen from China.
In addition, data obtained in the previous study of A. paronychioides is different
from the present findings. For instance, the pollen grains from Pakistan (Perveen &
-Qaiser 2002) were smaller (15.34 um) with 6—9 pores while those from this study are
bigger (21.85 um) with around 18 oval pores. Furthermore, the size of the pores in the
present study is about twice that of pores from the Pakistan pollen grains, 1.e., 6.06 um
and 3.44 um, respectively.
At the generic level, the pollen morphology of A/ternanthera species 1s close
to Pfaffia Mart. (Borsch 1998, Eliasson 1988), Zidestromia Standl. and Kyphocarpa
Schinz in having dodecahedric pollen grains and metareticulate sexine (Borsch 1998).
In fact, the current palynological data is generally in agreement with Borsch (1998) and
thus might help in the genus delimitation. For instance, the pollen of A/ternanthera can
be distinguished from Tidestromia by the sexine pattern. The pollen of 7: Januginosa
(Nutt.) Standl. is reported as devoid of spinules (Borsch 1998) and T. oblongifolia
(S.Watson) Standl. has a very narrow mesoporia which 1s triangular in cross-section
(Eliasson 1988). On the other hand, most of the A/ternanthera species in the present
study have one to two rows of microspines attached on the moderate mesoporia.
The number and shape of apertures appear to be an important key in
differentiating the Pfaffia and Alternanthera species. Instead of having spheroidal
grains with more than 20 pores as reported in Pfaffia (Borsch 1998), the pollen grains
of Alternanthera are dodecahedric with less than 20 pores. Most of the Pfaffia species
conform to the above characters except P. aurata (Mart.) T.Borsch, P. completa (Uline
& W.L.Bray) T.Borsch, P. costaricensis (Standl.) T.Borsch and P. densipellita T.Borsch
(Borsch 1995 & 1998). These species are reported also to have dodecahedric pollen
with 12-14 pores and metareticulate sexine. However, the distribution of microspines
on the sexine could be used to resolve this problem. The microspines are occasionally
arranged in an undulate row or side by side, as seen in P. aurata and P. costaricensis
(Borsch 1995) but distally and usually regularly arranged in most of the A/ternanthera
species.
Pollen morphology in Alternanthera 481
Conclusion
Pollen grains of A. sessilis, A. bettzickiana and A. paronychioides in Peninsular Malaysia
can be differentiated mainly by the number of apertures, number of ektexinous bodies
and distribution of perforations at the mesoporia. The apertures of 4. sessilis and A.
bettzickiana are pantoporate with 12 round pores whereas the pollen grains of A.
paronychioides have 18 oval pores. The sexine is metareticulate and tectum perforate
with unevenly distributed perforations at the top and base of the mesoporia, except in
the pollen grains of A. bettzickiana, in which the perforations are distributed unevenly
at the top of the mesoporia only.
Although this study has indicated that the palynology of the green and red
varieties of A. sessilis is remarkably similar, palynological data was only obtained
from specimens at a single habitat. As the green-leafed variety of A. sessilis could
be found from various habitats, e.g., in canals, ditches or wastelands, a survey on the
pollen morphology from those habitats should also be carried out. Besides, future
work should also include the transmission electron microscopy (TEM) study in order
to strengthen the hypothesis that the pollen morphology of the green and red varieties
of A. sessilis is not significantly different.
ACKNOWLEDGEMENTS. Financial assistance from the University of Malaya is gratefully
acknowledged. We thank Mrs. Patricia Loh and Mr. Roslee Halip for technical assistance.
References
Backer, C.A. (1949) A/ternanthera. Flora Malesiana, Ser. 1, 42: 91-94. Netherlands:
Noordhoff-Kolff N.V.
Balasuriya, B.M.G.K. & Dharmaratne, H.R.W. (2007) Cytotoxicity and antioxidant
activity studies of green leafy vegetables consumed in Sri Lanka. J. Natl. Sci.
Foundation Sri Lanka 35(4): 255-258.
Bashir, S. & Khan, M.A. (2003) Pollen morphology as an aid to the identification of
medicinal plants: Trianthema portulacastrum L., Boerhaavia procumbens Banks
ex Roxb. and A/ternanthera pungens Kunth. Hamdard-Medicus 46(1): 7-9.
Bhaskar, R.D., Subhashini, D.P., Aparanji, P., Rama, R.M. & Raghava, R.T. (2007)
In vitro evaluation of antioxidant properties and their capacity of A/ternanthera
sessilis and Tephrosia purpurea. Biosci. Biotech. Res. Asia 4(2): 551-560.
Borsch, T. (1995) The new combinations in Pfaffia (Amaranthaceae) from the New
World Tropics. Novon 5: 230-233.
Borsch, T. (1998) Pollen types in the Amaranthaceae. Morphology and evolutionary
significance. Grana 37: 129-142.
Borsch, T. & Barthlott, W. (1998) Structure and evolution of metareticulate pollen.
Grana 37: 68-78.
482 Gard. Bull. Singapore 63(1 & 2) 2011
Eliasson, U.H. (1988) Floral morphology and taxonomic relations among the genera
of the Amaranthaceae in the New World and the Hawaiian Islands. Bor. J. Linn.
Soc. 96: 235-283.
Erdtman, G. (1952) Pollen Morphology and Plant Taxonomy. Angiosperms. London:
Hafner Pub.
Erdtman, G. (1960) The acetolysis method. A revised description. Svensk Bot. Tidskr.
54: 561-564.
Erdtman, G. (1966) Pollen Morphology and Plant Taxonomy. Angiosperms. London:
Hafner Pub.
Fegri, K. & Iversen, J. (1950) Text-book of Modern Pollen Analysis. Copenhagen:
Eynar Munksgaard.
Goh, S.H., Chuah, C.H., Mok, J.S.L. & Soepadmo, E. (1995) Malaysian Medicinal
Plants for the Treatment of Cardiovascular Diseases. Petaling Jaya: Pelanduk
Publications.
Henderson, M.R. (1959) Malayan Wild Flowers, Dicotyledons. Kuala Lumpur:
Malayan Nature Society.
Jalalpure, S.S., Agrawal, N., Patil, M.B., Chimkode, R. & Tripathi, A. (2008)
Antimicrobial and wound healing activities of leaves of A/ternanthera sessilis
Linn. /ntern. J. Green Pharmacy 2(3): 141-144.
Kajale, L.B. (1940) A contribution of the embryology to the Amaranthaceae. Proc.
Nat. Inst. Sci. India 6(4): 597-625.
Kumaresan, P.T., Kumar, P.R.A.V., Rajasekaran,A.,Arivukarasu, R. & Balasubramanian,
S. (2001) Anti-bacterial activity of aerial parts of A/ternanthera sessilis. Indian J.
Nat. Prod. \7(1): 23-24.
Li, X., Huang, X. & Zhan, Y. (1993) Scanning electron microscopic observations on
the pollen morphology of Alternanthera from China. J. Wuhan Bot. Res. 11(2):
117-119.
Liang, Y.H., Long, H. & Liang, Q.F. (1978) Notes on pollen morphology of the plants
from Xisha Qundao. Acta Bot. Sin. 20(2): 166-178.
Mabberley, D.J. (2008) The Plant Book. Cambridge: Cambridge University Press.
Mittre, V. (1962) Pollen morphology of Indian Amaranthaceae. J. Indian Bot. Soc. 42:
86-101.
Miiller, K. & Borsch, T. (2005a) Phylogenetics of Amaranthaceae based on matK/trnK
sequence data-evidence from parsimony, likelihood and baysian analyses. Ann.
Missouri Bot. Gard. 92: 66—102.
Miiller, K. & Borsch, T. (2005b) Multiple origins of a unique pollen feature: stellate
pore ornamentation in Amaranthaceae. Grana 44: 266-281.
Nowicke, J.W. & Skvarla, J. (1979) Pollen morphology: the potential influence in
higher order systematics. Ann. Missouri Bot. Gard. 66: 633-700.
Perveen, A. & Qaiser, M. (2002) Pollen flora of Pakistan-X VIII. Amaranthaceae.
Pakistan J. Bot. 34(4): 375-383.
Rao, A.R. & Shukla, P. (1975) Pollen flora of Upper Gangetic Plain. New Delhi:
Today & Tomorrow Publishers.
Ridley, H.N. (1924) The Flora of the Malay Peninsula. London: Reeve & Co.
Pollen morphology in Alternanthera 483
Sanchez del-Pino, I., Borsch, T. & Motley, T.J. (2009) trnL-F and rpl/6 sequence
data and dense taxon sampling reveal monophyly of unilocular anthered
Gomphrenoideae (Amaranthaceae) and an improved picture of their internal
relationships. Syst. Bot. 34(1): 57-67.
Shyamala, B.N., Sheetal, G., Jyothi, A.L. & Prakash, J. (2004) Leafy vegetable extracts-
antioxidant activity and effect on storage stability of heated oils. Innovative Food
Sci. Emerg. Technol. 6: 239-245.
Song, C.L., Yun, H.L., Shyh, J.S. & Chung, C.L. (2006) Hepatoprotective effects of
Taiwan folk medicine: A/ternanthera sessilis on liver damage induced by various
hepatotoxins. Phytotherapy Res. 8(7): 391-398.
Turner, I.M. (1995) A catalogue of the vascular plants of Malaya. Gard. Bull. Singapore
47(1 &2): 1-757.
Png sy
pie, ) Le)
ae | ee
g ip ti®
to (<hr
pa th G9 ar Hs a
Ti Woo =o
J
ar iit Tac a
Ate? manag
i” a os] ol
j um = o
Tay O4 ig
ite BO an
Gardens’ Bulletin Singapore 63(1 & 2): 485-493. 2011 485
Flower biology of four epiphytic Malesian Gesneriads
Wiguna Rahman
Cibodas Botanic Gardens, Indonesian Institute of Sciences (LIPI),
Jl. Kebun Raya Cibodas, Sindanglaya, Cianjur 43253, Indonesia
wiguna.rahman@gmail.com
ABSTRACT. Floral traits, flowering events, nectar production and reproductive success of
Aeschynanthus horsfieldii R.Br., A. pulcher (Blume) G.Don, A. longiflorus A.DC. and Agalmyla
parasitica C.B.Clarke were observed for sympatric populations in the Cibodas area, Mount
Gede-Pangrango, West Java. All traits were significantly different among the species, but
were associated with a bird pollination syndrome. Many flowers of Aeschynanthus longiflorus
and Agalmya parasitica failed to develop mature stigmas. Agalmyla parasitica flowers take
a longer time to attract pollinators and receive pollen than the others and the filaments begin
to bend earlier than the others. Aeschynanthus pulcher produces more nectar than the other
species at the female phase, but the concentration was lower than in Aeschynanthus horsfieldii
and Agalmyla parasitica. These seem to be correlated with the reproductive success of the
respective species, with flowers of Aeschynanthus longiflorus and Agalmyla parasitica setting
fewer fruit than the other two species. Flower traits and pollination shift are discussed in light
of evidence that Aeschynanthus horsfieldii also attracts bumble bees (Bombus rucifes).
Keywords. Aeschynanthus, Agalmyla, Gesneriaceae, flower biology, Malesia, pollination
Introduction
Among 28 genera of Gesneriaceae that occur in Malesia, only Aeschynanthus Jack and
Agalmyla (Blume) G.Don have epiphytic representatives. Aeschynanthus comprises
approximately 160 species, while Aga/myla has about a hundred. The distribution areas
of these two genera overlap: Aga/myla (Hilliard & Burtt 2002) seems to be restricted
to Malesia and its distribution nested within that of Aeschynanthus (Mendum et al.
2001), a genus that is also significant outside of the region (e.g., Middleton 2007,
2009).
With a lack of direct observation of pollinators for every plant species,
pollination syndromes are usually inferred. However, Ollerton et al. (2009) and
Merxem et al. (2009) have recently cautioned that presumed pollination syndromes
do not always successfully predict the actual pollinators. Based on flower characters,
Aeschynanthus and Agalmyla have the characteristic association with bird pollination.
There is evidence that flowers of Aeschynanthus longiflorus, A. pulcher and three other
species of Aeschynanthus are usually visited by both Arachnothera spiderhunters and
sunbirds (Leeuwen 1937, McClure 1966). Aga/myla flowers have also been noted as
bird-visited (Hilliard & Burtt 2002). The nectar content of Aeschynanthus flowers
have the strength associated with bird pollination (Freeman et al. 1991).
486 Gard. Bull. Singapore 63(1 & 2) 2011
This paper compares the phenotypic traits of flowers of four plant species
from the presumedly bird-pollinated genera Aeschynanthus and Agalmyla, in view of
evidence for bee-pollination in one of the species, Aeschynanthus horsfieldii.
Material and methods
The epiphytic Gesneriaceae species studied were Aeschynanthus horsfieldii R.Br.
(section Microtrichium), Aeschynathus longiflorus A.DC. (uncertain sectional
affiliation), Aeschynanthus pulcher (Blume) G.Don (section Aeschynanthus), and
Agalmyla parasitica C.B.Clarke (section Aga/myla). All are widely distributed in West
Malesia. The observations were carried out at Cibodas, on the northern slope of Mount
Gede-Pangrango, West Java, from November 2009 to March 2010.
Flowers of these species were randomly tagged before they opened. For each
tagged flower, the day of flower opening was recorded and the flower was harvested
- following 0, 1, 3,5, 7, 10, and 13 days after opening. Some flowers, however, dropped
before harvesting. The sample size for each harvesting time was 5—10 flowers. For
each harvest time, the length of the calyx, corolla, flower tube, stamen, gynoecium;
the width of the flower mouth; and the diameter of the stigma were measured using
calipers (to +0.05 mm). Planar projection and en-face areas were measured followed
Dafni (1992).
The flowering event was observed using other flowers, with a sample size
of 27—50 flowers per species. These flowers were observed every day from January
to February, 2010. During this period and for each tagged flower, the day of flower
opening, convexing of the stigma, curvature and wilting of the filament, and the corolla
dropping, was recorded.
To observed nectar production, some sample flowers were bagged before their
opening, using flipped plastic. Nectar was extracted from detached flowers using 50 ul
micropipettes, at the staminate, sexual overlap and pistillate phases for each species. At
each flower phase, nectar was extracted at four different times, 0700—0800 hrs, 1000—
1100 hrs, 1300-1400 hrs, and 1600-1700 hrs. Sugar concentration in the nectar was
measured for each flower using a portable sugar refractometer (Kenko Refractometer,
0-80 % Brix).
A whole shoot bearing flowers was monitored through making a “flower map”
when it was difficult to individually tag every flower for observation. Flowers setting
fruits were counted and the percentage fruitset of the total flower number observed
was computed.
Results
Flower traits
Flower traits of the four study species are presented in Table 1. All flowers of
observed taxa are red (e.g., Fig. 1). The calyces are free, divided to the base, except
487
Flower biology of epiphytic Gesneriads
Table 1. Flower traits of the four study species. Values are expressed as mean + SD (with
n, sample size, in brackets); those marked by the same superscripts in each row are not
significantly different.
Statistical analysis
to enface area
ss Aeschynanthus Aeschynanthus Aeschynanthus Agalmyla
Flower trait ix ; < is
horsfieldii longifiorus pulcher parasitica F Pp
Flower colour red red red red
Calyx shape _ free, dividedto free, divided to free, divided
b cup-shaped
ase base to base
Flower attach- pendent erect erect erect
ment
~ Filament curled down curled down curled down curled back
curvature
Corolla length 27.9941.28 81.5842.54 —-63.9543.11. 43.7842.04" 1989.793 <0.001
(mm) (20) (30) (65) (20)
Corollatube -22.0340.99 _74.8442.42° -49.4922.71' 36.3342.70_2156.605 <0.001
length (mm) (20) (30) (65) (20)
Flower mouth 7.98+0.62 14.6321 42 16.29=1 28° 12.48+0.92° 266.055 <0.001
width (mm) (20) (30) (65) (20)
Filament 26.8841.37 95.1143.63 63.1129.28 64.4542.73' 284.820 <0.001
length (mm) (20) (30) (65) (20)
En-face area 0.9020.19 2.1740.48 4.42+0.66 1.04940.28 394.289 <0.001
(cm?) iy) (30) (65) (20)
Profile planar ~—«'1.71+0.21. 6.8920.66- 5.4540.63 3.854048 425.266 <0.001
area (cm?) (20) (30) (65) (20)
Ratio of 1.96+0.35. 3.3 140.83" 4.42+0.66 3.9421 28 $1.109 <0.001
profile planar (20) (30) (65) (20)
for Aeschynanthus pulcher, which has a cup-shaped calyx. Flowers of A. horsfieldii
are pendulous, while in the three others they are erect. The filaments of the three
Aeschynanthus species are curled downwards, whereas those of Agal/myla parasitica
are curled back. The longest corolla length, corolla tube length, and filament length
are found in A. Jongiflorus. The widest flower mouth, en-face area, and ratio of profile
planar to en-face ratio are found in 4. pulcher.
Due to protandrous development, gynoecium length in all four species
increased gradually after flower opening (Fig. 2). However, as much as 30.22 % (n
= 225) of Agalmyla parasitica and 80.89% (n = 178) of Aeschynanthus longiflorus
flowers failed to developed mature stigmas.
488 Gard. Bull. Singapore 63(1 & 2) 2011
Fig. 1. Bumble bee (Bombus rufipes Lep.) visiting and pollinating flowers of Aeschynanthus
horsfieldii R.Br.
80 80
60 60
40 40
20 20
0 1 3 5 7 (0) als! 0 1 3 5 7 if} ale?
Days after anthesis Days after anthesis
Fig. 2. Mean gynoecium length (+1 SE), in mm, of four species of Gesneriaceae after anthesis
(flower opening): Aeschynanthus pulcher (@), A. horsfieldii (mw), A. longiflorus (¢) and
Agalmyla parasitica ( &). Note the horizontal axes are not to scale.
Flower biology of epiphytic Gesneriads 489
4 (a) —7———— (b) af Hip (0)
a. - 3 3 | =
a 2 2 |__|
ai 1 —— J 1
0 5 10 15 20 0 4 8 12 2 fe) 2 4 6
4 4 =i (e) 4, (f)
3 3 -—_ 0 OTE
2 2, ——/ 2)
1 on nl
0 5 10 i5 0 3 6 9 12 15 4 8 12 1€
Duration (days) Duration (days) Duration (days)
Fig. 3. Boxplots of the duration (in days) of (a) flower longevity, (b) male phase, (c)
sexual overlap phase, (d) female phase, (e) filament curvature, and (f) stigma receptivity in
Aeschynanthus horsfieldii (@ Bar 4 in each plot), A. /ongiflorus (@ Bar 3 in each plot), A.
pulcher (@ Bar 2 in each plot) and Agalmyla parasitica ( Bar | in each plot).
Male Phase Sexual Overlap Phase Female Phase
500 + 4 4
A
{
4
B eae
0 “t r “ + 4 —+— + 4
= ai |
30 == t = |
Cc ; = SS SS ae
20 aaa | a a
‘4 ao =
10 Ls 2S
0 ' Y y i r a : == T IF Y T T
0700 1000 1300 1600 0700 1000 1300 1600 0700 1000 1300 1600
Time Time tan
Fig. 4. Nectar volume (11) (A), sugar concentration (% sucrose w/w) (B) and sugar amount
(mg) (C) in Aeschynanthus pulcher (—), A. horsfieldii (—), A. longiflorus (—) and Agalmyla
parasitica (—).
490 Gard. Bull. Singapore 63(1 & 2) 2011
Flowering event
Characteristics of the flowering event (flower longevity, duration of the male and
female phases, timing of filament curvature, and duration of stigma receptivity) are
significantly different across the species, except the sexual overlap phase (i.e., when
stamen functionality and stigma receptivity show an overlap), which showed weak
differences (Fig. 3). The longest flower longevity was observed in Agalmyla parasitica
(Fi), 14s; 18.506, p<0.001). The longest male phase was found in Aeschynanthus
longiflorus (F’, = 13.410, p<0.001), and the shortest female phase in A. horsfieldii
(F’, ga 77-314, p<0.001). The duration of the overlap phase is only weakly different
among the species (F, ,,=3.09; p<0.05). The sexual overlap phase is negative when
the stigma becomes receptive after filaments have wilted, or when flowers do not
display such overlap. The filaments of A. /ongiflorus were more slow to curve than in
the three other species (F', ,,.=16.955; p<0.001). The shortest stigma receptivity was
(3.115)
observed for A. horsfieldii Fo) no 007, P0001):
3.109)
Flower nectar production
Nectaries of the four species are located at the flower base. In the Aeschynanthus
pulcher flower, the corolla has a swollen (slightly bulbous) base which is not found
in the other three species. The pattern of nectar production showed in Fig. 4. Nectar
volume, and nectar sugar amount and concentration, vary over the time of day. There
were also significant differences in nectar volume between the species and between
flower phases, except in A. horsfieldii (Table 2).
The mean nectar sugar concentration was significantly different between
species, but not between the flower phase in each species, except Aga/myla parasitica
(Table 2). The highest nectar sugar concentration was found in Aeschynanthus
horsfieldii.
The mean nectar sugar amount was significantly different between species
in the overlap and female phases, but not in the male phase, when it was relatively
minimal (Table 2). The mean nectar sugar amount was also significantly different
between flower phases in Aeschynanthus pulcher and A. longiflorus, but not in A.
horsfieldii and Agalmyla parasitica. The highest nectar sugar amount was found in
Aeschynanthus pulcher flowers.
Reproductive success
More flowers of Aeschynanthus pulcher and A. horsfieldii successfully set fruit than
in A. longiflorus and Agalmyla parasitica (Table 3). This unequal fruitset between
the species could possibly indicate the presence of competition between species for
pollinator services.
Discussion
According to conventional interpretation of pollination syndromes, bird pollination
flowers have tubular shapes, are frequently red and odourless, and produce copious
49]
ytic Gesneriads
gy of epiphy
Flower biolo
100'0> LE OS
100'0> 8S°EC
cvL 0 [v0
d a
oNSHRIS
L00'0> — O8'CT
LOOIO== 28°C
[000> vce
d A
INSHRIS
100;0> = 8°89
100'0> = 80'&E
rlo0 SEAS
d al
onsHeysg
€80'0=4 °66° T=
(OZ) 89° €FSL'L
(07) SE'EFIO'L
(OZ) 99° T#BH'S
poiispsévd p]AUDS py
Z00'0=4 S894
(07) €6' 1F6L'€1
(O07) 6S IFIT'SI
(OZ) 77 1FI9'S I
poyispivd pjAU]ps py
S0'0>4 “b10°S=
(OZ) ZE'9TFIL-9S
(07) 68 OTFLE OF
(OT) Eb SIFSL' bE
poyispapd pjAUujo3 py
865 0=4 61S 0-4
(OZ) $8 IF16'9
S10°0=4 °€9 pH
(07) SS 9FIE'O1
(07) 88° S¥90'01 (07) Th €FSEL
(00) LV CFLE'S (07) I7'7#TS'9
SNAOYIBUO] SNYJUBUAYISAY — LIpjaifssoy snyjubudyIsap
(Sur) JunNOWRe Ie3ns IBIDON
pS10=4 * 9€6 [=H
(07) ST’ EFLS'91
(07) 8 E+96 91
811 0=4 '70 =a
(07) 89°7¥F 101
(07) Le-TFSO'TI
(07) LTFL8'I (OZ) 90°TF87 81
SNAOYIBUO] SNYJUBUAYISAP — LIpjalfSsoy snyjuvudyosop
(0%) UONR.NUDDUOD IWBNs eID
7S0'0=4 801 =H
(OZ) LE 8FET TH
(OZ) 1S S1¥#60°€b
(OZ) IL6FLL' bE
10° 0>4 “PLS: L=a
(07) 6S €h+90'S6
(07) 78° PE FPL S8
(07) TL 9TFOI'ES
SNAOYIBUO] SnyJUBUAYSep — Lpjaif{saoy snyjuvudyosay
({11) owinjoA,
100'0>4 °S 1 Lev
(07) 69 ITFLT'ZE
(OZ) €7' 11 F6h'7Z
(07) 67 bFFI'9
dayojnd snyjuvudyosapy
ZIP 0=4 °6' 04
(07) 16° 1+6r'T1
(07) ZO'TFIF'EI
(07) 8¢'7#96 71
dayond snyjuvudyosap
100°0>4 ‘€1 pr=H
(07) LI'S6F88 6ST
(07) 69° 9LFS6' $91
(07) SE 9TFLE'Sh
dayojnd snyjupudyosap
SHSHEYS
(Sw) oyeudy
(Sut) depssag
(Suu) ae
aseud
SULIOMO] 4
oNSHRIS
(0%) o] BUD 4
(%) de}1orO
(%) PW
ase
SULIOMO] 4
SHSHEIS
({v) oyeutay
({1) depsro
(1) [RW
aseyd
SULIOMO] 4
‘payeorpur (sjoyorsq Ul ‘U) ozIs o]duies pue (qs) suONeIAap piepurys
YIM SUBST d1v SOnjeA “SULIOMOY Jo sasvyd oyeutay pure dejs9A0 [enxos ‘aye oy) BuLNp ‘PouIspAnd vjAMjOSsp pur ‘snsopfisuo) “py ‘Upjayfssoy
‘Vy ‘dayond snyjupudyosay Jo (SW) 1e499U UL ISHS Jo JUNOW pur (M/M aSOIONS 0%) UONL.QUDOUOD AeBNs seJO9U “(Jr1) SUINJOA IRIDON *Z 9IqUI.
492 Gard. Bull. Singapore 63(1 & 2) 2011
Table 3. Percentage fruitset of Aeschynanthus pulcher, A. horsfieldii, A. longiflorus and
Agalmyla parasitica.
Species No. of flowers No. of fruits set % fruitset
Aeschynanthus pulcher 247 82 S8)ail)
Aeschynanthus horsfieldii 177 40 34.18
Aeschynanthus longiflorus 178 23 WROD
Agalmyla parasitica 225 50 DDD
amounts of nectar, with the stigma and anthers conspicuously exserted from the
floral tube. Although the flowers of all four species of Gesneriaceae observed have
this classic bird pollination syndrome, the flowers of Aeschynanthus horsfieldii were
frequently visited by bumble bees (Bombus rupifes Lep.) and could be said to show a
shift in floral presentation.
Shifts in flower traits in A. horsfieldii may include the shallower flower tube
length, narrower side advertisement, shorter flower longevity and higher nectar sugar
concentration. With a shallower flower tube length, the nectar is more easily foraged
by bees. The reduced flower tube length also decreases the potential effectiveness of
side advertisement (profile planar area). According to Dafni (1994), there are little
differences in the degree of side advertisement between bird flowers and large bee
flowers. Primack (1985) and Stratton (1989) have also shown that bee flowers have
shorter longevity than bird flowers. In Sinningiae (Gesneriaceae), bee flowers also
have higher nectar sugar concentration than bird flowers (Perret et al. 2001).
All four species observed have red flowers, although only A. horsfieldii appeared
to attract bumble bees. We do not know how the red colour in A. horsfieldii flowers is
compatible with bee vision, and suggest that perhaps flowers of A. horsfiledii could
have UV reflectance properties. Flowers which have red colour with UV reflectance
can attract bees (Chitka & Waser 1997).
In terms of reproductive success, the existence of co-flowering species with
similar syndromes should increase inter-specific competition and reduce pollination
success (Sargent & Ackerly 2008, Chitka & Shiirkens 2001). We suggest that each of
the four related species could have developed a different strategy to attract pollinators.
Aeschynathus pulcher has developed flowers with large side advertisement and which
produce high nectar volume and nectar sugar amounts. Aeschynathus horsfieldii has
flowers able to attract more than one pollinator class (birds and bumble bees) through
narrow side advertisement, high nectar sugar concentration and probably both nectar
and pollen as rewards.
Aeschynanthus longiflorus and Agalmyla parasitica could have a “flower
dimorphism syndrome”, sometimes apparently showing andromonoecy, when some
plants only present flowers with undeveloped gynoecia (i.e., with functionally male
flowers), while others present the usual hermaphrodite condition. From a population
perspective, producing more male flowers is a strategy to increase pollen transfer when
there are limitations in plant resources and pollinator visitation.
Flower biology of epiphytic Gesneriads 493
References
Chitka, L. & Schiirkens, S. (2001) Successful invasion of a floral market. Nature 411:
653.
Chitka, L. & Waser, N.M. (1997) Why red flowers are not invisible to bees. Israel J.
Plant Sci. 45 (2-3): 169-183.
Dafni, A. (1992) Pollination Ecology: A Practical Approach. Oxford.
Dafni, A. (1994) Note on side advertisement in flowers. Function. Ecol. 8: 136—138.
Freeman, C.E.. Worthington, R-D. & Jackson, M.S. (1991) Floral nectar sugar
compositions of some South and Southeast Asian species. Biotropica 23: 568—
574.
Hilliard, O.M. & Burtt, B.L. (2002) The genus Agalmyla (Gesneriaceae-
Cyrtandroideae). Edinburgh J. Bot. 59(1): 1-210.
Leeuwen, W.M_D. van (1937) Observation about the biology of tropical flowers. Ann.
Jard. Bot. Buitenzorg 48: 27-68.
McClure, H.E. (1966) Flowering, fruiting, and animals in the canopy of a tropical rain
forest. Malayan Forester 29: 182-203.
Mendum, M., Lassnig, P., Weber, A. & Christie, F. (2001) Testa and seed appendage
morphology in Aeschynanthus (Gesneriaceae): phytogeographical pattern and
taxonomic implications. Bot. J. Linn. Soc. 135: 195-213.
Merxem, D.G. de, Borremans, B., Jager. M.L. de, Johnson, T., Jooste, M., Ros, P..
Zenni, R.D.. Ellis. A-G. & Anderson, B. (2009) The importance of flower visitors
not predicted by floral syndrome. S. African J. Bot. 75 (4): 660-667.
Middleton. D.J. (2007) A revision of Aeschynanthus (Gesneriaceae) in Thailand.
Edinburgh J. Bot. 64(3): 363-429.
Middleton, D.J. (2009) A revision of Aeschynanthus (Gesneriaceae) in Cambodia,
Laos and Vietnam. Edinburgh J. Bot. 66(3): 391446.
Ollerton, J., Alarcon, R., Waser, N.M.. Price. M.V.. Watts, S.. Cranmer, L., Hingston,
A., Peter, C.L. & Rotenberry, J. (2009) A global test of the pollination syndrome
hypothesis. Amn. Bot. 103(9): 1471-1480.
Perret, M., Chautems,A., Spichiger. R., Peixoto, M. & Savolaine. V. (2001) Nectar sugar
composition in relation to pollination syndromes in Sinningieae (Gesneriaceae).
Ann. Bot. 87: 267-273.
Primack, R.B. (1985) Longevity of individual flowers. Ann. Rev. Ecol. Syst. 16: 15-27.
Sargent, R. & Ackerly D.D. (2008) Plant-pollinator interactions and community
assembly. Trends Ecol. Evol. 23: 123-130.
Stratton, D.A. (1989) Longevity of individual flowers in Costa Rican cloud forest:
ecological correlates and phylogenetic constraints. Biotropica 21(4): 308-318.
¥De
I .
"da
me,
a . — —
ne Le as fis | i
a
6 he (ape D iim myers
ar v0
7 see!
ws | a pe
7 i nie wea bor 7
4 7 : =
relist)
Wea
cit eLMaiip
Hed = et
we a . pr al ine
Lig at ae.
i rae) pot
( OTe ec.
Ait Ltn oan
1 ss ew,
per erty Geers
7) say ane
he pA (| Ronee
: Sim age 5
sine al
-<
=
a
ae 7
Gardens’ Bulletin Singapore 63(1 & 2): 495-498. 201] 495
The genus Premna (Lamiaceae)
and the presence of ‘pyro-herbs’ in the
Flora Malesiana area
Rogier de Kok
Herbarium, Library, Art and Archives,
Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3AE, U.K.
r.dekok@kew.org
ABSTRACT. The genus Premna consists of 14 species in the Flora Malesiana area. The
most common species in the region are all widespread. However, a series of morphologically
closely-related, rare and generally geographically more restricted species are present in the
region. These species can be characterised by three distinct morphological characters: 1) small
decussate scales at the base of the young twigs, 2) a calyx that always has four isomorphic
lobes, 3) a fruit that is clavoid in shape and almost single-seeded. The ecology and morphology
of Premna herbacea Roxb. is unique in the genus and is the first recognised ‘pyro-herb’ in the
Flora Malesiana area.
Keywords. Lamiaceae, Malesia, Premna, pyro-herbs
Introduction
The genus Premna was first described by Linnaeus in 1771 and occurs in the old
world tropics from Africa to China and south to Australia and the Pacific (Harley et
al. 2004). Following most other genera in the Lamiaceae, Premna is rich in species
and morphological variation on the Southeast Asian mainland and on the islands of
the Sunda shelf. It is widespread in the Pacific, but the number of species decreases
sharply south and east from New Guinea. Eighty percent of all Premna specimens in K
belong to only four species (P. serratifolia, P. odorata, P. trichostoma and P. tomentosa)
and these are all widespread. There are two groups of species in this genus. The first
group (P. serratifolia-group, see Table 1) has most of these common species: (P.
serratifolia, P. odorata and P. tomentosa), while species in the second group tend to
be rare and generally geographically more restricted (P. trichostoma group, see Table
The difference between the groups is mainly based on three distinct morphological
characters. The P. trichostoma-group has a series of small decussate triangular scales
at the base of the young twigs, while on older branches these scales usually have
fallen off, leaving a series of closely packed bract scars; a calyx that always has four
isomorphic lobes, its shape remaining largely intact when the flower develops and
when the fruits are formed: and a fruit that is clavoid in shape, almost single seeded
(four seeds present, but only one fully developing). While the P. serratifolia-group
496 Gard. Bull. Singapore 63(1 & 2) 2011
Table 1. Species of Premna in the Flora Malesiana area and their distribution in each
morphological group.
Taxon Distribution
P. serratifolia-group
P. odorata Blanco
P. pubescens Blume
P. serratifolia L.
P. sterculiifolia King & Gamble
P. tomentosa Willd
P. trichostoma group
P. clavata de Kok
P. decurrens H.J.Lam
P. herbacea Roxb.
P. interrupta Wall. Ex Schauer
P. oblongata Miq.
P. pallescens Ridl.
P. parasitica Blume
P. regularis H.J.Lam
India and China to Australia
Java, Sumatra, the Lesser Sunda Islands and the
Philippines
East Africa to Tahiti
Peninsular Malaysia
India and China to Australia, except Borneo
Sabah
Sumatra
Southeast Asia to Australia
India and China to Malaysia
Sunda Islands and Sulawesi
Borneo
Java and Bali
Philippines and New Guinea
P. trichostoma Miq. Myanmar and Vietnam to New Guinea
does not have these bracts at the base of the young shoots, the number of calyx lobes
varies from 0 to 5 and are almost always heteromorphic, and the fruits have four
mature seeds per fruit.
The ecology and the morphology of P. herbacea is of interest. The herbaceous
habit of this species 1s an illusion, as only the herb like twigs are visible above ground
(and are usually the only parts collected), but a short woody stem exists below or
near to the ground. Premna herbacea 1s in gross-morphological and ecological terms
very similar to the well known ‘pyro-herbs’, which occur in vegetation types adapted
to frequent fires. Given the many notes on herbarium labels stating that this species
occurs in vegetation which is frequently burned, the same factor may have been the
driving force behind the evolution of P. herbacea.
Pyro-herbs
Pyro-herbs are woody plants that survive frequent fires by reducing their woody parts
to underground structures and then only sprout herb-like branches each year in the wet
season. They are very common in parts of Tropical Africa and South America and are
rare in Asia. They are reported to be absent from Australia and South East Asia (White
1976). Since White’s overview article on pyro-herbs (or the suffrutescent habit), their
presence has been indentified in north Australia. At least three species belonging to the
Labiatae are now recognised to be pyro-herbs (Clerodendrum tatei (F.Muell.) Munir
Premna and pyro-herbs in Malesia 497
and Clerodendrum linifolia (Ewart & B.Rees) de Kok and Premna herbacea Roxb.
Given that the two essential elements of recognising a plant as a pyro-herb is the
woody underground parts (seldom present in herbaria) and a fire-regulated ecology
(seldom mentioned on herbarium labels), pyro-herbs are difficult to recognise without
extensive fieldwork. The case of C. /inifolia shows clearly the difficulty of recognising
these kinds of plants from herbarium material only. The species was first described as a
monotypical genus Hux/eya in the then also most entirely woody family Verbenaceae.
The main reason that it was described as a new genus was because the type specimen
consisted of only the herbaceous above-ground branches. This misconception lasted
until more detailed morphological, ecological, chemical and molecular research
revealed the true phylogenetic relationship of the genus and its survival strategy in
a habitat that burns almost annually (de Kok et al. 2000, Steane et al. 2004). Similar
ecological observations have now been made for the Australian populations of P.
herbacea (Munir 1984; de Kok, in press).
Absence or presences of pyro-herbs in Southeast Asia
There is not much literature about pyro-herbs in Southeast Asia. In his papers on the
ecology of the Indramajoe plains in West Java, van Steenis (1936) mentioned some
possible examples (see Table 2). On the other hand, White (1976) reports them to be
absent from South East Asia, and Henty (1982) mentions some possible examples
from Papua New Guinea from what he calls ‘the short lowland grasslands’ which are
maintained by frequent fires.
Conclusions:
1) The 14 species of Premna in the Flora Malesiana area can be divided into two
distinct groups: the P. trichostoma and P. serratifolia groups, based on morphological
characters.
2) Premna herbacea Roxb is the first recognised ‘pyro-herb’ in the genus in the Flora
Malesiana area.
Table 2. Possible ‘pyro-herbs’ occurring in Southeast Asia.
Taxa Sources
Butea monosperma (Lam.) Taub Van Steenis 1936
Crotalaria alata Buch.-Ham. ex D.Don Henty 1982
Crotalaria ferruginea Scheele Henty 1982
Crotalaria montana B.Heyne ex Roth Henty 1982
Dillenia sp. Van Steenis 1936
Fordia fruticosa Craib Van Steenis 1936
Grewia sp. Van Steenis 1936
Morinda sp. Van Steenis 1936
Phyllanthus emblica L. Van Steenis 1936
Premna herbacea Roxb. de Kok, in press
Ziziphus sp. Van Steenis 1936
498 Gard. Bull. Singapore 63(1 & 2) 2011
References
Davy, J.B. (1922) The suffrutescent habit as an adaptation to environment. J. Ecol. 10:
211-219.
Harley, R.M., Atkins, S., Budantsev, A.L., Cantino, P.D., Conn, B.J., Grayer, R.,
Harley, M. M., Kok, R.P.J. de, Krestovskaja, T., Morales, R., Paton, A.J., Ryding,
O. and Upson, T. (2004) Labiatae . In: Kubitzki, K. (ed) The Families and Genera
of Vascular Plants V1, pp. 167-275. Berlin: Springer.
Henty, E.E. (1982) Grasslands and grassland succession in New Guinea. In: Gressitt,
J.L. (ed) Biogeography and Ecology of New Guinea, vol 1, pp. 459-473. The
Hague: Junk Publisher.
Kok, R.P.J. de (in press) The genus Premna L. (Lamiaceae) in the Flora Malesiana
area. Kew Bull.
Kok, R.P.J. de, Grayer, R.J. & Kite, G.C. (2000) Relationships of the endemic Australian
genus Huxleya Ewart & Rees (Labiatae) based on fruit and flavonoid characters.
Austral. Syst. Bot. 13: 425-428.
Munir, A.A. (1984) A taxonomic revision of the genus Premna L (Verbenaceae) in
Australia. J. Adelaide Bot. Gard. 7: \—44.
Steane, D.A., Kok, R.P.J. de & Olmstead, R.G. (2004) Phylogenetic relationships
between Clerodendrum (Lamiaceae) and other Ajugoid genera inferred from
nuclear and chloroplast DNA sequence data. Molec. Phylogen. Evol. 32: 39-45.
Steenis, C.G.G.J. van (1936) Landschap en flora in Indramajoe. Trop. Natuur 25: 117—
118.
Steenis, C.G.G.J. van (1948) General considerations, Anthropomorphosis. Flora
Malaysiana, vol 4, pp. XXX VII-XX XVIII. Djakarta: Noordhoff.
White, F. (1976) The underground forest of Africa: a preliminary review. Gard. Bul.
Singapore 29: 57-71.
Gardens’ Bulletin Singapore 63(1 & 2): 499-505. 2011 499
A preliminary study on in vitro seed germination and
rooted callus formation of Tetrastigma rafflesiae (Vitaceae)
Netty W. Surya and M. Idris!
Plant Physiology Laboratory, Biology Department,
Faculty of Mathematics and Natural Sciences, Andalas University,
Padang, West Sumatra, Indonesia
‘uwakidris@gmail.com
ABSTRACT. /n vitro seed germination and induction of rooted callus formation was
investigated as a preliminary study on the propagation of Tetrastigma rafflesiae as potential
host plant material towards a sustainable conservation effort for Rafflesia. Seed germination
of T rafflesiae is epigeal with seedling emergence ranging 30-60 days after planting (dap),
regardless of light presence. After 60 days, 54-60% of seeds germinated in media treatments.
Callus formation began 7 dap in MS medium + 2 mg/L NAA, and 21 dap in MS medium + 2
mg/L 2,4-D. Browning of this medium due to phenolic compounds resulted from cutting a part
of the hypocotyl. Rooting of the callus was obtained after 21 days on MS medium + 2 mg/L
NAA, but was not evident with addition of 2 mg/L 2,4-D.
Keywords. In vitro seed germination, Tetrastigma rafflesiae, rooted callus
Introduction
Tetrastigma (Miq.) Planch. (Vitaceae) includes c. 100 species throughout tropical and
subtropical areas, 1n Asia and northern Australia. Some of those species are traditionally
used for medication in Indonesia and Malaysia, especially as leaf poultice in treating
fevers and headaches. In the Philippines, Tefrastigma is used to cure scabies. In
Vietnam, the leaf extract is used either internally or externally to treat headaches and
fevers. In addition, the fruits of some species can be eaten (Lemmens 2003).
Tetrastigma is also well known as the exclusive host of the parasitic Rafflesia
R.Br. (Nais 2001). Most Rafflesia spp. in Sumatra subsist parasitically on Tetrastigma
rafflesiae (Miq.) Planch., of which 7: leucostaphylum (Dennst.) Alston ex Mabb. is a
synonym; see Veldkamp (2008). A shortage of the host plant caused by, for example,
habitat reduction, can potentially affect the natural establishment and survival of
Rafflesia spp. (Attenborough 1995, Meijer 1997). The seriously damaged forest
environment in Sumatra makes it likely that viable Rafflesia populations remain only
in some protected areas (Zuhud et al. 1999, Sofiyanti et al. 2007). It may be possible to
help preserve the existence of Rafflesia in Sumatra by carrying out in vitro propagation
of the host plant, so that host-plant abundance and sites can be potentially increased.
As far as we know, there have been no reports about in vitro propagation
of Tetrastigma. Tissue culture techniques for 7efrastigma can potentially be adapted
from those used for the better investigated commercially important Vitis L. In vitro
propagation of Vitis, particularly grape, has been commercially carried out since a long
500 Gard. Bull. Singapore 63(1 & 2) 2011
time ago (Akbas et al. 2004, Salami et al. 2005, Alizadeh et al. 2010). Jaskani et al.
(2008) added the auxin NAA (approximately 2 mg/L) to the Murashige-Skoog (MS)
medium to induce rooted callus and embryo formation from grape leaf tissue.
This preliminary study aims to investigate the period of seed germination of 7,
rafflesiae on various media preparations, and the optimal conditions for in vitro seed
germination. It also attempts to induce rooted callus from pieces of hypocotyls by
using different types of auxin.
Materials and methods
Tetrastigma rafflesiae fruits were collected from the Andalas Botanical Garden for use
as an explants source.
The Murashige-Skoog (MS) basic medium was used, to which was added
modified active carbon, kinetin, NAA (1-Naphthalene Acetic Acid), or 2,4-D
(2,4-Dichlorophenoxyacetic acid), as required. Other substances added included 0.7%
agar and 3% sucrose. The culture medium acidity was controlled at pH 5 + 0.5. The
culture medium was heated until it boiled before being poured into sterilised culture
bottles. The bottles were then covered with aluminum foil and paper and secured with
rubber bands. Culture bottles were autoclaved for 15 minutes at 121°C and 15 psi
pressure.
Ripe fruits of T. rafflesiae were sterilised by soaking in 5% commercial
detergent solution for 20 minutes and washing them in flowing water for 5 min. The
fruits were then rinsed with 70% alcohol for 5 min, then commercial bleach (30%)
with 2 drops of Tween 20 for 5 minutes, before being washed three times with sterile
distilled water. Fruits were then peeled and seeds extracted and _ sterilised with 70%
alcohol for 3 min, and with 10% hypochlorite mixed with | drop of Tween 20 for 3
min before finally being washed with sterile distilled water for 5 min. These sterilised
seeds were then ready for implantation into the treatment media.
For in vitro seed germination experiments, 7. rafflesiae seed was germinated
on four media types: basic MS medium, MS + 1 g/L active carbon, MS + 0.5 mg/L
kinetin and MS + 0.5 mg/L kinetin + | g/L active carbon. The seed-implanted culture
media were then placed in two environments, a controlled incubation room (24°C +
2°C) with alternate 12 hours lighting (1000-1500 Lux), and a dark room. Type of seed
germination, period (range in days) of germination, and percentage germination were
recorded.
To study the effect of two types of auxin on in vitro induction of rooted callus
from excised hypocotyls of 7. rafflesiae, the hypocotyls were cut and implanted on
two types of media to induce rooted callus formation. Media employed were MS + 2
mg/L NAA + 0.5 mg/L kinetin, and MS +2 mg/L 2,4-D + 0.5 mg/L kinetin. The culture
media with hypocotyls were placed in a room with 12 hours lighting and in a dark
room, for 2 weeks, to induce callus formation. The time of callus formation, time of
rooted callus formation, type and colour of callus, and percentage of callus and rooted
callus formation, were recorded.
In vitro Tetrastigma germination and callus formation 501
Results and discussion
In vitro seed germination of Tetrastigma rafflesiae
Absence of light during germination did not affect the period of germination. However,
phenolic formation on the treatment media was higher in the presence of light (Fig. 1
A-B).
The type of seed germination is epigeal in which the hypocotyl is elongated
and the cotyledons are raised above the growth substrate (Fig. 1). The plants with this
type include cucumber, cotton, sesbania (Tischler et al. 2000), sunflower, pea and flax
(Klicova et al. 2004).
Table 1 shows that the time taken to first germination is 30 days after being
implanted on the medium for all treatments. The seed that germinated the highest
was seen on the MS medium with 0.5 mg/L kinetin added. Thus kinetin is capable of
promoting seed germination. Conversely, the lowest seed germination was seen on the
MS medium without growth regulators added; statistical significance, however, was
not tested in this preliminary trial.
The addition of active carbon does not affect the growth medium except that
it can reduce phenolics production by the explants. However, the phenolic compounds
produced do not appear to affect the in vitro seed germination of T. rafflesiae. Klicova
et al. (2004) assert that sunflower seed germination requires cytokinin-type growth
regulators to induce growth of the shoot with the use of 0.12% BA (6-benzyladenin).
Fig. 1. Tetrastigma leucostaphylum seed germination in two environmental conditions in MS
basic medium, after 8 weeks of treatment (A—B); and on four types of media with alternate
12-hours lighting, after 60 days (C-F). A. Complete darkness. B. Alternate 12-hours lighting.
C. MS basic medium. D. MS + 1 g/L active carbon. E. MS + 0.5 mg/L kinetin. F. MS + 1 g/L
active carbon + 0.5 mg/L kinetin. Photo by M. Idris.
502 Gard. Bull. Singapore 63(1 & 2) 2011
Table 1. Time of seed germination (days after planting (dap) in the medium) and percentage
germination after 60 days, in alternate 12 hours lighting.
No Treatment Germination (dap) % Germination Tissue formation /
medium
roots formed,
cotyledons raised,
hypocotyls elongated /
medium became brown
roots formed,
MS + 0.5 mg/L cotyledons raised,
9 tee
> kinetin oe 60 hypocotyls elongated /
medium became brown
roots formed,
7 :
3 MS + 1 g/L active 30-52 56 cotyledons raised,
carbon hypocotyls elongated /
medium normal
MS + | g/L active el anaes q
4 carbon + 0.5 mg/L 30-52 56 Se ae aaa
hypocotyls elongated /
kinetin ;
medium normal
Table 2. Period (days after planting, dap) and percentage of callus and rooted callus formation.
Treatment Period of callus Period of rooted % Callus % Rooted
formation (dap) callus formation formation callus
(dap) formation
MS + 2 mg/L NAA +
0.5 mg/L kinetin 7-10 21-30 100 80
MS + 2 mg/L 2,4-D +
0.5 mg/L kinetin 15—2] (nil) 100 0
Auxins and rooted callus induction from excised T. rafflesiae hypocotyls
MS medium augmented with 2 mg/L NAA was more effective at inducing rooted callus
formation compared to the MS medium with 2 mg/L 2,4-D (Table 2). In the NAA-
augmented medium, the elongation process and callus formation already began just
7 days after planting, and rooted callus formation began after 21 days after planting.
Meanwhile, in the 2,4-D-augmented medium, there was browning and the tissue in
fact shrank, and there was no sign of root formation (Fig. 2).
Jaskani et al. (2008) showed that addition of 2 mg/L NAA to the MS basic
medium promoted callus formation and rooted callus formation from grape leaf tissue.
AN
S
oS)
In vitro Tetrastigma germination and callus formation
Fig. 2. Formation of rooted callus on excised Tetrastigma leucostaphylum hypocotyls in
different media. MS + 2 mg/L NAA + 0.5 mg/L kinetin at 30 (A1) and 60 (A2) days after
planting, dap; and MS + 2 mg/L 2.4-D + 0.5 mg/L kinetin at 30 (B1) and 60 (B2) dap. Scale
bar = 0.5 cm. Photo by M. Idris.
Xu et al. (2005) added | mg/L 2,4-D to promote grape callus formation, but this did
not initiate root formation. They also added 1 mg/L NAA and 0.25 mg/L BA which
also formed callus without root formation. Akbas et al. (2004) added 1 mg/L NAA to
the medium to induce shoot formation of grape.
Such rooted callus formation derived from in vitro manipulation can be a source
of material to propagate the host plants of Rafflesia, and with which to investigate in
vitro infection by Rafflesia. Zhou et al. (2004) used Brassica napus L. as an in vitro
host material for infection by parasitic Orobanche L.. Kusumoto et al. (2007) used the
root of Trifolium pratense L. for infection by Orobanche minor Sutton in their studies
of the interaction between parasitic and host.
Conclusions
The basic MS medium augmented with 0.5 mg/L kinetin was effective in promoting in
situ seed germination of Tetrastigma rafflesiae. The formation of rooted callus can be
produced on MS medium by adding 2 mg/L NAA + 0.5 mg/L kinetin.
504 Gard. Bull. Singapore 63(1 & 2) 2011
ACKNOWLEDGEMENTS. The authors wish to thank the Herbarium Andalas (ANDA) of
Andalas University for help in identifying 7etrastigma rafflesiae used in this research, and the
Head of the Plant Physiology Laboratory for use of the facilities.
References
Akbas, F.A., Isikalan, G., Kara, Y. & Basaran, D. (2004) Comparison of the proliferation
of lateral buds of Vitis vinifera L. cv. Perle de Csaba during different periods of
the Year in in vitro conditions. Jnt. J. Agric. Biol. 6(2): 328-330.
Alizadeh, M., Singh, S.K. & Patel, V.B. (2010) Comparative performance of in vitro
multiplication in four Grape (Vitis spp.) rootstock genotypes. Jnt. J. Pl. Prod.
4(1): 41—S0.
Attenborough, D. (1995) The Private Life of Plants. London: BBC Books.
Jaskani, M.J., Abbas, H., Sultana, R., Khan, M.M., Qasim, M. & Khan, I.A. (2008)
Effect of growth hormones on micropropagation of Vitis vinifera L. cv. Perlette.
Pakistan J. Bot. 40(1): 105—109.
Klicova, S., Sebanek, J. & Vlasic, T. (2004) The effect of cytokinins and other plant
hormones on the growth of cotyledonary axilars of Flax (Linum usitatissimum),
Sunflower (Helianthus annuus) and Pea (Pisum sativum). Pl. Soil Environ. 50(4):
182-187.
Kusumoto, D., Goldwasser, Y., Xie, X., Yoneyama, K., Takeuchi, Y. & Yoneyama, K.
(2007) Resistance of Red Clover (Trifolium pretense) to the root parasitic plant
Orobanche minor 1s activated by salicylate not jasmonate. Ann. Bot. 100: 537—
544.
Lemmens, R.H.M.J. (2003) Tetrastigma (Migq.) Planchon. In: Lemmens, R.H.M.J.
& Bunyapraphatsara, N. (eds) Plant Resources of South East Asia No. 12 (3)
Medicinal and Poisonous Plants, pp: 400-401.Bogor: Prosea Foundation.
Meier, W. (1997) Rafflesiaceae. Flora Malesiana Series I, Vol. 13: 1-42.
Nais, J. (2001) Rafflesia of the World. Sabah Parks. Kota Kinabalu: Sabah Parks
Trustees.
Salami, A., Ebadi, A., Zamani, Z. & Ghasemi, M. (2005) Improvement in apex culture
in an Iranian Grapevine (Vitis vinifera L. ‘Bidaneh Sefid’) through fragmented
shoot apices. /nt. J. Agric. Biol. 7(3): 333-336.
Sofiyanti, N., Mat-Salleh, K., Purwanto, D. & Syahputra, E. (2007) A note on
morphology of Rafflesia hasseltii Suringar from Bukit Tiga Puluh National Park,
Riau. Biodiversitas 9(1): 257-261.
Tischler, S.R., Polley, H.W., Johnson, H.B. & Penningtont, R.E. (2000) Seedling
response to elevated CO2 in five epigeal species. /nt. J. Pl. Sci. 161(5): 779-783.
Veldkamp, J.F. (2008) The correct name for the Zefrastigma (Vitaceae) host of
Rafflesia (Rafflesiaceae) in Malesia and a (not so) new species. Reinwardtia 12
(4): 261-265.
Xu, X., Lu, J., Ren, Z., Wang, H. & Leong, S. (2005) Callus induction and somatic
embryogenesis in Muscadine and Seedless Bunch Grapes (Vitis) from immature
ovule culture. Proc. Florida State Hort. Soc. 118: 260-262.
In vitro Tetrastigma germination and callus formation 505
Zhou, W.J., Yoneyama, K., Takeuchi, Y., Iso, S., Rungmekarat, S., Che, S.H., Sato, D.
& Joel, D.M. (2004) /n Vitro infection of host roots by differentiated calli of the
parasitic plant Orobanche. J. Exp. Bot. 55(398): 899-907.
Zuhud, E.A.M., Hernidiah, N. & Hikmat, A. (1999) Pelestarian Rafflesia hasseltii
Suringar di Taman Nasional Bukit Tiga Puluh Riau-Jambi. Media Konservasi
VI(1): 23-26.
ss
7 glcee aoe
it) yar
ad
Hau) Au.
» may
es ory
a
J
>
Pol 2
x2
5
Gardens’ Bulletin Singapore 63(1 & 2): 507-517. 2011 507
Comparative anatomy of Grammitidaceae genera
in Peninsular Malaysia
A.T. Nor-Ezzawanis
Forest Research Institute Malaysia, 52109 Kepong, Selangor, Malaysia
ezzawanis@frim.gov.my
ABSTRACT. Grammitidaceae is represented by 12 genera and 52 species in Peninsular
Malaysia. The rhizome morphology and anatomy of Peninsular Malaysian Grammitidaceae
were studied to determine whether it can be used as a supplementary character in generic
delimitation. Two types of rhizome (creeping dorsiventral or erect), three types of stipe
arrangement on the rhizome (in horizontal rows, in whorls or spiral) and two types of stele
(solenostele or dictyostele) were identified.
Keywords. Anatomy, Grammitidaceae, morphology, Peninsular Malaysia, rhizomes
Introduction
Grammitidaceae is an important family for rainforest biodiversity in tropical montane
regions with over 750 species worldwide (Parris 2003, 2010). For Peninsular Malaysia,
12 genera, namely, Acrosorus Copel., Calymmodon C.Presl, Chrysogrammitis Parris,
Ctenopterella Parris, Dasygrammitis Parris, Oreogrammitis (Copel.) Parris, Prosaptia
C.Presl, Radiogrammitis Parris, Themelium (T.Moore) Parris, Tomophyllum (E.Fourn.)
Parris, Scleroglossum Alderw. and Xiphopterella Parris and 52 species are currently
recognised (Parris 2007, 2010). Lately, the morphological and molecular aspects of
this family have been intensively studied (Ranker et al. 2003, 2004; Schneider et al.
2004).
The Grammitidaceae have been treated in various ways taxonomically. Tryon
and Tryon (1982) treated the entire Grammitidaceae as a single genus based on spore
and sporangial characters while others established natural groupings based on the
type of the rhizome, rhizome scales, stipe and rachis, types of frond hairs, venation
patterns, soral arrangement and sporangial ornamentation (Parris 1983, 1986, 1995,
1997, 1998; Ranker et al. 2004). Parris (1995) also pointed out that neither cladistic
nor cytological studies were helpful in the generic inter-relationships in the family:
reticulate evolution possibly is one of the causes of failure.
For Peninsular Malaysia, all previous studies on this family were mainly
based on morphological characters, ecology and phytogeography (Holttum 1955;
Parris 1986, 1995, 1997, 1998, 2001, 2003, 2007, 2010). Rakotondrainabe & Deroin
(2006) have shown that rhizome anatomy is a useful character in generic delimitation
and can improve understanding of phylogeny of the family. Hence a study on rhizome
anatomy of selected species of Grammitidaceae in Peninsular Malaysia was conducted
508 Gard. Bull. Singapore 63(1 & 2) 2011
in order to determine whether it can be used as a supplementary character in generic
delimitation.
Material and methods
Fresh specimens of the rhizome were used for anatomical study. A list of specimens
used is attached in Appendix A. Transverse sections 25—50 um thick were made using
the sliding microtome following the methods used by Sass (1958). These sections were
stained with safranin and alcian green, then mounted with euparal. Anatomical data
analysis was made with a Leica Diaplan microscope equipped with a CCTV camera.
The photomicrographs were obtained from the camera through a computer using the
Analysis software.
Results and discussion
Out of the 12 genera of the Grammitidaceae in Peninsular Malaysia, 11 were studied.
The genus Chrysogrammitis Parris was excluded due to the limitation in obtaining
fresh specimens. Two types of rhizome, three types of stipe arrangement and two types
of vasculature were identified in this study (Table 1).
Two types of rhizome were observed, 1.e., creeping dorsiventral and erect.
Creeping dorsiventral rhizomes were found 1n Ctenopterella(Fig. 1:A1), Dasygrammitis
(Fig. 1: A2), Oreogrammitis (Fig. 1: A3), Prosaptia (Fig. 1: A4) and Themelium (Fig.
1: AS) while erect rhizomes occur in Acrosorus (Fig. 2: Al), Calymmodon (Fig. 3:
Al), Radiogrammitis (Fig. 2: A2), Scleroglossum (Fig. 3: A3 & A4), Tomophyllum
(Fig. 2: A3) and Xiphopterella (Fig. 3: A4). In comparison with Hovenkamp (1990),
the results in the current study agreed with the creeping dorsiventral type of rhizome
in Prosaptia. However for Dasygrammitis, instead of the radial type of rhizome found
by Hovenkamp (1990), the current study found only the creeping dorsiventral type of
rhizome for both D. brevivenosa and D. fuscata.
Three types of stipe arrangement were found in this study:
1) In horizontal rows, which occur in Crenopterella (Fig. 1: Al & B1), Dasygrammitis
(Fig. 1: A2 & B2), Oreogrammitis (Fig. 1: A3 & B3), Prosaptia (Fig. 1: A4 & B4)
and Themelium (Fig. 1: AS & B5);
2) In whorls, in Calymmodon (Fig. 3: Al & B1), Scleroglossum (Fig. 3: A2 & B2; A3
& B3) and Xiphopterella (Fig. 3: A4 & B4); and
3) Spirally in Acrosorus (Fig. 2: Al & B1), Radiogrammitis (Fig. 2: A2 & B2) and
Tomophyllum (Fig. 2: A3 & B3).
In the spiral arrangement, transverse sections of the rhizome show that the stipes
were in fact arranged in a more-or-less “2+1” arrangement i.e. two stipes were
at the same stage of development but there is another stipe in an another stage of
development. This proved that the three stipes are not exactly a true whorl.
Anatomy of Peninsular Malaysian grammitid ferns
Table 1. Rhizome characteristics of Grammitidaceae taxa studied.
509
Species Rhizome habit Stipe arrangement Vasculature
Acrosorus friderici-et-pauli erect spiral solenostele
Acrosorus streptophyllus erect spiral solenostele
Calymmodon curtus erect whorls dictyostele
Ctenopterella blechnoides creeping dorsiventral horizontal rows solenostele
Dasygrammitis brevivenosa __ creeping dorsiventral horizontal rows solenostele
Dasygrammitis fuscata creeping dorsiventral horizontal rows solenostele
Oreogrammitis adspersa creeping dorsiventral horizontal rows solenostele
Oreogrammitis congener creeping dorsiventral horizontal rows solenostele
Oreogrammitis malayensis creeping dorsiventral horizontal rows solenostele
Oreogrammitis reinwardtii creeping dorsiventral horizontal rows solenostele
Prosaptia alata creeping dorsiventral horizontal rows solenostele
Prosaptia contigua creeping dorsiventral horizontal rows solenostele
Prosaptia obliquata creeping dorsiventral horizontal rows solenostele
Radiogrammitis holttumii erect spiral solenostele
Radiogrammitis multifolia erect spiral solenostele
Scleroglossum pusillum erect whorls dictyostele
Scleroglossum sulcatum erect whorls dictyostele
Themelium tenuisectum creeping dorsiventral horizontal rows solenostele
Tomophyllum subminutum erect spiral solenostele
Xiphopterella hieronymusii erect whorls dictyostele
Xiphopterella sparsipilosa erect whorls dictyostele
Two types of vasculature were identified, 1.e., solenostele and dictyostele,
similar to those results of Ogura (1972), Bishop (1988, 1989), Rakotondrainibe &
Deroin (2006). Ogura (1972) also described the type of vasculature in the grammitids
Polypodium subpinnatifidum (currently known as Radiogrammitis subpinnatifida
(Blume) Parris) as solenosteles. The other type, the perforated dictyostele, occurs
in Ctenopteris sodiroi (currently known as Melpomene sodiroi (Christ & Rosenst.)
A.R.Sm. & R.C.Moran), Grammitis sp. and Scleroglossum sp. The results of the
current study are similar to his study. The solenostele was found in Acrosorus (Fig. 2:
B1), Radiogrammitis (Fig. 2: B2), Tomophyllum (Fig. 2: B3) (all with erect rhizomes
and the stipes arranged spirally); and Ctenopterella (Fig. 1: B1), Dasygrammitis (Fig.
1: B2), Oreogrammitis (Fig. 1: B3), Prosaptia (Fig. 1: B4) and Themelium (Fig. 1: B5)
(all with creeping dorsiventral rhizomes and the stipes arranged horizontally). The
dictyostele occurs in Calymmodon (Fig. 3: B1), Scleroglossum (Fig. 3: B2 & B3) and
Xiphopterella (Fig. 3: B4) (all with erect rhizomes and the stipes arranged in whorls).
Regarding the similarity in type of vasculature between genera, there are
two probabilities. The first is that the similarities are possibly caused by the close
510 Gard. Bull. Singapore 63(1 & 2) 2011
.
Fig. 1. Rhizome habit (Al—5S) and anatomy (B1—5) in Grammitid species of the Solenostele
group with creeping dorsiventral rhizomes and stipes arranged in horizontal rows. Al & BI.
Ctenopterella blechnoides; A2 & B2. Dasygrammitis fuscata; A3 & B3. Oreogrammitis
congener; A4 & B4. Prosaptia alata; A5 & B5. Themelium tenuisectum. Bar equals 200 tm in
B1, B2; 500 um in B4 & BS and 1000 um in B3. Abbreviations used: S = stipe, R = root, FT =
frond trace, RT = root trace.
Anatomy of Peninsular Malaysian grammitid ferns 511
Fig. 2. Rhizome habit (Al—3) and anatomy (B1-—3) in Grammitid species of the Solenostele
group with erect rhizomes and stipes arranged spirally. Al & B1. Acrosorus friderici-et-pauli;
A2 & B2. Radiogrammitis multifolia; A3 & B3. Tomophyllum subminutum. Bar equals 200 um
in B1—B3. Abbreviations used: S = stipe, R = root, FT = frond trace, RT = root trace.
relationship between genera. For the solenostele group in the current study (Acrosorus,
Ctenopterella, Dasygrammitis, _Oreogrammitis, Prosaptia, Radiogrammitis,
Themelium and Tomophyllum); the close resemblance between certain genera, e.g.,
(Oreogrammitis and Radiogrammitis; Ctenopterella and Themelium; Dasygrammitis
and Jomophyllum) has been discussed by Parris (2007). The genera Radiogrammitis,
a2 Gard. Bull. Singapore 63(1 & 2) 2011
Fig. 3. Rhizome habit (Al—4) and anatomy (B1-4) in Grammitid species of the Dictyostele
group with erect rhizomes and stipes arranged in whorls. Al & B1. Calymmodon curtus; A2
& B2. Scleroglossum pusillum; A3 & B3. Scleroglossum sulcatum;, A4 & B4. Xiphopterella
hieronymusii. Bar equals 200 um in B2—B4; 500 um in B1. Abbreviations used: S = stipe, R =
root, FT = frond trace.
we)
Anatomy of Peninsular Malaysian grammitid ferns 51
Acrosorus and Tomophyllum share similarities in having solenosteles and stipes
arranged spirally around the rhizome. However, on the basis of examining the close
relationships between genera, the similarity of these characters does not agree with
their placement in the phylogenetic tree in Ranker et al. (2004), where Radiogrammitis
is placed near to Oreogrammitis and Themelium (due to its close resemblance to
Oreogrammitis 1n many characters except its erect rhizome and the absence of
rhizome scales in many species); while Zomophyllum falls within the same group
with Calymmodon and Scleroglossum. However, not all the genera with solenosteles
have been sampled for the cladistic analysis in Ranker et al. (2004). Four genera not
yet included in the cladistic analysis are Acrosorus, Ctenopterella, Dasygrammitis
and Xiphopterella. Hence, an explanation of the relationship between the genera in
the solenostele group can only be proved once these four genera are included in the
analysis. For the dictyostele group in the current study, the close relationship between
Calymmodon and Scleroglossum was shown in the cladistic tree by Ranker et al.
(2004), while the position of Xiphopterella in the same clade based on rhizome form
and hair types, which was hypothesised in Parris (2007), agreed with the rhizome
anatomy results.
The second probability is that the similarities might be due to a mechanical
factor such as stipe arrangement on the rhizome. Ogura (1972) mentioned in general
that the dictyostele occurs in species with overlapping leaf gaps. This is clearly seen
with the occurrence of dictyosteles in genera with erect rhizomes and stipes arranged
in whorls (Calymmodon, Scleroglossum and Xiphopterella); as well as the genera with
creeping rhizomes and stipes arranged horizontally (Ctenopterella, Dasygrammitis,
Oreogrammitis, Prosaptia and Themelium) that all have solenosteles. However, the
occurrence of solenosteles in some species with erect rhizomes where the stipes are
arranged spirally around the rhizome is not clearly understood. This is probably due to
the non-overlapping of leaf gaps due to the spiral arrangement.
Ogura (1972) gave a brief description of the rhizome anatomy of the
Polypodiaceae. The cortex and pith consist mostly of parenchyma. Sclerenchyma may
occur in various forms and arrangements. In the current study, sclerenchyma around the
vascular bundles was found in Acrosorus friderici-et-pauli (Christ.) Copel. (Fig. 4A),
Calymmodon curtus Parris (Fig. 4B), Scleroglossum pusillum (Blume) Alderw. (Fig.
4C), Xiphopterella hieronymusii (C.Chr.) Parris (Fig. 4D) (all with erect rhizomes) and
Oreogrammitis malayensis Parris (Fig. 4E) (with a creeping dorsiventral rhizome).
One similarity between these plants is that almost all (except Acrosorus friderici-et-
pauli and Xiphopterella sparsipilosa) are small-sized plants with the lamina length
less than 10 cm. Hence, the sclerenchyma may function as supporting tissue in the
plants.
Conclusions
Similarities in the type of vasculature among genera in the current study are seen as
the result of the arrangement of stipes on the rhizome. It is consistent throughout the
514 Gard. Bull. Singapore 63(1 & 2) 2011
Fig. 4. Rhizome sections with red staining highlighting sclerenchyma tissue surrounding the
vascular bundles. A. Acrosorus friderici-et-pauli,; B. Calymmodon curtus; C. Scleroglossum
pusillum; D. Xiphopterella hieronymusii; E. Oreogrammitis malayensis. Bar equals 200 tm in
A, C-E; 500 um in B.
study that overlapping of leaf gaps influences the formation of stele type. Whether
the similarities are the result of close relationship among genera, it is only proven in
certain cases, especially in the dictyostele group. For the solenostele group, a larger
sampling is needed in the DNA analysis before such relationships can be assessed.
ACKNOWLEDGEMENTS. This study was carried out for the Flora of Peninsular Malaysia
Project funded by the Ministry of Science, Technology and Innovation (MOSTI) through the
National Council for Scientific Research and Development (MPKSN), under Project No. 01-
Anatomy of Peninsular Malaysian grammitid ferns S15
04-01-0000 Khas 2 entitled ‘Safeguarding the Forest Plant Diversity of Peninsular Malaysia’.
My deepest thanks are to my supervisor, the late Prof. Kamarudin Mat-Salleh for his guidance,
and I am much grateful to the FRIM (Forest Research Institute Malaysia) Training Committee
for financial support for my M.Sc. course. I am indebted to Dr. Barbara Parris from the Fern
Research Foundation, New Zealand, for advice on the current status of the Grammitidaceae
and to Drs. R. Kiew, L.G. Saw, R.C.K. Chung and E. Soepadmo for help in preparing the
manuscript. Heartfelt thanks go to Mr. Ahmad Damanhuri Mohammed and Mr. Razali Jaman
from the UKMB (Universiti Kebangsaan Malaysia, Bangi) Herbarium for their guidance and
sharing and to Dr. Khatijah Hussein, Dr. Noraini Talib, Mr. Mohd. Ruzi Abdul Rahman and
Puan Hajah Samiah Haji Kadri from UKM for their kind guidance in the anatomical work. I
should also like to thank the Curators of the herbaria at SING (for loan of specimens), KLU,
UKMB, BM, K and L (for access to their collections).
References
Bishop, L.E. (1988) Ceradenia, a new genus of Grammitidaceae. Amer. Fern J. 78(1):
1-S.
Bishop, L.E. (1989) Zygophlebia, a new genus of Grammitidaceae. Amer Fern J.
79(3): 103-118.
Holttum, R.E. (1955) A Revised Flora of Malaya. Vol 2. Singapore: Government
Office.
Hovenkamp, P.H. (1990) The significance of rhizome morphology in the systematics
of the Polypodiaceous ferns (sensu stricto). Amer. Fern J. 80(2): 33-43.
Ogura, Y. (1972) Comparative Anatomy of Vegetative Organs of the Pteridophytes. Ed.
2. Berlin: Gebrider Borntraeger.
Parris, B.S. (1983) A taxonomic revision of the genus Grammitis Swartz
(Grammitidaceae: Filicales) in New Guinea. Blumea 29: 13-222.
Parris, B.S. (1986) Grammitidaceae of Peninsular Malaysia and Singapore. Kew Bull.
41(3): 491-517.
Parris, B.S. (1995) Generic delimitation in Grammitidaceae (Filicales). In Dransfield, J.,
Coode, M.J.E. & Simpson, D.A. (eds) Plant Diversity in Malesia III. Proceedings
of the Third International Flora Malesiana Symposium, p. 171-176. Kew: Royal
Botanic Gardens, Kew.
Parris, B.S. (1997) Themelium, a new genus of Grammitidaceae (Filicales). Kew Bull.
52: 737-741.
Parris, B.S. (1998) Chrysogrammitis, anew genus of Grammitidaceae (Filicales). Kew
Bull. 53: 909-918.
Parris, B. S. (2001) Taxonomy of Malesian Grammitidaceae in relation to ecology
and phytogeography. In: Saw, L.G., Chua, L.S.L. & Khoo, K.C. (eds) Taxonomy:
the Cornerstone of Biodiversity. Proceedings of the Fourth International Flora
Malesiana Symposium 1998. Kepong: Forest Research Institute Malaysia
(FRIM). Pp. 155-160.
Parris, B.S. (2003) The distribution of Grammitidaceae (Filicales) inside and outside
Malesia. Telopea 10(1): 451-466.
Parris, B.S. (2007) Five new genera and three new species of Grammitidaceae
(Filicales) and the re-establishment of Oreogrammitis. Gard. Bull. Singapore
516 Gard. Bull. Singapore 63(1 & 2) 2011
58(2): 233-274.
Parris, B.S. (2010) Grammitidaceae. In: Parris, B.S., Kiew, R., Chung, R.C.K., Saw,
L.G. & Soepadmo, E. (eds) Flora of Peninsular Malaysia Series 1: Ferns and
Lycophytes, Volume |. Malayan Forest Records No. 48. Kepong: Forest Research
Institute Malaysia. Pp. 131-206.
Rakotondrainibe, F. & Deroin, T. (2006) Comparative morphology and rhizome
anatomy of two new species of Zygophlebia (Grammitidaceae) from Madagascar
and notes on the generic circumscription of Zygophlebia and Ceradenia. Taxon
55(1): 145-152.
Ranker, T.A., Geiger, J.M.O., Kennedy, S.C., Smith, A.R., Haufler, C.H. & Parris, B.S.
(2003) Molecular phylogenetics and evolution of the endemic Hawaiian genus
Adenophorus (Grammitidaceae). Molec. Phylogen. Evol. 26(3): 337-347.
Ranker, T.A., Smith, A.R., Parris, B.S., Geiger, J.M., Haufler, C.H., Straub,
S.C.K. & Schneider, H. (2004) Phylogeny and evolution of grammitid ferns
(Grammitidaceae): a case of rampant morphological homoplasy. Jaxon 53(2):
415-428.
Saas, J.E. (1958) Botanical Microtechnique. Ed. 3. Calcutta: Oxford & IBH Publishing
Co:
Schneider, H., Smith, A.R., Cranfill, R., Hildebrand, T.J., Haufler, C.H. & Ranker,
T.A. (2004) Unravelling the phylogeny of polygrammoid ferns (Polypodiaceae
and Grammitidaceae): exploring aspects of the diversification of epiphytic plants.
Molec. Phylogen. Evol. 31: 1041-1063.
Tryon, R.M. & Tryon, A.F. (1982) Ferns and Allied Plants, with Special Reference to
Tropical America. New York: Springer.
Appendix A. List of specimens used for rhizome anatomy study. Bkt = Bukit (Malay for Hill),
FR = Forest Reserve; G = Gunung (Malay for Mount)
Taxa Specimens examined
Acrosorus friderici-et-pauli Nor Ezzawanis FRI 52526, Cameron Highlands, G Brinchang,
(Christ.) Copel. c. 1729 masl, 17 Jan 2007 (KEP).
Acrosorus streptophyllus Nor Ezzawanis FRI 52395, Cameron Highlands, G Berembun,
(Baker) Copel. c. 1500 m asl, 19 Aug 2006 (KEP).
Calymmodon curtus Parris — Nor Ezzawanis FRI 52530, FRI 52531, Cameron Highlands, G
Brinchang, c. 1729 m asl, 13 Jan 2007 (KEP).
Ctenopterella blechnoides | Nor Ezzawanis FRI 54527, G Ledang, c. 1225 m asl, 15 Jan
(Grev.) Parris 2008 (KEP).
Dasygrammitis brevivenosa Nor Ezzawanis FRI 52525, Cameron Highlands, G Brinchang,
(Alderw.) Parris c. 1729 masl, 13 Jan 2007 (KEP).
Dasygrammitis fuscata Nor Ezzawanis FRI 52420, Cameron Highlands, G Berembun,
(Blume) Parris c. 1500 m asl, 19 Aug 2006 (KEP), Nor Ezzawanis FRI 52521,
FRI 52527, Cameron Highlands, G Brinchang, c. 1729 m asl,
13 Jan 2007 (KEP).
Anatomy of Peninsular Malaysian grammitid ferns
Oreogrammitis adspersa
(Blume) Parris
Oreogrammitis congener
(Blume) Parris
Oreogrammitis malayensis
Parris
Oreogrammitis reinwardtii
(Blume) Parris
Prosaptia alata (Blume) H.
Christ.
Prosaptia contigua (G.
Forst.) C. Presl
Prosaptia obliquata
(Blume) Mett.
Radiogrammitis holttumii
(Copel.) Parris
Radiogrammitis multifolia
(Copel.) Parris
Scleroglossum pusillum
(Blume) Alderw.
Scleroglossum sulcatum
(Kuhn) Alderw.
Themelium tenuisectum
(Blume) Parris
Tomophyllum subminutum
(Alderw.) Parris
Xiphopterella hieronymusii
(C.Chr.) Parris
Xiphopterella sparsipilosa
(Holttum) Parris
S17
Nor Ezzawanis FRI 54556, G Belumut, c. 1000 m asl, 22 to 24
Jan 2008 (KEP).
Nor Ezzawanis FRI 52374, Cameron Highlands, G Brinchang,
c. 1729 m asl, 18 Aug 2006 (KEP), Nor Ezzawanis FRI 52443,
Cameron
Highlands, G Berembun, c. 1500 m asl, 19 Aug 2006 (KEP).
Nor Ezzawanis FRI 54539, G Belumut, c. 1000 m asl, 22 Jan
2008 (KEP).
Nor Ezzawanis FRI 52360, Cameron Highlands, G Brinchang,
c. 1729 m asl, 18 Aug 2006 (KEP), Nor Ezzawanis FRI 52546,
Cameron Highlands, G Brinchang, c. 1729 m asl, 14 Jan 2007
(KEP), Nor Ezzawanis FRI 52593, Genting Highlands, G Ulu
Kali, c. 1767 m asl, 16 Feb 2007 (KEP).
Nor Ezzawanis FRI 54604, Berembun FR, Bkt Lantai, 9 Apr
2008 (KEP).
Nor Ezzawanis FRI 52428, Bkt Larut FR, G Hijau, c. 1300
m asl, 20 Aug 2006 (KEP), 7-L. Yao FRI 55924, Cameron
Highlands, Parit Falls, 10 May 2007 (KEP).
Nor Ezzawanis FRI 52432, FRI 52433, Bkt Larut FR, G Hyau,
c. 1300 m asl, 20 August 2006 (KEP), Nor Ezzawanis FRI
54460, Cameron Highlands, Ulu Telom, G Siku, c. 1486 m asl,
21 May 2007 (KEP).
Nor Ezzawanis FRI 54475, Cameron Highlands, Ulu Telom, G
Siku, c. 1486 m asl, 21 May 2007 (KEP).
Kueh H.L. FRI 52544, Taman Negara, G Tahan, 3 to 4 Feb
2007 (KEP)
Nor Ezzawanis FRI 52592, Genting Highlands, G Ulu Kalli, c.
1767 m asl, 16 Feb 2007 (KEP).
Nor Ezzawanis FRI 52582, Genting Highlands, G Ulu Kalli, c.
1767 m asl, 16 Feb 2007 (KEP).
Nor Ezzawanis FRI 52591, Genting Highlands, G Ulu Kalli, c.
1767 m asl, 16 Feb 2007 (KEP).
Nor Ezzawanis FRI 54482, Cameron Highlands, Ulu Telom, G
Siku, c. 1486 m asl, 21 May 2007 (KEP).
Nor Ezzawanis FRI 54464, Cameron Highlands, Ulu Telom, G
Siku, c. 1486 m asl, 21 May 2007 (KEP).
Kueh H.L. FRI 52560, Taman Negara, G Tahan, 3 to 5 Feb
2007 (KEP)
ll i WG
ai" as? ib ae
aA
ant Atha :
J enciin
my =.
Gardens’ Bulletin Singapore 63(1 & 2): 519-526. 2011 519
Ethnomedicinal study of the
Sundanese people at the Bodogol area,
Gede Pangrango Mountain National Park, West Java
Vera Budi Lestari Sihotang
Herbarium Bogoriense, Botany Division, Research Center for Biology,
Indonesian Institute of Sciences, Cibinong Science Center (CSC),
Jl. Raya Jakarta-Bogor Km 46, Cibinong, Bogor 16911, Indonesia
vera002(@lipi.go.id, verbudl@yahoo.com
ABSTRACT. Traditional medicine is often considered adequate even in present times,
especially when modern medical treatment is diffficult to obtain. Indonesia, a country with
rich biodiversity and a multicultural society, has a wealth of medicinal plant knowledge.
Observations on the ethnomedicinal practices of the Sundanese people, conducted in several
villages (Cipeucang, Ciwaluh, Lengkong Girang, Lengkong Hilir and Sungapan) around the
Bodogol area in West Java, are summarised.
Keywords. Bodogol, ethnomedicine, Gede Pangrango Mountain National Park, Indonesia,
medicinal plants, Sundanese people, West Java
Introduction
In traditional societies worldwide, plants feature importantly in the treatment of
many ailments, in particular infectious and parasitic diseases, diarrhoea, fever and
colds, as well as in birth control and dental hygiene. According to the World Health
Organisation (WHO), traditional medicine is a rather vague term for distinguishing
any ancient or culturally based health-care system from orthodox scientific medicine,
or allopathy. This includes systems that are currently regarded as indigenous or
unorthodox, alternative, folk, fringe, and even “unofficial”. Both the major Asian
systems (e.g., the Chinese, Ayurvedic, Unani, Unani Tibb medical systems), which are
comparatively well documented since ancient times, as well as the less widespread,
largely orally transmitted practices of other traditional communities, are included in
this understanding of traditional medicine. It is estimated (Farnsworth 1994, Cotton
1996) that about 64% of the world’s population depends on some form of traditional
medicine for their health care needs.
It was not until the 1970’s that a more organised study of traditional medicine
and herbal remedies received due academic attention. The World Health Organisation,
seeing the success of Chinese heterox healthcare programmes, began to encourage
studies and approaches that aimed at combining traditional and Western resources.
Effective ethnomedicinal research should consider related aspects, such as traditional
concepts of the origins of disease and local perceptions of the efficacy of particular
520 Gard. Bull. Singapore 63(1 & 2) 2011
treatments, applicable even when there is detailed documentation already available,
and certainly for orally transmitted practices. The study of ethnomedicine in a cultural
context is largely medical ethnobotany, involving the identification of botanical
species used in traditional remedies and even understanding folk classification of
medicinal plants (Balick & Cox 1996). Frequently, ethnobotanical inventories of plant
species used in healing are drawn up from interviews or surveys based on collection of
voucher specimens in a given area (Jain & Mudgal 1999). Some studies are based on
the identification of medicinal plants sold in markets.
Treatments with Old World medicinal plants were mentioned in the Materia
Medica of Dioscorides in the first century BCE. As scientists of western civilisations
became more interested in medicinal plants from Asia, Jacobus Bontius who worked
for the VOC (East India Company) conducted research on Javanese medicinal plants
in the 1600s, and the German naturalist Rumphius, who became deeply involved in the
study of plants and animals in Indonesia in the 1700s, also documented plants known
as cures. In the 1800s, during the British occupation, Thomas Horsfield assisted Raffles
in studying medicinal plants of Central and East Java. Their work inspired others, who
became engaged in ethnobotany (see Klokke 1998).
In fact, the study of medicinal plants should be emphasised for Indonesia,
which has a rich biodiversity and a multicultural society with long-standing knowledge
in the different forms of ethnomedicine and plant use. We carried out research into the
knowledge of medicinal plants in several villages around the villages of Bodogol, viz.,
Lengkong Girang, Ciwaluh, Cipeucang, Lengkong Hilir, and Sungapan. The study
was conducted from November 18th to 22th 2008.
Study site and overview of the Sundanese people
West Java occupies a total area of 35,746.026 square kilometres, consisting of 16
counties and 9 cities. West Java is known for its fertility and there is still an active
volcano. The region ranges generally from flat to undulating, to hilly and mountainous
in parts. Annual rainfall exceeds 2000 mm. The villages in the Bogor regency where
research was conducted include Sungapan village, Lengkong Hilir, Ciwaluh, Lengkong
Girang (a part of Wates Jaya village), and Cipeucang (including rural areas in Buncir
Sand). These villages engage much in rice cultivation, but medicinal plants can still be
found in the wild or are planted.
As a tribe, the Sundanese were forerunners of civilisation established in the
Indonesian archipelago, beginning with the founding of Salakanagara, the oldest
kingdom in Indonesia. Descendants of this Sunda Kingdom have founded other great
kingdoms in the archipelago, including Sriwijaya, Majapahit, Mataram Kingdom,
the Kingdom of Cirebon, and the Kingdom of Banten. Sunda is the culture of the
people who live in West Java. The Sundanese believe that they should possess the
ethos or character for a virtuous life in their culture, often described as Kasundaan
(““Sundanese-ness”’). Many Sundanese expressions representing good, white, clean, or
bright exist; and the moral characteristics of cageur (Sundanese for ‘healthy’), bageur
(good), bener (right), singer (introspective), and pinter (intelligent) have existed for
over a thousand years.
i)
Sundanese ethnomedicinal plant knowledge, Java 5
Sundanese people traditionally eat various kinds of plants and seeds obtained
from their gardens or fields. Commonly available plants that are sometimes used for
healing are called /alap.
Methods
Medicinal plant species were collected in five villages around Bodogol, and for 80%
of these, uses were documented through surveys and interviews (Table 1). Sometimes,
different local names for the same medicinal plant species were available from
different persons, and different tribes; this was also checked through further interaction
with the local shaman (paraji). The results obtained include those from the interview
and persons who have expertise in treatment, as well as elderly individuals who are
considered knowledgeable about medicinal plants. In recording the plant names and
uses, herbarium specimens were also taken for identification and as vouchers. Notes
on plant use also included the plant parts that were taken. A local resident served as
translator during interviews. Informants were asked to describe their knowledge about
traditional medicine; how medicinal plants were prepared for particular treatments;
how their knowledge was acquired; and how they disseminated their skills. The
literature was also consulted where possible.
Results and discussion
Perception about illnesses
Based on the interviews, the Sundanese perception is that disease can be caused by
behaviours which lead to imbalance in the body elements, in addition to diseases
caused by the supernatural, such as spirits. Imbalance in the body can be due to poorly
balanced diet or excessive and uncontrolled human emotion such as fear, hate or joy.
Members of the community who were interviewed accept that a person is truly
ill and in pain only when he is suffering from chronic disease or other health problems
that cause work activities to be disrupted. Even if a person has minor problems such as
colds, he is not considered ill if his work routine 1s not disrupted.
Plants in traditional medicine
The theory of disease (Foster & Anderson 1978) includes belief in natural health, causes
of disease and different types of medicine and healing techniques. In the traditional
theory of disease, pain is usually accepted as a consequence of some taboo being
broken, or an imbalance between hot and cold elements in the body. Local knowledge
is that which is traditionally owned and developed by a community in response to,
and in interaction with, its environment. According to Koentjaraningrat (1989), each
culture has a complex set of knowledge about nature, and the plants, animals, objects
and people around them that are abstracted into the concept, theory, and establishment
(koentjaraningrat). Traditional healers, in this case a shaman, datu, or teacher, provide
ye 72 Gard. Bull. Singapore 63(1 & 2) 2011
explanations and interpretations about the illness and the use of materials or herbs for
treating or curing diseases.
In this paper, the results represent part of the local knowledge of medicinal
plants of the Sundanese people in five villages around Bodogol. Basically, people in
all villages visited are still very dependent on the plants around them. In addition to
medicine, they also use it as animal feed for such as goats and sheep, for firewood,
and also to make household appliances. Among the five villages visited, three essential
aspects of traditional treatment using medicinal plants were noted. First was the
prominence of treatment after, or the recovery from, childbirth, which is usually done
by the paraji. Second was the existence of persons who believed that they had the
ability to cure diseases. Third was that treatment typically made use of plants around
the community, which can be easily obtained.
Treatment techniques are varied. There are plants that can be eaten as part of
the treatment. Sometimes, plant parts are brewed with hot water, grated and kneaded
and then applied onto an affected body part to be treated. Another use is by mixing
with other ingredients and then boiling to obtain a concoction or infusion that 1s taken
orally. A fourth way is to incorporate a medicinal herb into food dishes, for example,
by frying in mixture with egg. One other way is by soaking the plant material and
bathing with the infusion. Knowledge of such treatment methods is obtained from
parents and the community at large.
The 80 medicinal plants recorded around the five villages are listed in Appendix
A. An important use of medicinal plants is for recovery after childbirth. Often, a
woman who has just given birth will make a godogan consisting of roots and leaves of
buntiris (Kalanchoe pinnata Pers., Crassulaceae), babadotan (Ageratum conyzoides
L.), daun rane (Sellaginela plana (Desv.) Hieron.) and jawer kotok (Plectranthus
scutellarioides (L.) R.Br.). Plants that are often mentioned by informants include
buntiris and kumis kucing (Ortosiphon aristatus (Blume) Miq., Lamiaceae). Being
easy to obtain, buntiris is still often used to treat fever in children, while kumis kucing
is believed to cure various diseases. In these five villages, most community gardens are
planted with kumis kucing, which is sold to cities.
For 46 medicinal plant species recorded in Lengkong Girang Village, the uses
were for curing or treating fever (15 species), sprue (1 species), pain in the digestive
tract (3 types), high blood pressure (2 types), decreased vitality (8 types), skin diseases
(3 types), stroke (3 kind), injuries due to knives (1 species), toothache (1 species),
recovery after childbirth (5 types), and ambayen (1 species). These mainly utilised
leaves (37 species), roots (3 types), seeds (1 species), and tubers (4 types).
In Ciwaluh Village, 23 species of medicinal plants were recorded, used for
treating fevers (3 types), recovery after childbirth (8 types), diarrhea (1 species), knife
wounds (1 species), nausea and convulsions (1 species), pain when defecating and
urinating (2 types), hypertension (1 type), and as an “energy booster” (2 types) and
general medicine for all diseases (1 species). The plant parts used were leaves (17
species), roots (1 species), bulbs (2 types), and stems (2 species).
At Cipeucang Village, 29 species of medicinal plants were documented,
for treating bone fractures (4 types), diabetes (1 type), kidney problems (2 types),
Nn
i)
Lo
Sundanese ethnomedicinal plant knowledge, Java
cancer (2 types), low vitality (2 types), fevers (2 species), recovery after childbirth
(7 species), coughs (2 types), abdominal pain (2 types), constipation (1 types), skin
diseases (1 species), knife wounds (2 types); also generally for all diseases (3 types),
and for casting out spirits from young children (1 species). The parts used were leaves
and flowers.
From Lengkong Hilir Village, 16 species of medicinal plants were recorded.
These were for curing fevers (3 types), pain in the digestive tract (4 types), hypertension
(2 types), sprue (1 type), heatiness (1 species), low vitality (5 types), skin diseases (1
type), back pain (4 types), fatigue (3 types), and toothache (1 species).
At Sungapan Village, 17 species of medicinal plants were recorded, used for
treating fevers (1 species), pain in the digestive tract (2 types), gout (4 types), diabetes
(4 type), low vitality (4 species), recovery after childbirth (1 species), bleeding after
childbirth (3 types), dizziness (1 species), cough (1 species), anaemia in women (1
species), and sore eyes (1 species). All these utilised only leaves.
At the present time with more advanced medical facilities and many kinds
of medicines commercially available shops, drugstores and pharmacies, it is possible
that there should be increasing erosion of medicinal plant knowledge. However, in
many cases, people still remember traditional herbal treatments they were taught by
their parents. The interviews verified that old people knew more about what plants are
used as medicines than younger people, and clearly parents who used medicinal plants
attempted to pass this knowledge down to their children. According to Ms. Pupu, a
resident of Ciwaluh Village, if someone had knowledge that is not utilised, they will
not be rewarded after death.
The people who still use medicinal plants do not strongly believe in modern
drugs. It is also for economic reasons that they use medicinal plants, which are cheaper
and easier to obtain. Conversely, those who have switched over to modern medicine
say that modern medicine heals faster and was more convenient because it does not
require any processing by the user beforehand. Medicinal plants are usually the “first
aid” or a “last resort”. This can be gathered from the interviews with Ms. Yanti in
Cipeucang Village, who suffered from chronic cough. She visited the doctor but the
cough did not go away, and eventually tried capituheur (Mikania cordata (Burm.f.)
B.L.Rob.), a plant believed to cure coughs. After drinking an infusion regularly, the
coughing ceased. In contrast, Mr. Ujang, a resident of Lengkong Hilir Village, took
hantap (Sterculia rubiginosa Vent.) as a kind of “first aid” for a toothache.
Conclusion
Knowledge of medicinal plants in the five villages studied appears to be still substantial,
but perhaps just changing. It can be sensed, however, that when there is reduced
dependency on medicinal plants, knowledge of plant uses will become increasingly
reduced. Culture is dynamic and growing, and more changes will definitely come
(Naranjo 1997). There is certainly an urgent need to document this knowledge.
524 Gard. Bull. Singapore 63(1 & 2) 2011
References
Balick, M.J. & Cox, P.A. (1996) Plants, People, and Culture, The Science of
Ethnobotany. New York: Scientific American Library.
Cotton, C.M. (1996) Ethnobotany: Principles and Applications. Chichester: John
Wiley & Sons.
Farnsworth, N.R. (1994) Ethnopharmacology and Drug Development. In: Prance, G.T.
(ed) Ethnobotany and the Search for New Drugs. Chichester: John Wiley & Sons.
Foster, G.M. & Anderson, B. (1978) Medical Anthropology. New York: Wiley.
Jain, S.K. & Mudgal, V. (1999) A Hand Book of Ethnobotany. India: Bishen Singh
Mahendra Pal Singh.
Klokke, A.H. (ed) (1998) Traditional Medicine Among the Ngaju Dayak in Central
Kalimantan: The 1935 Writing of a Former Ngaju Dayak Priest. Borneo Research
Council, Inc.
Koentjaraningrat, R.M. (1989) Javanese Culture. U.S.A.: Oxford University Press.
Naranjo, P. (1997) The Urgent Need for the Study of Medicinal Plants. In: Schultes,
R.E. & Reis, S. von (eds) Ethnobotany: Evolution of a Discipline. Oregon:
Dioscorides Press.
Appendix A. List of 80 medicinal plant species recorded in the Bodogol area, West Java. The
villages studied are Cipecang (CP), Ciwaluh (CW), Lengkong Girang (LG), Lengkong Hilir
(LH) and Sungapan (S). A ‘+’ indicates that the species was recorded at a particular village.
Local Name
Akar eurih
Species, Family
Imperata cylindrica, Poaceae
CP’ “CW_EG_ EHESs
+ + +
Alpukat Persea americana, Lauraceae oF Ste
Antanan Centella aciatica, Apiaceae ap
Babadotan Ageratum conyzoides, Asteraceae aE +
Babawangan Peperomia pelucida, Piperaceae is
Baluntas Clerodendrum buchananii, Lamiaceae +
Bawang merah Allium cepa, Amaryllidaceae a
Bawang putih Allium sativum, Amaryllidaceae +
Bijangut Mentha arvensis, Lamiaceae +
Brotowali Tinospora crispa, Menispermaceae +
Buntiris Kalanchoe pinnata, Crassulaceae +
Capituheur Mikania cordata, Asteraceae +
Cecenet Physalis minima, Solanaceae F ie SF
Cengkeh Syzygium aromaticum, Myrtaceae a
Daun calincing Oxalis corniculata, Oxalidaceae +
Daun cengek Capsicum frutescens, Solanaceae
Daun jambu Psidium guajava, Myrtaceae ar
Daun kapuk Ceiba pentandra, Malvaceae +
Sundanese ethnomedicinal plant knowledge, Java
Daun katomas
Daun kopi
Daun mangga
Daun paria
Daun penurun tensi
darah tinggi
Daun putri malu
Daun rambutan
Daun rane
Daun saga
Daun sembung
Daun seureuh
Daun sirih
Daun sirsak
Daun waru
Daun wera hijau dan
merah
Gedang ganul
Gedebong
Hantap
Jambe/pinang
Jawer kotok
Jotang
Kaca piring
Kahitutan
Karas tulang
Katomas
Kaworo
Kecubung merah dan
putih
Keluwih
Ki beling
Ki koneng
Ki korejat
Ki rapet
Ki saat
Ki senok
Ki tajam
Ki urat
Kirinyuh pait
Kumis kucing
Kunci
Euphorbia heterophylla, Euphorbiaceae
Coffea arabica, Rubiaceae
Mangifera indica, Anacardiaceae
Momordica charantia, Cucurbitaceae
Pilea microphylla, Urticaceae
Mimosa pudica, Leguminosae
Nephelium lappaceum, Sapindaceae
Sellaginela plana, Sellaginellaceae
Abrus precatorius, Leguminosae
Blumea balsamifera, Asteraceae
Cymbopogon citratus, Poaceae
Piper betle, Piperaceae
Annona muricata, Annonaceae
Hibiscus tiliaceus, Malvaceae
Hibiscus rosa-sinensis, Malvaceae
Carica papaya, Caricaceae
Piper umbellatum, Piperaceae
Sterculia rubiginosa, Sterculiaceae
Arecha catechu, Arecaceae
Plectranthus scutellarioides, Lamiaceae
Spilanthes iabadicensis, Asteraceae
Gardenia jasminoides, Rubiaceae
Lasianthus inodorus, Rubiaceae
Turpinia montana, Staphyleaceae
Euphorbia heterophylla, Euphorbiaceae
Abelmoschus moschatus, Malvaceae
Datura metel, Solanaceae
Artocarpus camansi, Moraceae
Sericocalyx crispus, Acanthaceae
Arcangelicia flava, Menispermaceae
Laurentia longiflora, Campanulaceae
Ficus villosa, Moraceae
Artemisia vulgaris, Asteraceae
Abelmoschus manihot, Malvaceae
Clinacanthus nutans, Acanthaceae
Plantago major, Plantaginaceae
Eupatorium inulifolium, Asteraceae
Ortosiphon aristatus, Lamiaceae
Boesenbergia rotunda, Zingiberaceae
+
fo
+++ 4+
+
+
+
525
526 Gard. Bull. Singapore 63(1 & 2) 2011
Kunyit Curcuma longa, Zingiberaceae ar
Lampuyung Symphytum officinale, Boraginaceae se aR a5
Lempuyang Zingiber aromaticum, Zingiberaceae +
Lengkuas Alpinia galanga, Zingiberaceae aa
Monyenyen Laurentia longiflora, Campanulaceae +
nanangkaan Euphorbia hirta, Euphorbiaceae 3
Panglay Zingiber ottensii, Zingiberaceae ar ar
Pecah beling Strobilanthes crispus, Acanthaceae 35 oe
Pungpulutan Urena lobata, Malvaceae 4
Putat Planchonia valida, Lecythidaceae +
Rasamala Altingia excelsa, Hamamelidaceae aF
Remek daging Excoecaria cochinchinensis, + ar
Euphorbiaceae
Rendeu Staurogyne elongata, Acanthaceae ae =
Sambiloto Andrographis paniculata, Acanthaceae +
Sanagori Sida rhobifolia, Malvaceae + + -
Sawi langit Vernonia cinerea, Asteraceae =F
Seureuh Cymbopogon citratus, Poaceae ap +
Sintrong Erechites valerianifolia, Asteraceae a) ats
Tapakliman Pseudoelephanthopus spicatus, sr
Asteraceae
Tawulu Premna obongata, Lamiaceae + <5
Terong kori Solanum aculeatissimum, Solanaceae +
Walang Wedelia biflora, Asteraceae Se RE +H
Instructions for contributing authors (continued)
Title and authorship. The title should concisely describe the contents. If a scientific name is used, its
authority is normally excluded, but the family name would be provided. Authors’ names, affiliations
and postal / e-mail addresses are stated below the title. If more than one author, indicate “corresponding
author’. Avoid footnotes. A short running title (up to six words) should also be provided.
Abstract. The abstract is at most 100-300 words. It should concisely indicate the article’s contents without
summarising it; mentioning novelties and name changes. Keywords: Suggest at most eight keywords, in
alphabetical order.
Scientific names and author abbreviations. Genus and species names of organisms must be italicised
and followed by the authority (with family name in parentheses) when first mentioned in the text or
diagnoses. Standards for author abbreviations include:
Brummitt, R.K. & Powell, C.E. (1992) Authors of Plant Names. Kew: Royal Botanic Gardens, Kew.
Kirk, P.M. & Ansell, A-E. (1992) Authors of Fungal Names [Index to Fungi Supplement]. Wallingford:
CAB International.
Herbarium abbreviations (http://sciweb.nybg.org/science2/IndexHerbariorum.asp) follow
Holmgren, P.K., Holmgren, N.H. & Barnett, L.C. (eds) (1990) Index Herbariorum. Part I: The Herbaria
of the World [Regnum Veg. vol. 120]. New York: New York Botanical Garden.
Journal and book title abbreviations. For journals:
Bridson, G.D.R., Townsend, S.T., Polen, E.A. & Smith, E.R. (eds) (2004) BPH-2: periodicals with
botanical content: comprising a second edition of Botanico-Periodicum-Huntianum. Pittsburg: Hunt
Institute for Botanical Documentation.
For books:
Stafleu, F.A. & Cowan, R.S. (eds) (1976-88) Taxonomic Literature. 2nd edition. 7 vols [Regnum Veg.
vols 94, 98, 105, 110, 112, 115, 116].
Stafleu, F.A. & Mennega, E.A. (eds) (1992—) Taxonomic Literature. Supplements [Regnum Veg. vols
12521305 1:32):
(A useful source of verifying names of publications is
<http://asaweb.huh.harvard.edu:8080/databases/publication_index.html>. If in doubt, list full titles.
Other abbreviations and units of measurement. \f using standard abbreviations and acronyms, give the
full term on first mention. Dates are cited as: 1 Jan 2000. SI (metric) units of measurement are used and
spelled out except when preceded by a numeral; they are abbreviated in standard form: g, ml, km, etc.
Tables. Tables are numbered in arabic numerals in the order they are first mentioned in the text and carry
an indicative legend at the head. Tables are given at the end of the manuscript.
Illustrations. All drawings, maps, graphs and photographic images (individually or collected in a plate)
are to be numbered in arabic numerals in the order they are first mentioned in the text, as Fig. 1, Fig. 2,
etc. (plate components would be referred to in the text as Fig. 1A, 1B, Fig. 1A—D, etc.). If relevant, scale
bars should be used to indicate magnification.
When grouping photographs, the maximum page area 19.5 x 13 cm must be heeded. High resolution digital
images may be submitted as separate files (line drawings in black and white at 600 dpi, photographs at
300 dpi) sent electronically or in a CD. Do not embed images into the main text file.
References in the text. Citation in the text should take the form: King & Gamble (1886) or (King &
Gamble 1886), or King et al. (1886) if more than three authors to a work. Use 2000a, 2000b, etc. if
several papers by the same author(s) in one year are cited.
References listed at the end. There, works mentioned in the text are listed alphabetically as follows:
Dallwitz, M.J.. Paine, T.A. & Zurcher, E.J. (1999) User’s Guide to the DELTA Editor. http:
biodiversity.uno.edu/delta/ (accessed on 2 Aug. 2010).
Persson, C. (2000) Phylogeny of Gardenieae (Rubiaceae) based on chloroplast DNA sequences from the
rps16 intron and trnL(UAA)-F(GAA) intergenic spacer. Nordic J. Bot. 20: 257-269.
Ridley. H.N. (1930) The Dispersal of Plants Throughout the World. Ashford, U.K.: L. Reeve.
Smith, A.C. & Darwin, S.P. (1988) Rubiaceae. In: Smith, A.C. (ed) Flora Vitiensis Nova, A New Flora
of Fiji 4: 143-193.
References to web-based resources should include either a doi (digital object identifier) specification
or full URL mentioning also the date it was accessed. Use of DNA sequences from GenBank should be
acknowledged and the studies for which the sequences were generated should be cited.
Style of nomenclatural summaries. The following style is required:
Gardenia anisophylla Jack ex Roxb., Fl. Ind. ed. Carey & Wall. 2: 561 (1824).
Medinilla alternifolia Blume, Mus. Bot. 1: 19 (1849).
If authors include full bibliographic data for these works in the list of references at the end of the paper,
they should also be mentioned in the text briefly, e.g., “Nomenclatural references researched include
Blume (1849) and Roxburgh (1824).”
Homotypic synonyms should be provided in a block, stating the type at the end.
Front cover picture: Shorea albida canopy, Brunei Darussalam (Photo by K.M. Wong)
Back cover picture: Kopsia singapurensis, Singapore (Photo by Y.W. Low)
Flora Malesiana
was founded by Professor C.G.G.J. van Steenis (1901-1986)
who was enthralled by the vegetation and flora of the Malay Archipelago
and sought the most logical platform for an organised documentation
of the fascinating and valuable plant life of this region.
The Flora Malesiana Symposium was first held in Leiden
in 1989 to commemorate this eminent scholar, and is held once
every three years. The 8th Symposium was hosted by the
Singapore Botanic Gardens on 23-27 August, 2010.
From its initial objective, to obtain taxonomic documentation of the enormous botanical heritage
of Malesia (the distinct botanical region recognised by van Steenis, including the Malay Peninsula,
Sumatra, Java, the Lesser Sundas, Borneo, the Philippines, Sulawesi, the Moluccas, New Guinea
and smaller associated islands), the Flora Malesiana has come to represent an umbrella project that
includes not only revisions and flora writing. It now motivates and receives contributions from all
kinds of taxonomic and systematic work, involving identification, inventories and other biodiversity
assessments, checklists, and other biological studies on plants and their interactions. It is a huge flora
and a vast endeavour, one which has attracted the interest of many countries, especially those in the
flora area (Brunei, Indonesia, Malaysia, Papua New Guinea, Philippines, Singapore, Thailand), and
neighbouring territories spanning the Indo-Burmese, East Asian and Indo-Chinese regions across to the
Southwest Pacific and Australian regions, as well as institutions and individuals worldwide who wish
to participate in discovering and documenting one of the richest floras on Earth.
NATIONAL
PARKS
© BOTANY
Singapore Botanic Gardens, 1 Cluny Road, Singapore 259569 Tel: 64719921 Telefax: 64674832