UNIVERSITY OF

ILLINOIS LIBRARY

AT URBANA-CHAMPAIGN

BIOLOGY

'2001

5otany

EW SERIES, NO. 39

L Revision of the South American

Ipecies of Brunfelsia (Solanaceae)

imothy C. Plowmant

, Knapp and J. R. Press, Editors

ugust 31, 1998

ublication 1496

UBRAKY

wia.

Information for Contributors to Fieldiana

•

•

i

•

•

Referenci

cs for

-

•

uton Publ,-

..

'

,

and

•

Only

-.

>ciper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).

FIELDIANA

Botany

NEW SERIES, NO. 39

A Revision of the South American

Species of Brunfelsia (Solanaceae)

Timothy C. Plowmant

Department of Botany

Field Museum of Natural History

Roosevelt Road at Lake Shore Drive

Chicago, Illinois 60605-2496 U.S.A.

S. Knapp and J. R. Press, Editors

Department of Botany

The Natural History Museum

London SW7 5BD

England

Accepted July 2, 1996

Published August 31, 1998

Publication 1496

PUBLISHED BY FIELD MUSEUM OF NATURAL HISTORY

© 1998 Field Museum of Natural History

ISSN 0015-0746

PRINTED IN THE UNITED STATES OF AMERICA

n,S

Table of Contents

FOREWORD

Warfe Davis vii

EDITORS' PREFACE

S. Knapp and J. R. Press xi

I. MATERIALS AND METHODS 1

II. TAXONOMIC HISTORY OF THE GENUS 2

III. GENERIC RELATIONSHIPS 5

IV. GENERAL MORPHOLOGY 6

V. CYTOLOGY 16

VI. HYBRIDIZATION EXPERIMENTS 19

VII. ECOLOGY AND REPRODUCTIVE

BIOLOGY 20

VIII. TRENDS IN SPECIALIZATION 26

IX. GEOGRAPHICAL DISTRIBUTION AND

RELATIONSHIPS 29

X. TAXONOMIC TREATMENT 34

Brunfelsia Linnaeus 34

Artificial Key to the Sections, Species,

and Infraspecific Taxa of

Brunfelsia 35

Brunfelsia sect. Guianenses 38

1. Brunfelsia amazonica 38

2. Brunfelsia burchellii 40

3. Brunfelsia chocoensis 40

4. Brunfelsia clandestina 43

5. Brunfelsia guianensis 46

6. Brunfelsia martiana 49

Brunfelsia sect. Franciscea 51

1. Brunfelsia australis 53

8. Brunfelsia boliviano 57

9. Brunfelsia bonodora 59

10. Brunfelsia brasiliensis 61

lOa. Subspecies brasiliensis 62

lOb. Subspecies macrocalyx 66

1 1 . Brunfelsia chiricaspi 70

12. Brunfelsia cuneifolia 72

13. Brunfelsia dwyeri 74

14. Brunfelsia grandiflora 77

14a. Subspecies grandiflora 78

14b. Subspecies schultesii 80

15. Brunfelsia hydrangeiformis 84

15a. Subspecies hydrangeiformis 84

15b. Subspecies capitata comb, et

stat. nov. 86

16. Brunfelsia imatacana 89

17. Brunfelsia latifolia 91

18. Brunfelsia macrocarpa 93

19. Brunfelsia mire 95

20. Brunfelsia obovata 98

20a. Variety obovata 99

20b. Variety coriacea 101

2 1 . Brunfelsia pauciflora 103

22. Brunfelsia pilosa 108

23. Brunfelsia rupestris Ill

24. Brunfelsia uniflora 113

Dubious and Excluded Species 119

Nomina Nuda 119

XI. EDITORS' ACKNOWLEDGMENTS 119

XII. PLOWMAN'S ACKNOWLEDGMENTS 120

XIII. LITERATURE CITED 120

APPENDIX I: List of Exsiccatae 124

APPENDIX II: Brunfelsia sect. Brunfelsia 132

Description 132

Names of Brunfelsia sect. Brunfelsia 132

APPENDIX III: Additional References 133

INDEX . 134

List of Illustrations

Timothy Plowman vi

Brunfelsia hydrangeiformis subsp. hydran-

geiformis, type xii

1. Brunfelsia americana 3

2. Root of Brunfelsia grandiflora subsp.

schultesii 8

3 . Floral structure of Brunfelsia 11

4. Brunfelsia densifolia (Knapp s.n.) from

Puerto Rico 12

5. Brunfelsia lactea (Knapp s.n.) from

Puerto Rico 14

6. Fruit of Brunfelsia americana 17

7. Methona themisto, type specimen 24

8. Caterpillars of Methona confusa on

Brunfelsia chiricaspi leaf 25

9. Distribution of section Brunfelsia 31

10. Distribution of sections Guianenses

and Franciscea 32

1 1 . Brunfelsia amazonica 39

12. Brunfelsia burchellii 41

13. Brunfelsia chocoensis 42

14. Brunfelsia clandestina 44

15. Distribution of species in section

Guianenses 46

16. Brunfelsia guianensis 47

17. Brunfelsia martiana 50

18. Manaca (Brunfelsia uniflora) 52

19. Brunfelsia australis 55

20. Distribution of Brunfelsia australis,

B. boliviana, and B. bonodora 56

21. Brunfelsia boliviana 58

22. Brunfelsia bonodora, illustration of

lectotype 60

in

23. Distribution of Brunfelsia brasiliensis

and B. chiricaspi 63

24. Brunfelsia brasiliensis subsp. bras-

iliensis 64

25. Brunfelsia chiricaspi 71

26. Brunfelsia cuneifolia 73

27. Brunfelsia dwyeri 75

28. Distribution of Brunfelsia cuneifolia,

B. dwyeri, and B. grandiflora 77

29. Brunfelsia grandiflora subsp. grandiflo-

ra and subsp. schultesii 79

30. Brunfelsia hydrangeiformis subsp. hy-

drangeiformis, illustration of type 85

3 1 . Isotype of Brunfelsia hydrangeiformis

subsp. capitata stat. nov 87

32. Brunfelsia imatacana 89

33. Distribution of Brunfelsia hydrangei-

formis, B. imatacana, and B. mire 90

34. Brunfelsia latifolia, illustration of

type 92

35. Brunfelsia macrocarpa 94

36. Brunfelsia mire 96

37. Distribution of Brunfelsia latifolia, B.

macrocarpa, B. obovata, and B.

pilosa 99

38. Brunfelsia obovata var. obovata 100

39. Brunfelsia obovata var. coriacea 102

40. Brunfelsia pauciflora 104

41. Distribution of Brunfelsia pauciflora

and B. rupestris 105

42. Brunfelsia pauciflora cv. "Lindeniana" 109

43. Brunfelsia pilosa 110

44. Brunfelsia rupestris 112

45. Brunfelsia uniflora, illustration of type 114

46. Distribution of Brunfelsia uniflora 116

List of Tables

1 . Historical classification of Brunfelsia .... 6

2. Chromosome counts in Brunfelsia 18

3. Chromosome numbers and size in Brun-

felsia 19

4. Pollinators of Brunfelsia grandiflora 23

5. Brunfelsia species recorded as larval food

plants for species of Methona 26

6. Trends in specialization of Brunfelsia ... 30

7. Flowering and fruiting times of Brunfel-

sia uniflora 117

IV

Tim Plowman

Foreword

The idea for this monograph emerged from

somewhat unusual circumstances. On March 29,

1942, ethnobotanist and plant explorer Richard

Evans Schultes arrived at the Kofan Indian village

of Conejo, on the banks of the Rio Sucumbios in

the remote rain forests of southern Colombia. In

the midst of his first sojourn in the Amazon, with

the goal of surveying medicinal plants and arrow

poisons, Schultes found himself immersed in a

world of phytochemical wizardry unlike anything

he had ever known. Psychoactive and toxic plants

touched every aspect of the lives of the Kofan.

The most important and sacred of their plants was

the hallucinogen Banisteriopsis caapi, or yage,

the "vine of the soul." When Schultes inquired

how often the people imbibed yage, his infor-

mant's response suggested that the question had

no meaning. During illness, of course, and in the

wake of a death. In times of need or hardship. At

certain passages in life. When a young body of

six has his initial haircut, or when he kills for the

first time. And naturally, the shaman suggested, a

youth will drink yage at puberty, when his nose

and ears are pierced and he obtains the right to

wear the tail feathers of the macaw. As a young

man, he may drink it at his leisure to improve his

hunting technique or simply to flaunt his physical

prowess. The message that Schultes received was

that the Kofan took yage whenever they felt like

it. At least once a week, and no doubt on any

occasion that warranted it, such as the eve of his

own departure from Conejo a fortnight later. With

the people dancing, the men facing the women in

long lines moving forward and then back, turning

from side to side, the dancers stamping the ground

lightly in time with the drums, this solitary stu-

dent of plants, a man who in time would inspire

a generation of ethnobotanists, took yage with, as

he would recall 50 years later, "the whole damn

village."

The next morning, his face still painted and his

mind swirling with the sounds of chants, Schultes

paddled up the Sucumbios in a dugout canoe. His

head throbbed. Unbeknownst to him, the prepa-

ration had contained bark of tsontinb"k'o, the cold

fever tree, together with su-tim-ba-che, a root said

by the Kofan to cause a "drunk worse than yage."

Indeed, it had. Despite his discomfort, Schultes

made a note to investigate the admixtures at

length. Thirty years later he did, by dispatching

his protege, Timothy Plowman, to the Putumayo.

After 15 months in the field in South and Central

America and the Caribbean, Tim Plowman

emerged with a definitive study, a monographic

revision of the genus Brunfelsia, including de-

scriptions of several new species and varieties.

The plant that Schultes drank at Conejo and that

gave him such a headache is now known as Brun-

felsia grandiflora subsp. schultesii.

Tim Plowman approached botanical research

with the intensity and rigor of a scientist, the cour-

age of his mentor, and the flair of a poet. His first

Colombian excursion set the pattern of his career.

Arriving in Bogota, he traveled south to the valley

of Sibundoy, where he contacted Pedro Juajibioy,

a Kamsa Indian healer who as a boy guided

Schultes into the upper Putumayo. Tim and Pedro

retraced that journey to reach a Kofan settlement

on the Rio Guamues, also visited by Schultes in

1942. The Kofan recognized B. grandiflora as

tsontinba"k"d. Another of Schultes's students,

Homer Pinkley, who spent a year among the tribe

in 1965, had written that their shaman occasion-

ally ingested the plant in order to diagnose dis-

ease. This observation supported other reports

dating back to the 19th century that suggested the

plant might be hallucinogenic. Plowman wanted

to know. At Santa Rosa on the Rio Guamues he

found B. grandiflora commonly cultivated in

house yards, but he also met an old shaman who

brought him from the forest a rare but related

plant named for the tapir because of its inordinate

strength as a drug. Plowman immediately recog-

nized the plant as a new species, which he later

named Brunfelsia chiricaspi after the Quechua

word meaning cold tree. He asked the old man to

prepare it, but the shaman refused. He described

the plant as a dangerous messenger of the forest

and disavowed any knowledge of taking it for vi-

sions. Tim persisted. Eventually the shaman

agreed, although reluctantly and only on the con-

dition that Pedro also drink the preparation.

The drug, an extract from the bark, was murky

brown and bitter to swallow. Tim felt the effects

within ten minutes: a tingling sensation such as

one feels when the blood rushes back to a limb

that has fallen asleep. Only in this case the sen-

sation grew to a maddening intensity, spreading

from his lips and fingertips toward the center of

his body, progressing up his spine to the base of

his skull in waves of cold that flooded over his

consciousness. His breathing collapsed. Dizzy

vn

with vertigo, he lost all muscular control and fell

to the mud floor of the shaman's hut. In horror he

realized that he was frothing at the mouth. An

hour passed. Paralyzed and tormented by an ex-

cruciating pain in his stomach, he remained only

vaguely aware of where he was — on the earth,

face to face with three snarling dogs fighting over

the vomit that had spread in a pool around his

head.

The shaman, noticing his plight, did what sha-

mans normally do under such circumstances: He

went to bed. Desperate to escape the sensation,

half-blinded by the drug, and incapable of walk-

ing, Tim and Pedro stumbled and crawled through

the forest for two hours until finally, toward dawn,

they reached the village of San Antonio, where

they were staying in an abandoned jail. As light

came to the forest, they crawled into their ham-

mocks, where they remained, motionless, for two

days. Pedro Juajibioy, whose experience as a folk

healer had taken him on a thousand flights of the

spirit, summed up the experience succinctly: "The

world was spinning around me like a great blue

wheel. I felt that I was going to die."

This story reveals something of the character

of this remarkable young scholar, a man of ad-

venture, grace, generosity, kindness, modesty, and

honor, whose academic achievements in time

might well have surpassed even those of his fa-

mous mentor, Richard Evans Schultes, the "Fa-

ther of Ethnobotany." Sadly, Tim's life was cut

short by cruel circumstances. His untimely death

at the age of 45 from AIDS on January 7, 1989,

severed a career of immense promise. Already far

on the way as one of the most discerning, original,

and effective naturalists of the century, Tim was

a gentleman, a friend of everyone, an understand-

ing and devoted teacher, an intellectual of extraor-

dinary depth, and a tireless and demanding re-

searcher happy to share his experience and coun-

sel with whomever sought his advice.

Tim Plowman's interest in and love of plants

developed when he was a child growing up in the

temperate woodlands surrounding Harrisburg,

Pennsylvania. An avid collector even as a boy, his

passion for plants grew into the central metaphor

of his life. After attending college at Cornell Uni-

versity, he went as a graduate student to the Bo-

tanical Museum of Harvard University. Such was

his promise that even before enrolling in graduate

school, Tim was dispatched by Schultes to the

Amazon on an expedition that would define the

course of his professional life. In the fall of 1966

Tim returned from Brazil flush with excitement

and fully committed to spending the rest of his

life in pursuit of the mysteries of the tropical rain

forest.

His dissertation, with its thorough bibliographic

research, precise and deliberate taxonomic and

ethnobotanical insights, elegant prose, and exqui-

site line drawings, set the standard of excellence

by which he would operate throughout his pro-

fessional career. By the time his Ph.D. was offi-

cially conferred, in 1974, Tim was already deeply

involved in the project for which he will always

be remembered, a 15 -year effort to decipher the

complex taxonomy of Erythroxylum and to study

the ethnobotany of coca, the sacred leaf of the

Andes and the notorious source of cocaine. Of

Tim's 80 published scientific papers, 46 are relat-

ed to his work on Erythroxylum, and his position

as the world's authority on the genus enabled him

to speak eloquently and powerfully in defense of

the traditional use of coca by beleaguered indig-

enous peoples of the Andes and northwest Ama-

zon.

In 1978 Tim left Harvard for the Field Museum

of Natural History, where he became tenured in

1983 and was appointed curator in 1988. If Tim

grew up at the Botanical Museum at Harvard, he

came into his own at the Field Museum, and his

years there were both the happiest and most pro-

ductive of his remarkable career. His interdisci-

plinary interests in systematics, ethnobotany, and

ethnopharmacology led him to interact with an in-

creasingly diverse group of scholars that included

not only fellow botanists but also archaeologists,

phytochemists, ethnographers, and pharmacolo-

gists. In addition to carrying out an active scien-

tific research program as co-principal investigator

of the National Science Foundations "Projeto Flo-

ra Amazonica," he served on the editorial boards

of numerous journals, including Flora Neotropi-

cal Monographs, Advances in Economic Botany,

Journal of Psychoactive Drugs, and Journal of

Ethnopharmacology. Between 1984 and 1988 he

was co-editor-in-chief of the Journal of Ethno-

pharmacology and the scientific editor of Fieldi-

ana. He was vice president of the Beneficial Plant

Research Association, a fellow of the Linnean So-

ciety, and a member of many professional soci-

eties, including the American Society of Plant

Taxonomists, the Society of Economic Botany,

the Council of Biology Editors, the Society of

Ethnobiology, and the New England Botanical

Club. As chairman of the botany department of

the Field Museum of Natural History (1986-

1988), Tim secured a substantial increase in Na-

Vlll

tional Science Foundation funding for the herbar-

ium and developed a new facility for the curation

of economic collections. His enthusiasm, spirit of

cooperation, professional rigor, and passionate

commitment to botany proved infectious, and un-

der his leadership, morale within the botany de-

partment soared.

Credentials alone, however, present but a shad-

ow of the man who affected so many lives in such

profound ways. For Tim, life was but a vehicle

for seeking understanding and for expressing free-

dom. If there is a word to describe Timothy Plow-

man it would be freedom, and he lived with the

conviction that every person had the right to pur-

sue his or her own path unshackled by the burdens

of social convention. Equally at ease in the tran-

quil world of plants and the society of people,

Tim was charismatic, and those privileged to have

spent time with him often developed a respect that

bordered on reverence, for he was a true renais-

sance scholar, a man out of time, whose breadth

of interests and passions went far beyond the

boundaries of his beloved field of botany.

But it is as a botanist and intrepid plant explorer

that Tim will be best remembered. He spent more

than five years of his life in the most remote and

inhospitable regions of the Andes and Amazon,

making more than 15,000 collections of unsur-

passed quality. He always considered his time in

the field a privilege, and he never failed to re-

member his fellow botanists toiling away in the

less romantic confines of the herbaria. Tim

seemed to have a Rolodex in his head that re-

corded the name of every specialist in every

group of plants, and he was constantly on the

lookout for specimens that might prove useful to

a distant colleague. He collected everything. His

voucher specimens were not only complete but

aesthetically beautiful, and whenever possible he

augmented them with invaluable collections of

live material. Living plants, many new to science

and collected first by Tim, may be found in bo-

tanical gardens throughout the world.

In the rain forests of the Amazon Tim felt the

fullness of life. He marveled at the thousand

themes, the infinitude of form, shape, and texture

that so clearly mocked the terminology of tem-

perate botany. He always traveled in the forest as

a student, and his commitment to ethnobotany

grew in part from his direct experience with the

indigenous peoples who understood the plants in

ways that he believed he could only hope to em-

ulate. To be in the forest, he said, was to be in

Eden, and to say the names of the plants was to

recite the names of the gods. He believed that all

forms of life were manifestations of the sacred.

Hence, for Tim, biological and cultural diversity

represented far more than the foundation of sta-

bility; they were articles of faith, fundamental

truths that indicated the way things were supposed

to be.

Tim had a special affinity for Indians, and his

uncanny ability to gain their trust and confidence

was one measure of the deep respect he had for

their way of life. He empathized with their world

view, which defined man as but one element in-

extricably linked to the whole of creation. It was

this unique cosmological perspective, he believed,

that enabled the Indians to comprehend implicitly

the intricate ecological balance of the forest he

loved so dearly. Tim viewed with pain, dismay,

and increasing anger the other world view, in

which man stands apart, that now threatens the

forest with devastation. It was one of his fondest

hopes that the lessons of ethnobotany might ulti-

mately facilitate a dialogue between these two

world views such that folk wisdom might temper

and guide the inevitable development processes

that today ride roughshod over much of Earth.

The many of us who loved him as a brother and

respected him as a colleague can do no better ser-

vice to his memory than continue our own strug-

gles to make his dream a reality.

— Wade Davis

IX

Editors' Preface

Tim Plowman not only was a "botanist's bot-

anist," he also inspired a whole generation of

young taxonomists with his enthusiasm for plants

in both the field and herbarium.

The study of the systematics of the South

American species of Brunfelsia was the subject of

his Ph.D. research while he was at Harvard Uni-

versity. After receiving his Ph.D. in 1974, Tim

worked mainly on Erythroxylum, but Brunfelsia

remained an area of special interest for him. He

was preparing a major revision of the group, using

his dissertation as a basis and backing up with

further study and extensive field work, when he

died in 1989.

In 1994 Tim's colleagues at the Field Museum

of Natural History in Chicago sent the manu-

script, together with various notes and archival

material, to one of us (S.K.) to assess whether the

work could be brought to completion. We are very

pleased to have been able to do this, both because

it will make available this extensive study of an

interesting group of Solanaceae and because it

will serve as a tribute to the work of a well-liked

and respected colleague.

In taking the working manuscript and bringing

it to publication, we tried throughout to keep in

mind Tim Plowman's concept of the group and

changed his work as little as possible. Most of the

manuscript remains essentially unaltered, but

where necessary we rearranged some sections and

augmented others in line with Tim's own publi-

cations, notes, and records. The only major de-

parture from this approach is where we felt a need

to update the manuscript in the light of works

published since Dr. Plowman's death. Such up-

dates usually take the form of explanatory notes

in the text that are clearly indicated as being ed-

itors' contributions.

To this end, the key was revised to incorporate

additional species published (by Plowman) since

the original manuscript was written. The order of

specimen citations is Plowman's. In some places

reasoning is not clear, but we chose not to change

the order. Specimens not cited in the original work

but seen by Dr. Plowman and recorded in his

notes and in the herbarium at the Field Museum

were added; they are marked with an asterisk and

are inserted in order of collector at the end of each

state or province. These later records are included

in the distribution maps. Infraspecific taxa are not

indicated on the maps because Plowman's later

annotations were only to specific rank and could

not be integrated with distributions of infraspecif-

ic taxa. Additional references, some of them

Plowman's own subsequent publications, were in-

serted; like the additional specimens, these are

marked in the bibliography with an asterisk.

Several new combinations remained unpub-

lished at the time of Tim's death and are validated

here. The authority for these combinations should

be simply Plowman, as they represent Tim's as-

sessment of the status and rank of the taxa and

should be attributed solely to him.

The additional bibliography in Appendix II lists

references to publications that postdate the origi-

nal manuscript and that we believe will provide a

better understanding of the current state of knowl-

edge in the genus.

This revision of Brunfelsia is the product of

Tim Plowman's research, but we accept sole re-

sponsibility for any errors of interpretation or pre-

sentation.

— S. Knapp

J. R. Press

XI

Opposite: Brunfelsia hydrangeiformis subsp. hydrangeiformis (originally as Fran-

ciscea hydrangeiformis), illustration of type. Illustration reproduced from J. E.

Pohl, Plantarum Brasiliae Icones et Descriptiones (1826).

A Revision of the South American

Species of Brunfelsia (Solanaceae)

Timothy C. Plowmanf

S. Knapp and J. R. Press, Editors

I. Materials and Methods

Beginning in the fall of 1968, I began studying

the genus Brunfelsia L. in the field, both in South

America and in the West Indies. During trips

varying in duration from a few weeks to several

months, I studied many of the species of Brun-

felsia in their natural habitats, giving special at-

tention to population variability and ecological

conditions. At this time I collected extensive ma-

terial for morphological, cytological, and chemi-

cal studies, including numerous herbarium speci-

mens and living material in the form of seeds and

cuttings for propagation in the greenhouse. I also

took this opportunity to investigate the medicinal

properties and uses of the plants among native

people who still possess an intimate knowledge of

plant medicines. I learned about these aspects of

the plants not only by inquiring among native

healers but also by taking the medicines myself

(Plowman, 1977).

Many of the species I had hoped to see in the

field could not be found in localities where they

were formerly collected, due to both the sporadic

occurrence of the plants and the destruction of

their habitats by man. At other times, the plants

were found but without much sought-after flowers

or fruits.

To acquire a historical perspective of the genus,

to properly typify specific names, and to view as

much morphological variation as possible, I bor-

rowed herbarium specimens from many institu-

tions in Europe, South America, and the United

States. These specimens, numbering about 3,000

[the number at the time of his thesis in 1974, al-

though many more subsequently were examined

by Plowman — Eds.}, provided the basis for the

study. This material was organized according to

the recommendations outlined by Leenhouts's

(1968) A Guide to the Practice of Herbarium Tax-

onomy, which offered many helpful suggestions.

Some 55 morphological characters were ob-

served and measured or scored on a character

flow sheet. For many specimens it was not pos-

sible to record all the characters, especially those

of the fruits, which are rarely, if ever, collected,

and those of the corolla, which is usually poorly

pressed and often glued fast to the herbarium

sheet, precluding dissection of the stamen and pis-

til. The characters and their variation were ana-

lyzed according to methods outlined and dis-

cussed in Benson (1962), Davis and Hey wood

(1963), Lawrence (1965), and Anderson (1967),

among others. Descriptions of the various taxa

were prepared using characters observed in her-

barium specimens and in living plants. In many

cases the color of the flowers and other evanes-

cent characters were not known with certainty; as

a result, the completeness of the description varies

somewhat from species to species.

Pollen was acetolyzed according to the proce-

dure of Erdtman (1960). Unacetolyzed pollen

grains were also examined with the aid of a scan-

ning electron microscope. Cytological material

was collected in the field and in the greenhouse

and fixed in Carney's solution (absolute alcohol:

glacial acetic acid, 3:1), then stored in 95% al-

cohol. Chromosomes were stained with acetocar-

mine.

Bulk material for chemical analysis was col-

lected in the field and dried in the open sun or in

special drying ovens. This material was sent to

Mr. Jan-Erik Lindgren at the Department of Tox-

icology, Karolinska Institute!, Stockholm, Swe-

den, and to Dr. John Leary, Massachusetts Col-

lege of Pharmacy, Boston, Massachusetts. Ex-

tracts of these plants were tested for alkaloids and

general constituents using precipitation reactions

and gas and thin-layer chromatography [see Plow-

man, 1977; Lloyd et al., 1985; Schultes & Raf-

fauf, 1990, 1991— Eds.]. In addition, simple tests

for alkaloids were performed by Prof. Robert F.

Raffauf and myself on fresh plants growing in the

greenhouse. Extracts were prepared from fresh

FIELDIANA: BOTANY, N.S., NO. 39, AUGUST 31, 1998, PP. 1-135

material of five species and tested for alkaloids by

precipitation reactions.

Sixteen species were grown in the greenhouses

of the Biological Laboratories, Cambridge, Mas-

sachusetts; the Arnold Arboretum, Jamaica Plain;

and the Botanical Research Station of Northeast-

ern University in Woburn, Massachusetts. [Other

specimens were cultivated at the Field Museum

of Natural History, Chicago, Illinois, in later

years. — Eds.] Material was propagated from both

seeds and cuttings collected in the wild or from

cultivated plants. Crosses were attempted between

as many species as possible. The living plants

were also a valuable source of those morpholog-

ical characters that are lost in dried specimens and

permitted observations to be made on the growth

and development of various organs, particularly

the inflorescence.

A considerable amount of time in this study

was devoted to searching the botanical, chemical,

and pharmacological literature for references to

Brunfelsia. Only those works that are cited in the

text are included in the bibliography.

All of the specimens cited in this work were

examined by me unless they are denoted by "non

vidi." The use of brackets in the citation of a col-

lection number indicates that some uncertainty ex-

ists or, in the case of type materials, that the num-

ber appearing on the sheet was not given in the

original citation of the specimen.

Finally, the use of taxonomic characters de-

serves some explanation. Of the 24 species of

Brunfelsia that are recognized in South America,

four are considered to possess one variety each in

addition to the typical variety. The category of

subspecies is here used for geographical or eco-

logical races that show morphological differenti-

ation from the typical form of the species. It

should be noted, however, that the degree of dif-

ferentiation and isolation varies from one species

to another, so that the subspecific concepts are not

entirely equivalent. Furthermore, variants exhib-

iting considerable differentiation exist within sev-

eral subspecies (e.g., B. grandiflora subsp. schul-

tesii Plowman, B. brasiliensis (Spreng.) L.B. Sm.

& Downs subsp. brasiliensis). Subsequent work-

ers in the genus may choose formally to recognize

these as variants at infraspecific rank.

II. Taxonomic History of the Genus

The genus Brunfelsia was first described by

Charles Plumier in 1703 in his Catalogus Plan-

tarum Americanarum, commemorating the Ger-

man herbalist Otto Brunfels (1464-1534). Brun-

fels was a Carthusian monk and physician who,

in 1530, published an herbal, Herbarum Vivae Ei-

cones, that, as the name indicates, included illus-

trations of living plants. This early work is out-

standing for the high quality, accuracy, and nat-

uralistic aspect of the woodcuts, which for the first

time departed from the often copied and stylized

illustrations of earlier herbals (Arber, 1912). Un-

fortunately, Plumier misspelled the genus name as

"Brunsfelsia" an error that has tenaciously per-

sisted in the literature [see D'Arcy, 1989 — Eds.].

Simultaneously with his short description and

diagnosis, Plumier published a rather good illus-

tration of the flower and fruit of the plant, which

is now known as B. americana L. The illustration

was taken from a larger, more complete drawing

that, along with more detailed descriptions in

manuscript form, is kept at the Museum National

d'Histoire Naturelle in Paris, France. The com-

plete plate was published with a more complete

description in 1756 in Burman's edition of Plu-

mier's Catalogus Plantarum Americanarum (Fig.

1).

In the second edition of Genera Plantarum

(1742), Linnaeus included Brunfelsia, based on

Plumier's description. He classified Brunfelsia

among the Pentandria Monogynia, indicating that

the plant had five stamens (an aberrant condition

that occasionally occurs in B. americana). It

would thus seem that Linnaeus had access to

specimens since Plumier failed to mention this

character and it is not indicated in his drawing.

Although there is no known specimen of B. amer-

icana in extant Linnaean herbaria, he did indicate

with a symbol in the fifth edition of Genera Plan-

tarum (1754) that he knew the genus from her-

barium material.

In a letter to David Van Royen in 1767, Lin-

naeus mentioned that he did, in fact, see a speci-

men of the Brunfelsia of Plumier: " Brunsfelsiam

Plumieri vidi exsiccatam, at ilia forte a Tua specie

diversa erat, ideoque nihil certi de ea potui pron-

FIG. 1. Brunfelsia americana. From Burmann's edition of Plumier's Catalogus Plantae Americanarum (1703).

FIELDIANA: BOTANY

TAB . LXV.

PLOWMAN: REVISION OF BRUNFELSIA

unciare, nee Adansonii compilationes mihi suffec-

ere" (Bonnet, 1895). Van Royen had sent Linnae-

us a specimen (no. 257.1, LINN) that Linnaeus

recognized as being different from Plumier's spe-

cies. This plant is now referable to B. undulata

Swartz.

The valid publication of the genus dates from

1753, when Linnaeus published B. americana in

Species Plantarum, in which he followed the orig-

inal misspelling of the genus. This was apparently

an orthographical oversight on his part with which

he was inconsistent in later works. For example,

in the following year (1754), he used the correct

spelling in Genera Plantarum but misspelled it in

subsequent editions of Species Plantarum. In the

International Code of Botanical Nomenclature

(Greuter et al., 1994, p. 297), Brunfelsia is listed

in Appendix IIIA as a conserved spelling [see also

D'Arcy, 1989— Eds.} .

The next important date in the history of the

genus was 1826, when Pohl published seven new

species from Brazil as the genus Franciscea,

which he placed in the order Scrophularinae, close

to BrowalHa. Each of Pohl's seven new species

was accompanied by an excellent illustration

(frontispiece, Figs. 22, 28, 31, 39, this work). Ad-

ditional species were soon added to Franciscea

by Chamisso and Schlechtendal (1827) and Hook-

er (1828).

However, as early as 1829, David Don pub-

lished his observations on this group, in which he

maintained that Franciscea did not differ signifi-

cantly from Brunfelsia. He united the two genera

under the earlier name Brunfelsia but published

only one new combination, B. uniflora (Pohl) D.

Don. Don also expressed the opinion, precocious

for the time, that Brunfelsia belonged in the So-

laneae (Solanaceae) rather than the Scrophulari-

nae, an idea suggested even earlier by Antoine de

Jussieu (1791).

Most authors of the time ignored Don's conclu-

sions and continued to recognize Brunfelsia and

Franciscea as distinct genera, primarily on the ba-

sis of the nature of the capsule and the length of

the corolla tube (Bentham, 1835; G. Don, 1837;

Endlicher, 1839; Meisner, 1840).

In 1846, however, in his monograph on the

Scrophulariaceae for de Candolle's Prodromus,

Bentham united Franciscea with Brunfelsia, ob-

serving that although the francisceas were distin-

guished by some authors, especially by the dry

capsule, he himself had looked in vain for defin-

itive structures; B. latifolia (Pohl) Benth. and B.

guianensis Benth. had small flowers like the rest

of the francisceas but capsules no less fleshy than

those of B. americana, and in several species the

fruit was completely unknown.

Bentham's treatment of Brunfelsia included 19

species, eight of which were formerly considered

Franciscea. He divided the genus into three

groups without rank: Multiflorae and Uniflorae

(Franciscea) and Longiflorae (Brunfelsia), placed

among the Salpiglossideae. This work bears the

distinction of being the only revision of Brunfelsia

that has attempted to encompass the entire genus.

Bentham considered the Salpiglossideae to be a

"suborder" (subfamily) of the Scrophulariaceae

and noted their close affinity with the capsular

Solanaceae, with which, he added, they might be

better associated.

Shortly after Bentham's revision appeared,

John Miers began publishing a series of papers on

the Solanaceae consisting of taxonomic treat-

ments, descriptions of new genera and species,

and a major reorganization of the tribes into two

families, the Solanaceae and the Atropaceae.

These replaced the older concept of Solanaceae

and included the Salpiglossideae of Bentham. The

text of Miers 's work was originally published in

The Annals and Magazine of Natural History be-

tween 1849 and 1857 and later appeared bound

together in two volumes accompanied by 87

plates under the name Illustrations of South Amer-

ican Plants (Miers, 1855, 1857). This extensive

work remains a major source of information for

solanaceous plants.

Miers (1849) erected the new family Atropa-

ceae as a distinct group with imbricate or plicate

corolla estivation that was intermediate between

the Solanaceae (with valvate estivation) and the

Scrophulariaceae (with imbricate estivation). The

group included ten tribes, the last of which he

named the Brunfelsiae. Miers firmly believed that

Brunfelsia and Franciscea should be maintained

as distinct genera, citing some ten differences that

he considered diagnostic. Among these were the

habit of the plants, the relative length and shape

of the corolla, the color of the corolla, the shape

of the anthers, the size and structure of the fruit,

and the number of seeds per fruit.

Several points of difference indicated by Miers

between Brunfelsia and Franciscea no longer

stand when one considers the wide range of vari-

ation and many additional species that are now

recognized. For example, he stated (1849, 1850)

that brunfelsias are trees attaining 20 ft in height

whereas francisceas are small shrubs. In fact, both

low shrubs and small trees occur in about equal

FIELDIANA: BOTANY

numbers in both groups. Miers considered the

fruit of Brunfelsia a fleshy drupe containing an

indehiscent putamen and that of Franciscea a de-

hiscent capsule that splits along four vertical su-

tures. It is now clear that within both Brunfelsia

and Franciscea, dehiscent and indehiscent, thin-

and thick-walled, and fleshy and dry fruits occur.

On the basis of limited material, Miers failed to

recognize the overall similarities between these

groups. However, several differences that he

pointed out remain valid and serve to distinguish

Brunfelsia and Franciscea as distinct sections of

the genus.

Perhaps the most disturbing fact in Miers 's

study of Brunfelsia is his omission of B. guianen-

sis, a species that Bentham (1846) described and

commented on in the Prodromus. In Miers's oth-

erwise complete list of new combinations, in

which he transferred most of Bentham's new

brunfelsias to Franciscea, B. guianensis is con-

spicuously missing. Since B. guianensis possesses

several features that are intermediate between

Brunfelsia and Franciscea, it seems that Miers

perhaps simply ignored it because it blurred the

sharp distinction he wanted to demonstrate.

The view of Bentham, to unite Franciscea with

Brunfelsia, eventually prevailed over Miers's

ideas, perhaps because of the wide distribution

and general acceptance of de Candolle's Prodro-

mus. Although some of Miers's conclusions have

not gained acceptance in light of more recent

knowledge, his extensive observations are none-

theless valuable to any taxonomist interested in

the Solanaceae. For example, Bentham and Hook-

er (1873) drew heavily from Miers's data in de-

scribing the Solanaceae for Genera Plantarum.

There has been only one other major study of

Brunfelsia since the time of Miers and Bentham.

This is found in J.A. Schmidt's treatment of the

Scrophulariaceae in Martius's Flora Brasiliensis in

1864. Although only Brazilian species were in-

cluded, then numbering about ten, Schmidt made

many sound taxonomic judgments and correctly re-

duced many names to synonyms. Until the present

time, this work remained the standard reference for

Brunfelsia in Brazil, which possesses the largest

number of species of any country in South Amer-

ica. However, Schmidt's treatment is now out of

date, as may be expected, and not without error.

In an attempt to clarify certain nomenclatural

problems regarding his new species Brunfelsia

mire Monach., Monachino (1953) offered a brief

discussion of the names of the South American

species. This was a work of limited value and

provided little help in solving the taxonomic dif-

ficulties in the genus, being based primarily on

the erroneous conclusions of others.

Besides the original descriptions of the plants,

which for certain species are the only information

available, only floristic works have provided ac-

counts of the species in particular areas. Foremost

among these, which are by no means of uniform

quality, are the following works: Flora of the Brit-

ish West Indian Islands (Grisebach, 1861); Botany

of Porto Rico and the Virgin Islands (Britton &

Wilson, 1925); Flora de Cuba (Liogier, 1957);

Flora of Peru (Macbride, 1962); Flora Ilustrada

Catarinense, Solandceas (Smith & Downs, 1966);

Flowering Plants of Jamaica (Adams, 1972); and

Flora of Panama (D'Arcy, 1974).

Table 1 summarizes the history of classification

of Brunfelsia, including generic, tribal, and family

placement. It should be noted under the heading

"Family" that the groupings of the earlier bota-

nists were not equivalent to our modern families

of plants. These are included for historical pur-

poses only.

III. Generic Relationships

Brunfelsia is a distinct genus that is not readily

confused with other genera. It belongs to the tribe

Salpiglossideae of the Solanaceae, as circum-

scribed by Bentham (1835, 1846, under Scrophu-

lariaceae), Bentham and Hooker (1873), Wettstein

(1895), and Baehni (1946). This group is consid-

ered the most advanced tribe of the Solanaceae1

and resembles Scrophulariaceae in several charac-

ters, such as the zygomorphic corolla with some-

what imbricate estivation, the number of stamens

reduced to four or two, and the capsular fruits. It

is distinguished from the latter by intraxylary phlo-

em (Metcalfe & Chalk, 1950) and from other So-

lanaceae by the estivation of the corolla and the

number of stamens.

The closest relatives of Brunfelsia are Browal-

lia L., a weedy herbaceous genus of two or pos-

1 Although traditionally placed in the tribe Salpiglos-

sideae and considered closely related to Browallia and

Salpiglossis, results of recent phylogenetic studies using

chloroplast DNA (Olmstead & Palmer, 1992; Olmstead

& Sweere, 1994; Olmstead et al., in press) indicate that

Brunfelsia may be better placed in a monophyletic group

with the genera Petunia, Fabiana, and Nierembergia.

The morphological evidence remains to be assessed. —

Eds.

PLOWMAN: REVISION OF BRUNFELSIA

TABLE 1 . Summary of historical classification of Brunfelsia.

sibly three heteromorphic species, and Streptoso-

len Miers, a monotypic Andean shrub genus.

[Plowmania Hunz., a monotypic genus from mon-

tane Guatemala, first described as a species of

Brunfelsia, is also considered to be closely related

to Brunfelsia (see Hunziker & Subils, 1986). —

Eds.] Browallia is distinguished from Brunfelsia

by its herbaceous habit; by the filaments, which

are greatly dilated and ciliate at the apex; by the

broad, lamellate stigma, which has recurved mar-

gins, and by the dry capsule, which is dehiscent

by four sutures. Streptosolen differs from Brun-

felsia in having the corolla tube twisted at the

base, the limb bright red-orange, and the filaments

thickened at the apex. This shrub grows at high

altitudes in the Andes from Colombia to Peru and

is pollinated by hummingbirds.

The other genera of the Salpiglossideae are pri-

marily annuals found in southern South America:

Salpiglossis Ruiz and Pavon, Schizanthus Ruiz

and Pavon, Leptoglossis Benth., Reyesia Clos, Pe-

tunia Juss., and Schwenkia L. (which also occurs

in Africa). In addition, three genera of Australian

shrubs are also placed in the tribe: Duboisia R.

Br, with indehiscent, baccate fruits, and Antho-

cercis Labill. and Anthotroche Endl., with dehis-

cent, capsular fruits.2

2 Since the time of the dissertation, these genera have

been segregated as the endemic tribe Anthocercideae

(Haegi, 1981). and additional genera have been de-

scribed.— Eds.

Hutchinson (1969) elevated the Salpiglossideae

to the rank of family (Salpiglossidaceae) with the

explanation that the genera of this group consis-

tently arrived next to Scrophulariaceae in his key

to the families of plants of the world. This deci-

sion does not seem justified in view of the exten-

sive evidence from morphology, anatomy, and

chemistry, which links the Salpiglossideae with

the Solanaceae.

The origin and relationships of the Salpiglos-

sideae remain obscure. Baehni (1946) derived the

group from two sources, the Solaneae via the sub-

tribe Goetzeineae and the Atropeae via Para-

bouchetia Baillon. In the past, relationships at the

generic and tribal levels in the Solanaceae were

studied using relatively few characters, such as

the curvature of the embryo, corolla estivation,

number of anthers, or type of fruit. A definitive

and in-depth evaluation of these higher relation-

ships is needed in which many characters are con-

sidered simultaneously, including those from mor-

phology, chemistry, anatomy, cytology, and other

areas [see footnote 1 — Eds.}

IV. General Morphology

Habit

All known species of Brunfelsia are woody

shrubs, a few of which will become small trees.

FIELDIANA: BOTANY

No herbaceous members are known. In section

Brunfelsia, certain species may reach 10-12 m in

height (B. jamaicensis (Benth.) Griseb., B. amer-

icana, B. lactea Krug & Urb.). These are some-

what atypical, however, and most are only 2-3 m.

Species of sections Franciscea and Guianenses

are mostly medium-sized shrubs 1-3 m tall that

form a component of the understory of the forest.

Brunfelsia grandiflora D. Don of the western Am-

azon may become a small tree 4-5 m tall or, in

the Brazilian state of Acre, may become lianoid

in habit, attaining up to 10 m. Brunfelsia latifolia

is the most diminutive species, often flowering

when only 30 cm tall and rarely reaching 1 m in

height.

Root

The roots of Brunfelsia are diffuse in the seed-

lings and young plants. However, mature plants

often develop a deep, woody taproot. This is es-

pecially true in species of section Brunfelsia,

which often grow in rocky areas. In section Fran-

ciscea, the central root may be more or less

branched (Fig. 2).

Branching

Species of Brunfelsia may have one central

trunk (B. densifolia Krug & Urb., B. mire, B. chir-

icaspi Plowman) or they may be branched to a

greater or lesser extent from near the base, at

times producing two or more main trunks (B. aus-

tralis Benth., B. grandiflora, B. brasiliensis). In

some cultivated species, root suckers often appear

around the main trunk and form extensive clonal

plants (B. nitida Benth., B. undulata Sw., B. aus-

tralis).

In section Brunfelsia, the branches are often

few and subvirgate, with few leaves at the end of

a long, spindly branch (B. undulata, B. splendida

Urb.). In sections Franciscea and Guianenses, the

branches are more often diffuse or spreading and

tend to arch over with age.

The branching pattern in Brunfelsia is sympo-

dial. When the terminal bud ceases to grow, the

shoot is perpetuated from one or more subtermi-

nal axillary buds or, rarely from the bracts of the

inflorescence. In the new shoot, several prophylls

are produced, followed by a gradual transition to

mature leaves.

Leaf

The leaves of Brunfelsia are alternately ar-

ranged with 5/13 phyllotaxy. They may be scat-

tered along the branchlets or crowded toward the

apex, rarely subverticillate through periodic short-

ening of the internodes (B. mire, B. hydrangeifor-

mis (Pohl) Benth., B. rupestris Plowman). The

petiole is very short in relation to the leaf blade,

varying from 1 to 10 mm in length. It is semiter-

ete in section and canaliculate, and frequently be-

comes transversely cracked or corrugate with age

or drying.

The leaves are simple with an entire margin.

There is much variation in their size and shape,

which provide important characters for distin-

guishing the species. In form, the blade may be

elliptic to lanceolate; obovate to oblanceolate; or,

rarely, spathulate, ovate, linear, or linear-lanceo-

late. Certain species are noted for great variability

of leaf shape, sometimes in the same individual

(B. uniflora (Pohl) D. Don, B. americana, B. lac-

ted). In other species (B. portoricensis Krug &

Urb., B. densifolia), the leaves of seedlings and

young plants differ radically from the foliage of

adult plants. At the apex, the leaves may be blunt,

acute, or acuminate. It is of ecological interest that

the leaves of species of wet, tropical forests are

often acuminate in the form of a modified "drip

tip" (B. grandiflora, B. macrocarpa Plowman, B.

chocoensis Plowman, B. amazonica Morton).

In size, leaves may be 1-30 cm long and 0.1-

15 cm wide. Leaf size may vary considerably

within a species (e.g., B. grandiflora, 6-23 cm

long) and is strongly influenced by local environ-

mental conditions such as available light and

moisture.

The leaf texture is mostly firm-membranaceous

to subcoriaceous, often soft and rubbery to the

touch. A few species (B. splendida Urb., B.

dwyeri D'Arcy, B. obovata var. coriacea J. A.

Schmidt, B. rupestris) are noted for their shiny,

thick-coriaceous leaves. Coriaceous leaves also

occur in certain populations of B. uniflora and B.

brasiliensis that grow in unusually dry or exposed

situations.

In color, the leaves are light to dark green and

paler green beneath. Dark purple pigmentation

due to anthocyanins is present in the young flush

growth of several species (B. pilosa Plowman, B.

plicata Urb., B. undulata). The leaves of B. splen-

dida are particularly striking, being very dark

green, almost black, and very shiny.

The nervation in Brunfelsia consists of the mid-

PLOWMAN: REVISION OF BRUNFELSIA

FIELDIANA: BOTANY

rib, which is completely submerged on the adaxial

surface and prominent beneath, and three to 12

pairs of pinnate, lateral nerves. Each nerve

spreads out from the costa and arcuately anasto-

moses with the next near the margin. The veinlets

are finely reticulate and may be obscure or some-

what prominulous. The number of lateral nerves

may be used to separate certain species (e.g., B.

martiana Plowman from B. guianensis), but the

angle of divergence and degree of arching are of-

ten variable. Brunfelsia hydrangeiformis is unique

in having the midrib, nerves, and major veinlets

sunken into a shallow depression or sulcus on the

adaxial surface, creating a very uneven surface

texture.

Indument

Most Brunfelsia species are essentially glabrous

throughout, although the young leaves and flower

buds may be sparsely puberulent in their earlier

stages, becoming glabrous with age. In the pu-

bescent species, trichomes are usually encoun-

tered in the twigs, leaves, pedicel, calyx, and co-

rolla tube. In section Franciscea, the indument is

best developed in B. boliviano Plowman, B. bras-

iliensis, B. cuneifolia J.A. Schmidt, B. hydrangei-

formis, and B. pilosa, and to a much lesser extent

in certain populations of B. uniflora, B. pauciflora

(Cham. & Schltdl.) Benth., and B. obovata Benth.

In section Brunfelsia, the following species may

be pubescent: B. americana, B. sinuata A. Rich.,

B. membranacea Urb., and B. jamaicensis. Spe-

cies of section Guianenses are entirely glabrous

but occasionally have sparse glandular hairs on

the petioles and pedicels (B. burchellii Plowman,

B. clandestina Plowman) and ciliate or villous

bract margins (B. chocoensis, B. martiana).

Uniseriate trichomes are found in both section

Brunfelsia and section Franciscea. They may be

either glandular or nonglandular, containing from

two to ten cells. Glandular trichomes often have

a single terminal, globose cell or may have one

to several biseriate, terminal cells (B. uniflora).

Several types of trichomes may occur together in

the same organ, or only one type may be found.

In Venezuelan populations of B. uniflora, the ca-

lyx and pedicel may bear short glandular tri-

chomes, while the twigs and leaves may be pilose

to villous with nonglandular hairs and a sharp de-

marcation at the attachment of the pedicel.

In section Brunfelsia (B. americana, B. sinu-

ata), simple, branched trichomes also occur on the

leaves. These have not been observed in section

Franciscea.

The presence, location, and degree of pubes-

cence appear to be highly variable in some species

(B. uniflora, B. pauciflora, B. americana). At-

tempts have been made in the past to distinguish

varieties solely (or essentially so) on the basis of

the presence or absence of trichomes (B. uniflora

var. pubescens (Benth.) R. E. D. Baker & N. W.

Simmonds, B. americana var. pubescens Griseb.,

B. pauciflora var. calycina (Benth.) J. A. Schmidt,

B. brasiliensis var. acuminata (Pohl) L. B. Sm. &

Downs). Numerous intergradations in degree of

pubescence are found in these species, and lack-

ing correlation with other, more stable characters,

it would be unwise to recognize these as "varie-

ties." Indeed, variation can be observed within

duplicates of a single collection. [This is also

commonly the case in other Solanaceae (see

Knapp, 1989). — Eds.] The greater development of

pubescence is sometimes associated with plants

growing in drier or more exposed habitats, sug-

gesting that they may be environmentally induced

forms or ecotypes. In any case, the unstable nature

of these variants precludes their receiving formal

taxonomic status. In other species, however, the

type and location of trichomes, correlated with ad-

ditional characters, can be useful in delimiting

subspecies (B. brasiliensis, B. hydrangeiformis).

In relatively few species is the indument of di-

agnostic value at the specific level. Brunfelsia pi-

losa is characterized by long, pilose trichomes on

the twigs, leaves, and calyces. Brunfelsia bras-

iliensis has a distinctive yellow-brown pubescence

that is generally distributed in the plant. Brunfel-

sia cuneifolia usually has a villous calyx, a lower

leaf surface, and twigs, although the type collec-

tion is nearly glabrous. Brunfelsia boliviana has

villous-ciliate leaf margins and villous twigs.

Inflorescence

The inflorescence may be strictly terminal, ax-

illary, or a combination of terminal and axillary

in the same plant.

FIG. 2. Root of Brunfelsia grandiflora subsp. schultesii.

PLOWMAN. REVISION OF BRUNFELSIA

The inflorescence in Brunfelsla is basically a

cincinnus or some reduction or variation thereof

(Danert, 1958). In Diccionario de Botdnica, Font

Quer (1965) provided a succinct definition of the

cincinnus as a uniparous cymose inflorescence in

which the medial plane of each branchlet is placed

transversely with respect to the medial plane of

the subtending bract of the main axis (pseudaxis)

and with the lateral branchlets alternately dis-

posed to the right and left. It is a scorpioid cyme

in which the various pseudaxes (branchlets) do

not all fall in the same plane.

The cincinnal cyme is a primitive type of inflo-

rescence in the Solanaceae in that the lateral

branchlets are essentially undifferentiated from the

primary inflorescence axis. In view of the other-

wise advanced position of Brunfelsia in the family,

this is a striking feature [but see Olmstead & Palm-

er, 1992— Eds.].

In Brunfelsia species with well-developed in-

florescences, the cincinnus is elaborated as fol-

lows: a single flower terminates the main or lat-

eral axis and is subtended by one to three (rarely

more) small, caducous bracts, each of which

bears a bud in its axil. From one or more of these

bracts arise lateral branchlets, which are usually

highly reduced, each of which bears a terminal

flower and usually two tiny bracts, one on each

side. Each successive branchlet is formed at right

angles to its subtending bract. In secondary and

higher orders of branching, only one of the two

bracts forms a branchlet, alternating to the right

and left. This gives the older branches of the in-

florescence a zigzag pattern, produced by the re-

maining pedicel scars (e.g., B. latifolia, B. gran-

diflord).

The most common form of inflorescence in the

genus is few-flowered, compact, and sessile, with

the branchlets of the cincinnus very reduced. In

some species, the individual branchlets are often

difficult to distinguish, so condensed are the

flower clusters. Two species (B. grandiflora, B.

brasiliensis) are known to develop open, lax in-

florescences in which the branches may be elon-

gated to greater or lesser extent. In other species

(B. mire, B. hydrangeiformis), many flowers are

produced but without elongation of the inflores-

cence branchlets, resulting in a very dense, cap-

ituliform cyme. In still others, the inflorescence

consists of a single flower (B. americana, B.

dwyeri, B. macrocarpa, B. uniflora) or one to

three flowers (B. pilosa, B. australis, B. cunei-

folia, B. rupestris).

Great plasticity is found in the expression of

the inflorescence in a single species and is due

to both environmental and genetic factors. Spe-

cies that frequently produce many flowers (B.

grandiflora, B. hydrangeiformis) in a well-de-

veloped inflorescence may have only a few flow-

ers per cluster. Other species that usually are sin-

gle-flowered (B. pilosa, B. amazonicd) will oc-

casionally develop two or three flowers. Under

normal forest conditions, B. chiricaspi bears up

to seven flowers in the inflorescence; however,

one collection (Kennedy 1386), growing in full

sun in a cut-over place, had more than 20 flow-

ers. Similar light-related responses have been ob-

served in B. grandiflora (Plowman 2535). This

variability, evident to some extent in most of the

species, limits the usefulness of the number of

flowers per inflorescence as a taxonomic char-

acter, particularly in the construction of keys.

Flower

The flowers of Brunfelsia species are com-

posed of the following parts: a pedicel; a five-

lobed calyx; a long-tubular corolla with a five-

lobed limb; two pairs of stamens inserted in the

upper part of the corolla tube; and a pistil, the

style of which nearly equals the tube. There are

major differences among the flowers of the three

sections of the genus; however, within a section,

the flowers are remarkably similar with only a

few minor exceptions. Differences that do occur

and that are important in distinguishing the spe-

cies are found in the relative sizes of the flowers

and their parts.

The flowers are slightly zygomorphic with bi-

lateral symmetry. This is observable in the calyx,

which occasionally has unequal lobes; in the cur-

vature of the corolla tube; in the unequal limb;

and in the didynamous stamens. The stamens and

style are curved toward the "back" or posterior

side of the flower, in line with the uppermost co-

rolla lobe, and the anthers and stigma are brought

very close to that side of the tube. This construc-

tion is probably advantageous in pollination be-

cause the proboscis of the pollinating insect (Lep-

idoptera) is most likely to descend along this part

of the tube (see Fig. 3).

The flowers of section Brunfelsia are very large

with a long tube and a usually showy white limb

that fades to yellow (Figs. 4, 5). They are ex-

tremely fragrant in the evening and apparently are

pollinated by night-flying moths (Sphingidae). In

section Franciscea the flowers are smaller and

10

FIELDIANA: BOTANY

GUIANENSES BRUNFELSIA FRANCISCEA

FIG. 3. Floral structure of the sections of Brunfelsia.

have a shorter tube with a purple limb and a white

eye at the center. They are usually not fragrant

and are pollinated by day-flying butterflies from

several different families. Species of section Gui-

anenses have small whitish or greenish flowers

with a short tube and a relatively narrow limb.

They are said to be fragrant, but the pollinators

are unknown.

Calyx

The calyx is actinomorphic or slightly zygomor-

phic, in form tubular to campanulate and often

somewhat inflated. It varies in length from 3 to 35

mm. In section Brunfelsia the calyx is very short

compared to the corolla tube; in sections Francis-

cea and Guianenses it is usually about half as long,

PLOWMAN: REVISION OF BRUNFELSIA

11

12

FIELDIANA: BOTANY

although in some species (B. pauciflora, B. hy-

drangeiformis) it may equal the corolla tube.

The calyx teeth or lobes are triangular, ovate or

lanceolate, and valvate in bud. In section Brun-

felsia they are often blunt with a few ciliate tri-

chomes. In sections Franciscea and Guianenses

the teeth are acute or acuminate, and the ciliate

margins are much reduced or obsolete.

The calyx is glabrous or may be furnished with

glandular (B. boliviana, B. pauciflora, B. uniflord)

or nonglandular (B. pilosa, B. brasiliensis) tri-

chomes. In color it is usually some light shade of

green to yellowish (B. undulata, B. grandiflord)

or purplish (B. pilosa, B. jamaicensis).

The size and shape of the calyx, as well as its

pubescence, are taxonomically important in spe-

cific delimitation. In most species the calyx is te-

rete, but in a few (B. obovata, B. boliviana, B.

cuneifolia, B. nitidd) it is five-angled at the sinus-

es and concave at the lobes.

The nature of the calyx in fruit also provides

useful characters. The calyx is usually persistent

in fruit but scarcely accrescent. It may become

thicker and firmer in texture and in section Fran-

ciscea frequently bears lenticular outgrowths near

the base. It may split along one or more sides to

accommodate the developing fruit, persisting only

at the base, or it may increase in girth to partially

or completely enclose the fruit at maturity (B. hy-

drangeiformis, B. pauciflora, B. brasiliensis).

Brunfelsia macrocarpa is unique in the genus in

having a large, truly accrescent calyx that be-

comes thick and leathery, completely enclosing

the fruit.

Corolla

The corolla differs significantly in the three sec-

tions of Brunfelsia, as mentioned above (Fig. 3).

In section Brunfelsia the corolla tube is very long,

from 6 to 24 times as long as the calyx. It be-

comes gradually dilated near the apex but never

curved or constricted at the mouth (Fig. 5). The

limb usually is held at an angle to the tube. The

five lobes may be narrow or broad, plane or un-

dulate to crenate at the margin. In color the co-

rolla tube is greenish or white; the limb is usually

pure white when it first opens, turning yellow with

age. Brunfelsia americana and B. lactea may be

streaked with purple. Anomalous species are B.

cestroides A. Rich, and B. purpurea Griseb. of

Cuba, the flowers of which are said to be purplish

red.

Section Franciscea has a much shorter tube

that is only one to five times as long as the calyx.

It is briefly inflated near the apex to include the

anthers and stigma, then slightly constricted to

form a raised, fleshy ring. The tube may be white

or violet, and the limb is deep to pale violet, fad-

ing to pale lavender or pure white with age. The

ring or eye at the center may be white or yellow.

The limb in section Franciscea may be large and

varies from 1 .5 to 8 cm in diameter. The lobes are

broadly rounded, overlapping at the lateral mar-

gins.

The corolla tube in section Guianenses is short,

only two to three times as long as the calyx, and

gradually dilated or inflated toward the apex.

There is no marked constriction as in section

Franciscea. The flowers in section Guianenses are

white or greenish white and apparently do not

change color with age. The corolla limb in section

Guianenses is relatively small, rarely exceeding

25 mm in diameter. The lobes are quite narrow

and convex; the lateral margins appear to be re-

flexed and are completely folded under in pressed

specimens. This is also observed in B. densifolia

of section Brunfelsia (Fig. 4).

The estivation of the corolla lobes in bud is a

combination of quincuncial and imbricate. In

quincuncial estivation, two of the lobes lie wholly

to the outside, two wholly on the inside, and one

lobe has one side to the inside and one to the

outside. In section Brunfelsia the uppermost lobe

is wholly to the inside, one of the lower lobes is

wholly to the outside, and the other three are im-

bricate with one side in and one side out.

Only the size of the corolla and the shape of

the lobes are useful as taxonomic characters.

Within each section the corolla is quite similar.

Characters such as color and the shape of the cen-

tral eye in section Franciscea are diagnostic in the

living state, but these are unfortunately lost in

preparation of specimens and rendered useless to

the herbarium taxonomist.

Androecium

The four stamens occur in two pairs in the up-

per part of the corolla tube, alternating with the

FIG. 4. Brunfelsia densifolia (Knapp s.n.) from Puerto Rico.

PLOWMAN: REVISION OF BRUNFELSIA

13

14

FIELDIANA: BOTANY

upper three corolla lobes. A fifth stamen is oc-

casionally produced in Brunfelsia americana, but

this is an irregular condition. The glabrous fila-

ments are fused to the corolla tube for most of

their length and are free for only 2-5 mm. The

anterior pair is longer, exceeding the stigma, and

may be briefly exserted from the tube (B. obovata,

B. chocoensis); the posterior pair, attached to the

"back" of the tube, is shorter than the style. Both

pairs are curved at the apex to project the anthers

toward the back of the tube. In section Brunfelsia

the filaments may be curved somewhat in the re-

gion of the stigma (B. americana, B. lactea),

The anthers are medifixed to the filaments and

dehisce by a single longitudinal slit. In section

Brunfelsia the anthers are oblong-reniform and

may appear to be slightly bilobate, although the

cells are confluent. At anthesis, each pair may

tend to adhere to each other, apparently from the

sticky nature of the pollen. Brunfelsia densifolia

is somewhat anomalous in having both pairs of

anthers reaching the mouth of the tube; the ante-

rior pair is much reduced in size, although it pro-

duces a small amount of pollen. They are not

completely sterile, as implied in Urban's original

description (Krug & Urban, 1897). Slight inequal-

ity in anther size is not unknown in other species

(cf., B. martiand) and hardly merits the formation

of a new genus as proposed by Urban (1906). In

sections Franciscea and Guianenses the anthers

are unilocular and more or less semicircular, ap-

pearing round or reniform in outline.

Pollen

The pollen grains of all three sections of Brun-

felsia are spheroidal.3 Those of section Brunfelsia

(B. americana, B. undulata) and section Guianen-

ses (B. chocoensis, B. martiand) appear triporate

or possibly tricolporate. Those of section Fran-

ciscea are tricolpate or tricolporoidate (B. chiri-

caspi, B. grandiflora, B. latifolia, B. pilosd). They

are all medium-sized grains 30-38 to 34-42 jim

in diameter.

The surface of the pollen was examined by

means of a scanning electron microscope using

3 Pollen grains of Brunfelsia species were acetolyzed

according to the procedure of Erdtman (1960).

unacetolyzed grains. These revealed low-relief

scrobiculate or reticulate exine patterns. Although

the sculpturing appears to differ slightly from spe-

cies to species, the preparations were not suffi-

ciently clear to permit a detailed analysis, nor

were enough species observed to permit general-

ization.

Gynoecium

The ovary is ovoid or conical and is inserted

on a short, nectar-producing disc that may be ob-

scure in dried specimens or appear merely as a

slight thickening at the base of the ovary. The

ovary is bilocular with axile placentation, bearing

numerous anatropous ovules. The thin dissepi-

ments occasionally do not extend to the apex of

the ovary, breaking down early in fruit formation

and leaving the placenta attached only at the base.

The fruit then may have only one locule.

The style is filamentous and nearly equals the

corolla tube in length. It may be somewhat thick-

ened and curved at the apex, especially in sections

Franciscea and Guianenses, or suberect, as in

several species of section Brunfelsia.

The stigma is usually included between the four

anthers in the uppermost part of the tube. Only

rarely does it project to the mouth of the tube (B.

densifolia). The stigma is short and briefly bifid

in sections Franciscea and Guianenses and faces

the posterior wall of the corolla tube. It has the

appearance of an open forceps at anthesis with the

stigmatic tissue included between the lobes. After

pollination, the lobes close together. In section

Brunfelsia the stigma is more shallowly lobed and

may appear subcapitate and convex.

Fruit

There has been considerable misinterpretation

of the fruit of Brunfelsia from both a morpholog-

ical and a phylogenetic point of view. This has

occurred, I believe, because early botanists had

very little material to work with and drew their

conclusions from only a few specimens. The fruit

varies considerably from group to group, and dif-

ferent fruit types may be found in otherwise

closely related species. In some cases, the nature

FIG. 5. Brunfelsia lactea (Knapp s.n.) from Puerto Rico.

PLOWMAN: REVISION OF BRUNFELSIA

15

of the pericarp and surface of the fruit may be

important in differentiating species; fruit charac-

ters should be used whenever possible. Unfortu-

nately, the fruits of many species are poorly or

not at all known.

The fruit of Brunfelsia is essentially a capsule,

not a berry as suggested by Miers ( 1 850) and oth-

ers in the past. There is, however, great variation

in the texture of the pericarp, which may be thin-

walled and brittle, leathery, fleshy or hard and in-

durated (bony). In shape it ranges from globose

to ovoid, and in size, from 1 to 5 cm long. There

may be two locules or only one by degeneration

of the dissepiments. The surface may be smooth

or rough; reticulate or areolate; and green, brown,

or bright yellow or orange. The pericarp may be

differentiated into three distinct layers or may be

more or less fused into one. In all cases the en-

docarp is a thin, somewhat cartilaginous layer that

on drying becomes brittle. The outer layers of the

fruit wall often separate from the endocarp when

the fruit rots. The thin endocarp may then split in

half to release the seeds.

The uncertainty and controversy concerning the

nature of the fruit in Brunfelsia (see "Ecology and

Reproductive Biology") reflect the scarcity of

fruiting material for many species. In reality, there

is a wide spectrum of variation in the texture of

the pericarp, which has led different authors to

classify the fruit variously as a berry, a capsule,

or a drupe. The fruits of certain species are indeed

difficult to classify, demonstrating the limitations

in our terminology. When the full range of forms

is seen, however, it is clear that all the fruits of

Brunfelsia are modifications of the capsule. Dif-

ferent types of capsule are found in each of the

three sections of the genus, and no type is unique

to any one section.

In section Brunfelsia the fruit is usually leath-

ery or fleshy and brightly colored. The calyx per-

sists only at the base. In B. americana and B.

densifolia, the valves are fleshy and yellow and

orange, respectively. The valves split apart at ma-

turity in a true dehiscence (Fig. 6). Brunfelsia ja-

maicensis is green at maturity, with a thick, bony

pericarp. Brunfelsia maliformis Urb. has a large,

fleshy to leathery fruit with a distinct reticulate-

areolate surface. It apparently does not open ex-

cept by rotting. (B. lactea has orange valves that

rot, allowing the seeds to be shaken out. — Eds.).

Some of the Cuban species have fruits that are

thin-walled, indehiscent, and usually brightly col-

ored (B. nitida, B. clarensis Britton & P. Wilson).

These may function as a berry in dispersal by an-

imals.

In section Franciscea the fruit is usually a rel-

atively small capsule that is green or brown and

has a coriaceous pericarp that becomes thin and

brittle on drying. The persistent calyx encloses the

fruit, at least in part. These capsules may split

open at maturity, but only sparingly. It may be

that many species are completely indehiscent and

their seeds are released only upon rotting on the

forest floor. Two anomalous species are B. dwyeri

(from Panama) and B. macrocarpa (from South

America), in which the fruits have a leathery and

a fleshy pericarp, respectively. The calyx of B.

macrocarpa increases greatly in size and itself be-

comes fleshy, completely enclosing the bright yel-

low fruits.

In section Guianenses we find fleshy, yellow

fruits in two of the six species, with the short ca-

lyx present only at the base. The fruits of B. ama-

zonica and B. burchellii are thin-walled and dry

at maturity; the fruit of B. martiana is as yet un-

known.

Seed

The seeds are oblong or ovoid and often an-

gular or compressed on one or more sides. They

range in size from 2 to 13 mm long and in di-

ameter from 1 to 7 mm. The surface texture is

reticular-pitted and reddish brown in color. The

surface pitting is formed by many tightly com-

pressed surface outgrowths, resembling globulose

hairs, that slough off as the seeds reach maturity.

[See Soueges, 1907; Edmonds, 1983; Lester &

Durands, 1984. These hairs are lignified thicken-

ings of the lateral testal cell walls. — Eds.] Each

of these hairs leaves a tiny pit in the surface. The

embryo is relatively large and straight and is em-

bedded in endosperm. The cotyledons are flat and

ovate to elliptic. The size of the seeds may be

useful as a taxonomic character, but their shape

and number per capsule appear to be quite vari-

able.

V. Cytology

The chromosomes of very few species of Brun-

felsia have been investigated. Four counts for the

genus are known from the literature, but these are

of limited value at the specific level because no

voucher herbarium specimens were prepared.

16

FIELDIANA: BOTANY

FIG. 6. Fruit of Brunfelsia americana.

PLOWMAN: REVISION OF BRUNFELSIA

17

TABLE 2. Chromosome counts in Brunfelsia.

Species

Chromosome no.

Author

B. americana

B. americana

B. australis

B. calycina (= B. pauciflora)

B. dwyeri

B. americana

B. australis

B. calycina (- B. pauciflora)

B. dw\eri

2n = 22

n = 11

n = 12

In = 22

2n = 22

2n = 22

n = 11

n = 12

2n = 22

2n = 22

Bhaduri (1933)

Madhavadian (1968)

Ratera (1944)

Darlington and Janaki Ammal (1945)

Davidse (1981)

Bhaduri (1933)

Madhavadian (1968)

Ratera (1944)

Darlington and Janaki Ammal (1945)

Davidse (1981)

These findings are summarized in Table 2. On the

basis of this evidence, Darlington and Wylie

(1955) hypothesized that the base number for

Brunfelsia is x = 11.

The mitotic chromosomes of Brunfelsia are dif-

ficult to prepare because the young root tips pro-

duced in flower pots tend to be hard and brittle,

making squashes very difficult. Treatment in 1%

HC1 softens the material somewhat, but despite

this treatment, no good mitotic figures could be

found in three species examined (B. australis, B.

latifolia, B. pauciflora). Sufficient seeds of Brun-

felsia were not available to test the possibilities of

preparing squashes from the roots of young seed-

lings.

A second method of obtaining chromosome

preparations is to examine pollen mother cells un-

dergoing meiosis. Young flower buds were col-

lected both in the field and from cultivated spec-

imens in the greenhouse and immediately placed

in Carney's solution for at least 24 hr and refrig-

erated when possible. Material was collected

about midday on sunny days. It was difficult to

obtain a sufficient number of buds because brun-

felsias generally have few flowers per inflores-

cence, and these are usually in various stages of

maturation; not infrequently only one flower per

inflorescence per day will be available for cytol-

ogy. Compounded by the fact that most species

flower only sporadically and some not at all under

greenhouse conditions, securing adequate material

was a task.

Meiosis occurs in Brunfelsia when the buds are

very small (about 1.0-2.5 mm long) and well be-

fore the corolla emerges from the calyx. The

young anthers were teased out in 2% acetocar-

mine and gently heated, then squashed with even

pressure. Relatively few countable meiotic figures

were obtained out of numerous attempts. The re-

sults obtained in five species and one subspecies

are summarized in Table 3.

All counts were obtained from meiotic figures

with the exception of B. pauciflora, which appar-

ently represents a somatic division in a tapetal

cell. The meiotic figures obtained were fixed in

late anaphase I or telophase II. The chromosomes

are small, from 2.0 to 5.5 (jtm (cf., Swanson,

1957), and exhibit some size variation between

the different species examined. In the case of B.

nitida and B. americana, the smaller size is due

to the fact that only two chromatids are present

(telophase II). The somatic chromosomes of B.

pauciflora appear greatly enlarged. This is prob-

ably the result of endomitosis, which produces a

high degree of endopolyploidy within the cell.

Unfortunately, no normal mitotic cells could be

obtained for this species.

The chromosomes obtained from these species

were very compact in the stages observed, sug-

gesting a high degree of contraction due to chro-

mosomal coiling. No details of chromosomal

structure could be ascertained from the meiotic

figures. It is possible, however, to see the centro-

mere in the somatic chromosomes of B. pauciflo-

ra.

In an extensive investigation, Gottschalk (1954)

examined the structure of the pachytene chromo-

somes of many species of Solanaceae, with par-

ticular reference to the presence and position of

heterochromatic and euchromatic areas. Because

of the high degree of contraction, he was unable

to obtain pachytene chromosomes in what he re-

ferred to as B. macrophylla (probably B. pauciflo-

ra). He did, however, observe the presence of

chromocenters in the interphase nucleus, indicat-

ing that the chromosomes of this genus are dif-

ferentiated into euchromatic and heterochromatic

regions. Similar results were found in species in

18

FIELDIANA: BOTANY

£

u

related genera (Browallia grandiflora sensu Gott-

schalk (1954) (probably B. speciosa Hook.) and

Nierembergia coerulea Gill, ex Miers).

From the preliminary results obtained in this

study, it may now be safely predicted that the base

chromosome number in Brunfelsia is x = 11. Ra-

tera's (1944) count of n = 12 in B. australis was

not confirmed because no flowering material was

available for this species. This count, however,

should be repeated at the earliest possible oppor-

tunity. Both section Brunfelsia and section Fran-

ciscea have 11 pairs. No material from section

Guianenses was available.

The most common chromosome number for the

Solanaceae is 2n = 24. There has been consider-

able debate about whether it can be assumed that

12 pairs represent the base number for the family.

Goodspeed (1954) presented evidence that the

primitive number for the family may be six pairs

(2n = 12), from which higher numbers were de-

rived by secondary association.

Related genera of the tribe Salpiglossideae have

various chromosome numbers. Four genera

(Streptosolen, Salpiglossis, Browallia, Brunfelsia)

have 1 1 pairs (2n = 22). Lower numbers occur in

Petunia (2n = 14), Nierembergia (2n = 18), and

Schizanthus (2n = 20). The endemic Australian

genus Duboisia has 30 pairs (2n — 60) (Darling-

ton & Wylie, 1955; Madhavadian, 1968).

The Salpiglossideae are thought to be derived,

at least in part, from the Cestreae, in which tribe

the following chromosome numbers have been

derived: Cestrum L. and Vestia Willd., eight pairs

(2n = 16); Fabiana Ruiz and Pavon, nine pairs

(2n = 18); and Nicotiana L., eight, nine, ten, 12,

18, 19, 20, 21, and 24 pairs, which represents a

well-known aneuploid series (Goodspeed, 1954).

Because the cytology of several important genera,

such as Juanulloa Ruiz and Pavon, Markea Rich.,

and Sessea Ruiz & Pavon of the Cestreae and An-

thocercis and Anthotroche of the Salpiglossideae,

have not been studied, it would be risky to draw

any phylogenetic conclusions on the basis of

chromosome number alone. It seems reasonable,

however, to suggest that Brunfelsia and other gen-

era of the Salpiglossideae originated from an an-

cestral 12-paired precursor through aneuploidy.

VI. Hybridization Experiments

As an aid to investigating the breeding systems

and interspecific relationships in Brunfelsia, a pro-

PLOWMAN: REVISION OF BRUNFELSIA

19

gram of artificial hybridization was undertaken.

Plants obtained from both wild and cultivated

sources were grown at the Biological Laboratories

Greenhouse, Harvard University, and at North-

eastern University Greenhouses in Woburn, Mas-

sachusetts.

There are several inherent difficulties in at-

tempting artificial hybridization in woody peren-

nials such as Brunfelsia. Although germination is

relatively rapid, the time required from seed to

flower varies from a minimum of 6 months to

several years. In Brunfelsia flowering may be spo-

radic or rare under greenhouse conditions, and in

most species only a few flowers per plant are open

at any one time.

Seventeen clones, representing ten species of

the genus, were available for making crosses, in-

cluding five each from section Franciscea and

section Brunfelsia. Other species were grown but

could not be induced to flower. A major problem

in making crosses was the lack of simultaneous

flowering in desired species. Some species (B.

americana, B. pauciflora) flowered almost contin-

uously throughout the year; others flowered only

occasionally or not at all. Attempts to store pollen

under refrigeration for future crosses were unsuc-

cessful.

Pollination was accomplished by excising the

corolla limb and uppermost part of the tube to

expose the anthers and pistil. Pollen was then

transferred using a small brush. The anthers usu-

ally dehisce before the complete unfolding of the

corolla lobes and scatter the pollen on the inside

of the tube. The stigma is receptive just before the

flower opens and remains so for 2-3 days.

Because of the small number of plants avail-

able, the results of the breeding experiments were

generally inconclusive. The majority of crosses

made, whether inter- or intraspecific, produced no

fruit. The flowers withered and persisted (B.

americana, B. australis) or dropped off in 1-3

days (B. densifolia, B. lactea, B. pauciflora).

The data obtained from the breeding program

are summarized in the following paragraphs. The

actual records of the crosses are not included be-

cause they are lengthy and, for the most part, neg-

atively repetitious. The following species were in-

cluded in the program: B. americana, B. australis,

B. densifolia, B. grandiflora, B. lactea, B. latifol-

ia, B. nitida, B. pauciflora, B. pilosa, and B. un-

dulata. Reciprocal crosses were made between

these where possible.

All species were artificially self-pollinated, al-

though selfing appears to take place naturally. The

dehiscing anthers lie very close to the stigma, and

loose pollen may easily fall upon it. However, no

viable seed was produced from any of the self-

pollinated individuals, indicating that these spe-

cies are self- incompatible. Fruits were occasion-

ally produced spontaneously in B. pilosa and B.

densifolia, but these contained completely aborted

seeds.

Mature fruit containing viable seeds was pro-

duced in the following intraspecific crosses be-

tween nonclonal individuals within the following

species: B. latifolia X B. latifolia, B. pauciflora X

B. pauciflora, B. americana X B. americana, and

B. densifolia X B. densifolia. However, even in

these species many of the crosses did not take or

produced aborted or hollow seeds. This may be

attributed to various factors. In some cases failure

to set seed may have been due to fluctuations in

greenhouse conditions or the use of certain pes-

ticides. Certain plants (B. americana, B. pauciflo-

ra) were of cultivated origin and may have been

clones of individuals highly selected for ornamen-

tal characters; reproductive irregularities are not

uncommon in these forms. Lastly, these failures

may have been the result of innate incompatibility

factors as yet uninvestigated in the genus.

The failure of interspecific crosses suggests that

all the species investigated are reproductively iso-

lated. In one case this may be significant in view

of specific delimitation. No viable seeds were pro-

duced in crosses between B. australis and B. pi-

losa, two species of southern Brazil that were for-

merly considered conspecific with B. uniflora.

This lends support to the view that B. australis

and B. pilosa should be considered distinct spe-

cies. Additional crosses involving these two spe-

cies as well as B. uniflora would be useful in con-

firming and further clarifying the relationships

among them.

VII. Ecology and Reproductive

Biology

Habitat

Brunfelsia species are found in a variety of hab-

itats. Species of section Brunfelsia of the Antil-

lean region are frequently encountered in rocky

woodlands or thickets on limestone (B. malifor-

mis, B. membranacea, B. splendida) or serpentine

soils (B. densifolia, B. shaferi Britton & P. Wilson,

B. sinuatd), often in relatively dry situations. Oth-

20

FIELDIANA: BOTANY

er species grow in humid, montane rain forests

(B. portoricensis) and in cloud forests at higher

elevations (B. lactea, B. jamaicensis). Most of the

species of this section are limited in ecological

tolerance and restricted to relatively small and lo-

cal areas. Only B. americana, which also has the

widest ecological amplitude, occurs on more than

one island (see Appendix II for a list of species

in section Brunfelsia and their distributions).

The basic ecology of section Brunfelsia, partic-

ularly the Cuban species, is virtually unknown. As

in the case of many plants of limited distribution,

many of the Antillean species are now in grave

danger of extinction because of the destruction of

their habitats. Several species in Jamaica are

known from a few local populations (e.g., B. ja-

maicensis, B. membranacea, B. plicata, B. splen-

didd) that are rapidly being encroached upon by

human activities. The only known wild population

of B. undulata, on the north coast of Jamaica, was

recently eliminated by the construction of a cattle

path. Brunfelsia picardae Krug. & Urb. of Haiti

has not been collected since 1928. Little is known

of the many Cuban endemics, the majority of

which are rare and known from only one or a few

collections.

The three Puerto Rican endemics are currently

protected from man's destruction because they are

now all included within the boundaries of national

parks. However, the only known population of B.

americana in Puerto Rico is precariously located

in a pasture near Coamo.

If the incredible increase in human population

in the West Indies continues at its present rate, the

sheer pressure of human needs will surely lead to

destruction of the few remaining natural areas.

Many of the Antillean species of Brunfelsia have

large, showy flowers of potentially great horticul-

tural value. Several of these have already been

successfully introduced into cultivation (B. amer-

icana, B. nitida, B. undulata), but the remainder

are in danger of being lost completely. In addi-

tion, several of these plants have been shown to

contain alkaloids or to be of interest medicinally.

In this study, an effort was made to distribute

seeds and cuttings of as many wild species as pos-

sible to tropical botanical gardens and conserva-

tories to preserve them in cultivation.

The six species of section Guianenses all grow

in low-elevation tropical rain forests in continen-

tal South America; they are found in the shrubby

stratum of the understory and at the margins of

rivers. Species of this group have not been studied

in the field, and little is known of their biology.

They are mostly known from only a few collec-

tions.

The South American species of section Fran-

ciscea are primarily shrubs or small trees of the

understory of humid tropical rain forests; those

species that have adapted to other environments

are considered more specialized. This group

grows primarily in mountainous regions, up to

about 2,000 m of altitude. Exceptions are found

in several species that appear to be restricted to

very low elevations (B. australis, B. imatacana

Plowman, B. latifolia, B. macrocarpa) and in B.

uniflora, which reaches 3,300 m in the Bolivian

Andes.

A number of species of section Franciscea

have become specialized in their ecological re-

quirements. Brunfelsia latifolia grows only on

sandy restingas, a maritime formation of low

shrubs on the coast of southeastern Brazil. Brun-

felsia obovata is adapted to semiaquatic environ-

ments and grows in marshes (brejos), often in

standing water, or along flooded river margins

from southeastern to central Brazil. In southern

Brazil, B. cuneifolia and B. pilosa are restricted

more or less to the region of the Araucaria forests

of Parana, Santa Catarina, and Rio Grande do Sul,

and B. rupestris grows only in the campo rupestre

habitat of the Serra do Espinha9o in Minas Gerais,

Brazil. Brunfelsia australis, the southernmost spe-

cies of the genus, grows in gallery forests and in

small patches of woodland found on raised

ground in the low-lying Parand River basin.

Two species have a wider ecological amplitude

and are found in varied habitats, including sec-

ondary forest growth and disturbed areas. Brun-

felsia brasiliensis grows in the humid forests of

Rio de Janeiro, especially in well-drained sites

such as outcrops and rocky peaks. It also flour-

ishes in the drier mountainous states of Minas

Gerais and Sao Paulo, often persisting in second-

ary growth woodlands and in gallery forests.

Brunfelsia uniflora is a wide-ranging species that

occurs in the rain forests of the Serra do Mar; in

the drier central and northeastern states of Brazil;

on the dry coast of Venezuela, and, lastly, in the

high-altitude cloud forests of the Bolivian Andes.

This is perhaps the most widespread species in

Brazil because of its ability to survive in disturbed

habitats.

Certain species of sections Franciscea and Gui-

anenses are of such local occurrence that they

may also be considered endangered. This is par-

ticularly true of B. latifolia, which is precariously

restricted to the region around the city of Rio de

PLOWMAN: REVISION OF BRUNFELSIA

21

Janeiro, and of B. dwyeri, on the upper slopes of

Cerro Jefe in central Panama. Both of these areas

are rapidly being developed for housing and farm-

ing, respectively. A large portion of Cerro Jefe

has already been decimated for building poultry

farms. Little is known of the effect of human ac-

tivities on other South American species because

they have not been studied in the field. The danger

of their extinction is less serious, however, than

for the West Indian species because of the usually

wider distribution of the continental species.

Life Cycle

The seeds of Brunfelsia species germinate with-

in 2-4 weeks of planting and are relatively short-

lived. They will not tolerate desiccation. Most

species are slow-growing woody shrubs that vary

in the time required to mature from seedling to

flowering plant. Under greenhouse conditions

plants mature from 6 months (B. latifolid) to 2 or

more years after germination. Several species are

known to flower while the plants are still small,

even though they may eventually become trees in

nature. Little is known about the longevity of in-

dividual plants, but there is evidence that they

grow for at least 20 years (B. grandiflora). Veg-

etative reproduction may also occur to a limited

extent by the rooting of broken branches on the

forest floor. This has been observed in B. chiri-

caspi (Plowman 2081).

The flowering period in Brunfelsia may be

short, lasting only a few weeks, or extend more

or less continuously throughout most of the year.

Flowering specimens have been collected from

the type locality of B. grandiflora subsp. schul-

tesii in January, May, and September. The major-

ity of the South American species flower between

October and December.

The fruits of Brunfelsia require several months

to a year for ripening and may remain on the plant

even after the seeds are fully mature, until the

fruits are dried and brown. Dehiscence of the

fruits is discussed under "Dispersal Mecha-

nisms."

Pollination Biology

The flowers of Brunfelsia are well adapted for

insect pollination, specifically by moths and but-

terflies. No other kinds of insects have been ob-

served to visit the flowers. The three basic forms

of the corolla found in the genus have apparently

evolved through selection for specific groups of

pollinators: section Franciscea for butterflies, sec-

tion Brunfelsia for hawk moths, and section Gui-

anenses probably for other moths [or perhaps

bees — Eds.]. All the species have a disc at the

base of the ovary that secretes abundant nectar in

the base of the tube. So far as is known, they are

all self-incompatible.

Species of section Franciscea have flowers

with a relatively short tube and a brilliant violet

limb with a white eye at the throat, which pre-

sumably serves as a nectar guide. The somewhat

raised, fleshy rim at the throat may function as a

landing platform for the visiting insect. The violet

color of the flower fades to pale lavender or pure

white with age and may signal the insect that the

flowers are no longer receptive. The flowers are

rarely scented in this group, although two species

(B. australis, B. imatacand) are exceptionally fra-

grant with a scent resembling that of jasmine.

Butterflies in four different families have been

collected on the flowers of B. grandiflora, includ-

ing species of Euptychia (Nymphalidae), Eurema

(Pieridae), Euselasia (Lycaenidae), and the Cal-

podes group (Hesperidae) (see Table 4). In south-

ern Colombia the most frequent visitor of these

was an unidentified species of Euptychia.4

Species of Brunfelsia sect. Brunfelsia, in con-

trast, possess a much longer corolla tube than

those of section Franciscea. They are usually

white upon opening and fade to yellow with age.

They bloom at night and are highly perfumed, be-

ginning to produce scent just before sunset. The

fragrance of a few flowers is capable of pervading

a relatively large area of forest. These flowers

have no landing platform as found in section

Franciscea, which suggests that the insect hovers

while feeding.

Although no insect pollinators have been seen

on the flowers of the Antillean species, these are

typically pollinated by moths. In view of the

length of the corolla tube (up to 14 cm), it may

be said with some certainty that hawk moths

(Sphingidae) are responsible for pollination in this

group. Similarly constructed, fragrant white flow-

ers in other families of plants are known to be

pollinated by hawk moths, such as Ipomoea alba

Garcke (Convolvulaceae), Aquilegia longissima

A. Gray (Ranunculaceae), Habenaria leucophaea

A. Gray (Orchidaceae), Hippobroma longiflora G.

4 Identified by Prof. John Burns, Museum of Com-

parative Zoology, Harvard University.

22

FIELDIANA: BOTANY

TABLE 4. Pollinators of Brunfelsia grandiflora.

Don (Campanulaceae), and species of Randia L.

(Rubiaceae), Quisqualis L. (Combretaceae), and

Brugmansia Pers. (Solanaceae).

The pollinators of section Guianenses are not

known. The flowers in this group have a short

tube but are white and have no landing platform;

both B. amazonica and B. chocoensis are reputed

to be fragrant, suggesting that they may also be

pollinated by moths. This remains to be confirmed

in the field.

Dispersal Mechanisms

The fruits of Brunfelsia are basically capsular

and of two general types: a more or less thin-

walled capsule that becomes dry and brittle with

age or a more or less thick-walled capsule, the

valves of which may be fleshy, leathery, or bony

at maturity. Both types of fruit occur in all three

sections of the genus.

The fruits of most species of Brunfelsia do not

dehisce to release the seeds as is common in most

capsules; at best they are tardily dehiscent. They

frequently remain on the plant well after the seeds

are ripe, becoming dry and brown, in which state

of brittleness they may then break apart on the

plant [but see discussion of fruit types in Solanum

in Symon (1979) — Eds.]. Alternatively, the whole

capsule may fall to the ground, where the outer-

most layer of the pericarp rots away, leaving the

thin, brittle endocarp. This eventually breaks in

half along the lines of dehiscence to release the

seeds.

Most of the South American species of Brun-

felsia have capsules that are green or brown at

maturity and included in the dark green calyx. As

a result, the fruits are very inconspicuous; fruiting

specimens are almost never collected. In the ab-

sence of any pronounced color, odor, or edible

flesh, it seems unlikely that the fruits could be

attractive to animals. These species apparently

have no special adaptations for dispersal, a situ-

ation that has been pointed out for many shrubs

and trees of the primary rain forest (Richards,

1952).

The thin-walled, capsular species may be dis-

persed over short distances by water. The air spac-

es created by placental breakdown and the lack of

dehiscence permit these fruits to float. After they

have fallen to the ground, they may be carried

away from the parent plant during heavy rains,

especially along river courses.

Other species in South America and in the West

Indies have rather large, orange or yellow fleshy

fruits that are probably dispersed by birds or

mammals. Examples of these are B. chocoensis,

B. densifolia, B. guianensis, B. macrocarpa, and

B. nitida. Two of these (B. chocoensis, and B.

macrocarpa) are said to be edible and probably

have a sweet flesh that would attract animals.

However, animal dispersal has not been observed

in the field.

Plant-Insect Relationships

In addition to pollinators, other insects are in-

timately associated with the genus Brunfelsia.

One of the most interesting is the butterfly genus

Methona (Rhopalocera, Nymphalidae), the larvae

of which feed exclusively on the leaves of Brun-

felsia species. This and related groups have been

studied by Dr. Gerardo Lamas of the Museo de

Historia Natural, Lima, Peru (Lamas, 1973).

Methona (Fig. 7) is a genus of medium-sized

butterflies (wing span of 7-10 cm) consisting of

seven species and nine subspecies. The larvae are

PLOWMAN: REVISION OF BRUNFELSIA

23

FIG. 7. Methona themisto (type specimen at BMNH).

gregarious and warningly colored (Fig. 8). The

species of Methona are all tropical forest insects

and are distributed in the same areas of South

America where Brunfelsia occurs. The greatest

concentration of Methona species is found in the

Andes; three species occur in the Amazon basin,

two occur in northeastern Brazil, and one occurs

in southeastern Brazil. A disjunct population, con-

sidered a distinct subspecies of M. confusa, lives

only on Cerro Jefe in Panama (Lamas, 1973).

Until the present, the documentation of the

Brunfelsia-Methona relationship has been patchy,

primarily because of the lack of voucher speci-

mens of the host plants. Larvae of M. themisto

megisto have been collected on B. australis in Mi-

siones Province of northeastern Argentina (Plow-

man 2731} as well as in Buenos Aires and Uru-

guay. Reports from the latter two sites were made

from cultivated plants that were incorrectly iden-

tified as B. pauciflora, B. uniflora, and B. latifolia,

even though none of these species are known this

far south in the wild or in cultivation. The same

species of Methona has been collected in Tucu-

man in western Argentina. Although no species

of Brunfelsia occurs there naturally, B. australis

is cultivated in gardens and must serve as the

host.

Other species of Methona have also been found

on Brunfelsia species: M. singularis in Paraiba in

northeastern Brazil and M. confusa confusa in

Manaus in the Amazon basin. Without voucher

specimens, it is difficult to say which species of

Brunfelsia were involved.

The geographical distribution of Methona ex-

hibits a striking parallel to that of the South Amer-

ican brunfelsias. The butterfly, however, does not

FIG. 8. Caterpillars of Methona confusa on leaf of Brunfelsia chiricaspi.

24

FIELDIANA: BOTANY

PLOWMAN: REVISION OF BRUNFELSIA

25

TABLE 5. Brunfelsia species recorded as larval food

plants for species of Methona (Lepidoptera: Nymphali-

dae: Ithomiinae).

M. M. M. M.

Species confusa megisto singularis themisto

B. amazonica •

B. australis •

B. dwyeri •

B. grandiflora • •

B. pauciflora • •

B. mire •

B. martiana •

B. uniflora • •

occur in the West Indies, suggesting that it arose

in South America after Brunfelsia sect. Brunfelsia

became cut off from its South American ances-

tors. Not enough data are available at the present

time to indicate whether there may be any spe-

cies-specific relationships between Brunfelsia and

Methona. A summary of the species of Brunfelsia

fed upon by the larvae of Methona is presented

in Table 5 [more detailed information can be

found in Drummond and Brown (1987) — Eds.].

The capsules of Brunfelsia species in South

America are also fed upon by Lepidopteran lar-

vae. The flesh of the young fruit and the seeds

are eaten by tiny larvae5 of the family Gelechi-

idae. These have been found in the following spe-

cies: B. australis (Plowman 2723), B. brasiliensis

subsp. macrocalyx (Dusen) Plowman (Plowman

2891, 2906), and B. grandiflora subsp. schultesii

(Plowman 2404). Further identification of the in-

sects has not been possible. They apparently lay

their eggs on or near the developing fruits. The

larvae then burrow through the pericarp and de-

vour the seeds and placental tissue.6

Protection from this predation may be the ori-

gin of certain specializations in the genus. The

fruits of several species (B. macrocarpa, B. hy-

drangeiformis, B. pauciflora) are completely en-

closed in the often coriaceous calyx, which may

offer some defense against the insect larvae. The

thick, leathery valves in the capsules of B. dwyeri

may serve a similar function.

A third insect-plant relationship has been ob-

5 Identified by Dr. Donald Duckworth, U.S. National

Museum. Washington, D.C.

6 The taxon-specific relationships between members

of the Gelechiidae and other Solanaceae have been well-

documented in the temperate zone but not in the tropics

(see Foott, 1967; Scholtz, 1978; Solomon & McNaugh-

ton, 1979) — Eds.

served in southern Colombia in B. grandiflora

subsp. schultesii. Two species of ants7 in the gen-

era Campanotus (Formicinae) (Plowman 2416)

and Pheidole (Myrmicinae) (Plowman 2151) form

temporary bivouacs among the fruit clusters of the

plants. Associated with the ants are two different

Homopteran insects (one in the family Psillidae,

the other unidentified) that feed on the plant and

apparently are tended by the ants. Pheidole man-

ufacture a nest of "carton" (cellulosic material)

among the fruits that may serve to protect their

charges. This may represent a symbiotic relation-

ship between the ants and Homopterans, but it is

not clear what advantage, if any, this may give to

the plant.

VIII. Trends in Specialization

In attempting to understand the evolutionary re-

lationships among the sections and species of

Brunfelsia, it is necessary to determine which

characters are primitive or unspecialized and

which are derived and specialized. Levels of spe-

cialization may be observed in several organs of

the plants, including leaves, flowers, and fruits, as

well as in habitat preferences and other ecological

factors. Because the species of section Brunfelsia

have not been studied in depth, the following dis-

cussion is limited primarily to the South American

species of section Guianenses and section Fran-

ciscea.

Of the South American brunfelsias, the flowers

of section Guianenses are the least specialized

and possibly ancestral to section Brunfelsia. The

corolla tube is short in section Guianenses and

open at the mouth with relatively narrow corolla

lobes. The corolla is white or greenish white, sug-

gesting pollination by moths.

Section Guianenses occurs primarily in the

Amazon basin with disjuncts in Choco and Bahia.

There is a distinct tendency in this group toward

axillary, few-flowered inflorescences on the older

branchlets, although terminal flowers are also pro-

duced. Two species (B. chocoensis, B. guianensis)

have capsules with fleshy valves, and two others

(B. amazonica, B. clandestina) have a dry, thin-

walled capsule. The fruit of the remaining two

species (B. martiana, B. burchellii) is unknown.

Evolution in the type of fruit, however, shows no

7 Identified by Prof. E. O. Wilson, Department of Bi-

ology, Harvard University.

26

FIELDIANA: BOTANY

overall direction because different types of fruit

(dry vs. fleshy, etc.) have probably arisen repeat-

edly under local selection pressures.

Section Guianenses is a small group that occurs

in a relatively uniform environment with few ma-

jor differences separating the species. It is there-

fore not surprising that trends of specialization are

difficult or impossible to detect within the group.

Section Franciscea apparently arose as a sec-

tion by becoming isolated in the Andean region,

where the flowers coevolved with butterflies. The

flowers are bright purple or violet with a distinct

white or yellow eye. The corolla tube is short and

somewhat constricted at the apex into a thickened

ring. The limb is often large and showy with

broad, rounded lobes.

The inflorescence of section Franciscea varies

from a many-flowered, lax or compact cyme to a

solitary terminal flower. Most of the species bear

a few-flowered terminal cyme; the number of

flowers per inflorescence may vary considerably

in a single species (see "General Morphology").

Evolution in the inflorescence must first of all

be considered in terms of the entire plant and pol-

lination strategy. Species with several to many

flowers in the inflorescence tend to have fewer

flower-bearing branches (B. chiricaspi, B. hydran-

geiformis, B. mire). Species with only one to three

flowers are generally more profusely branched,

with many of the branchlets bearing flowers (B.

australis, B. cuneifolia, B. uniflora}. In the first

case, the flowers are aggregated into a large single

mass (see Fig. 30); in the second, the entire shrub

may be diffusely covered with many single flow-

ers (see Fig. 45). Both modes of organization

must be equally attractive to pollinators, although

there may be different butterflies associated with

the different forms of inflorescence. This associ-

ation, however, has not been observed.

There is also a tendency for the one-flowered

species to have larger flowers (B. macrocarpa, B.

pilosa). Among the species with several (one to

ten) flowers, there are many intermediate situa-

tions with respect to the number of flowers per

branch, the number of branches, and the size of

flowers. In adapting to local ecological conditions,

available pollinators, etc., the species of section

Franciscea have evolved a mosaic pattern of in-

florescence organization within which these gen-

eral trends can be perceived.

Attempts to ascertain the "primitive" condition

of the inflorescence, i.e., whether it may have

been one-, few-, or many-flowered, branched or

unbranched, etc., leads to several problems. In

terms of flower and fruit characters, two species

appear to be least specialized, namely B. uniflora

and B. grandiflora. Both of these occur in the An-

dean region, as well as in eastern Brazil, indicat-

ing, as shown below [and in Plowman (1979) —

Eds.], that both species are relatively old for the

genus. However, B. uniflora, as its name suggests,

has but a single flower terminating the branch; B.

grandiflora may have few to many flowers in a

more or less branched inflorescence. We cannot,

therefore, decide which of these may have a more

primitive inflorescence solely on the basis of as-

sociated characters.

There has been considerable controversy over

the origin of inflorescence types in the Angio-

sperms (for discussion see Constance, 1955; Law-

rence, 1965). According to Lawrence, the inflores-

cence is a branch system in which a solitary flower

is the terminal unit of each branch. Some workers

believe that this single terminal flower is the prim-

itive inflorescence from which more complex

forms were derived through increased lateral

branching. Others, including Rickett (1944), whose

views have been widely repeated, interpreted the

sequence in the opposite direction. Rickett consid-

ered the primitive condition to be a compound,

branched inflorescence, specifically a compound

dichasium, from which few- or solitary-flowered

inflorescences were derived through suppression of

lateral branches. He derived various inflorescence

types, including the cincinnus, by reduction and

condensation of the branch system. The solitary-

or two- to three-flowered cyme could easily be de-

rived from the cincinnus by mere suppression of

the lateral branchlets. Accordingly, Lawrence con-

cluded that "the modern or present-day inflores-

cence represented by a solitary flower in almost all

plants is a product of reduction, a reduction that

may have resulted from any one of many multi-

flowered inflorescence types, whose origin may be

indicated only by comparative studies of inflores-

cences of related taxa."

One genus in the Solanaceae has been thor-

oughly studied from this point of view. Good-

speed, in The Genus Nicotiana (1954), postulated

that an open paniculate inflorescence (a broad

thyrse) is primitive in this group, from which a

wide array of inflorescences has been derived, in-

cluding glomerules and false racemes.

In conclusion and with reservations, I would

designate the pluriflorous, branched inflores-

cences of B. grandiflora as the probable ancestral

type for section Franciscea, from which fewer-

flowered forms such as B. bonodora (Veil.) J.F.

PLOWMAN: REVISION OF BRUNFELSIA

27

Macbr. and B. uniflora have been derived by re-

duction. Condensed, many-flowered inflores-

cences such as are found in B. mire and B. hy-

drangeiformis could easily be derived from that

of B. grandiflora by a shortening of the axis with

a concomitant increase in the number of lateral

branches.

Specializations in the calyx of Brunfelsia have

occurred both between and within the sections. In

section Brunfelsia the calyx is usually small and in

fruit appears only at the base. The same is true of

section Guianenses with the exception of B. ama-

zonica and B. clandestina, in which the calyx par-

tially encloses the thin-walled fruit. In section

Franciscea, the latter situation is the rule: the calyx

persists tightly appressed to the developing fruit,

enclosing only the lower half of the fruit (B. aus-

tralis, B. grandiflora) or far exceeding it in length

(B. hydrangeiformis, B. pauciflord). Even in the

fleshy-fruited members of this section (B. macro-

carpa, B. dwyeri), the calyx envelops the fruit and

in the former species becomes very large. The ca-

lyx in these species remains attached to the fruit

when it falls from the plant.

The relative length of the calyx at anthesis var-

ies considerably both within species (B. bras-

iliensis) and among them. However, a longer ca-

lyx in relation to the corolla tube indicates in-

creased specialization, possibly as protection

against robber bees or other insects that lance the

base of the corolla tube to obtain nectar. Another

feature of the calyx that is considered derived is

the inflated and five-angled condition found pri-

marily in B. boliviano, B. cuneifolia, and B. ob-

ovata.

The characters of the corolla were discussed

above. The corollas of each of the three sections

are distinctive, as demonstrated above, but show

little variation or evolution within each section.

However, minor differences do occur. In section

Brunfelsia there has been some differentiation in

the size of the corolla limb relative to the tube

and in the color of the corolla. In section Fran-

ciscea. although somewhat variable in size and to

a lesser degree in color, the corolla is relatively

uniform, as it also is in section Guianenses.

Equally conservative is the form and relative

size and position of the stamens and pistil. Within

each section, there has been apparently little

change. A tendency toward exsertion of the up-

permost pair of anthers has arisen convergently in

all three sections, but the adaptive significance of

this, if any, is not known.

Several different types of fruit are found in the

genus Brunfelsia and are distributed in all three

sections. These have been classified variously as

dehiscent or indehiscent berries; as fleshy, leath-

ery, or dry capsules; or as drupes. After examin-

ing most of the species for which fruits are

known, I would suggest that the basic or primitive

fruit in the genus is a tardily dehiscent capsule

with leathery valves. This basic form has under-

gone several modifications in response to partic-

ular environmental conditions and selection pres-

sures for dispersal.

One of the most generally encountered derived

characters is the loss of dehiscence, which may

be partial or complete, in both thin- and thick-

walled types. However, in many species, and es-

pecially in section Franciscea, the septicidal su-

tures (lines of dehiscence) remain on the fruit.

There may be two additional sutures that divide

the valves, apparently a vestige of previous four-

way dehiscence. This mechanism is seen today in

capsules of other genera in the Salpiglossideae,

such as Salpiglossis, Browallia, and Streptosolen,

as well as in Nicotiana and Sessea of the Cestreae.

In Brunfelsia dwyeri the sutures are cruciform and

prominently indented in the surface of the fruit.

A septicidal line of dehiscence is readily observed

in the endocarp of many species, which is usually

cartilaginous and brittle on drying. As the mature

fruit rots, the endocarp, which is often more du-

rable than the mesocarp, remains and eventually

breaks along the sutures through the agencies of

weathering and predation. This I believe to be a

vestige of the ancestral capsular form.

It is of interest here to compare the view of

other workers in the field of Angiosperm evolu-

tion. Hutchinson (1969) believed that there is a

general morphological trend within the Angio-

sperms in which the capsule has preceded the ber-

ry and drupe. Corner (1964), in summarizing his

"Durian Theory," stated that "the primitive fruits

of flowering plants were dehiscent arillate carpels

or, in the syncarpous ovary, capsules. These fruits

had bright red or orange walls and black seeds

covered with red or orange edible arils." From

this original type he derived the dry capsule with

dry seeds, indehiscent types that led eventually to

the fleshy berry, and the stone fruits and nuts by

lignification of part or all of the fruit wall.

Similar trends can be readily observed in the

Solanaceae operating at various taxonomic levels,

including tribe, subtribe, genus, and species. The

family is characterized by two main fruit types,

the berry and the capsule. Both types occur in

four of the five tribes of Wettstein (1895). The

28

FIELDIANA: BOTANY

fifth tribe, represented solely by Nicandra, bears

a "dry berry."

Besides Brunfelsia, other genera show hetero-

morphism in their fruits. The genus Lycium, mo-

nographed by Hitchcock (1932) and pointed out

by Croizat (1962), has mostly fleshy, tomato-like

berries except in two species, L. ameghinoi Spe-

gazzini and L. californicum Nutt., which bear cap-

sules. An anomalous species, L. macrodon A.

Gray, has a curious intermediate type of fruit.

A second example is found in the genus Cestrum

and its close relative Sessea. Francey (1935, 1936)

emphasized that the only "good" character to sep-

arate these two genera is the nature of the fruit,

which in Cestrum is baccate and in Sessea a four-

valved capsule. However, even this character, he

adds, breaks down in S. farinosa (Urb. & Ekman)

Francey, the young fruit of which is baccate, and

in C. strigillatwn Ruiz & Pavon and C. lindenii

Dunal, in which the young fruit bears a cruciform

suture at its apex, a feature typical of Sessea.

In that flowers and fruits tend to evolve as in-

dependent units (Davis & Hey wood, 1963), it is

not surprising to find both capsules and berries at

different levels in an evolutionary scheme based

primarily on floral morphology. Irrespective of

their relative primitiveness, the occurrence of two

types of fruits in the same subtribe or genus in-

dicates differential evolutionary rates. The fruit

has continued to change under varying conditions

of climate, habitat, and selection for dispersal, and

presumably different types of fruits may arise

more than once, indicating a mosaic rather than a

directional pattern of evolution. Constance (1955)

summarized this phenomenon as follows: "anal-

ogous dehiscent or indehiscent, dry or fleshy, and

one- or several-seeded fruit types can arise any-

where along this sequence, presumably as a result

of selection pressure of biological demands." De-

tailed morphological studies of fruits in all three

sections of Brunfelsia along with ecologically ori-

ented field work will add much to our understand-

ing of the dynamic relationships between these

basic fruit types.

Specialized features are also found in the vege-

tative organs, which in many cases can be related

directly to environmental conditions. Leaf size and

texture are particularly variable according to exist-

ing climatic conditions. Species of humid tropical

forests typically have relatively large leaves (B.

chiricaspi, B. hydrangeiformis, B. imatacana, B.

pauciflora, B. plicatd). Those of the more exposed,

drier habitats frequently exhibit smaller, often more

coriaceous leaves (B. brasiliensis, B. latifolia, B.

nitida, B. obovata, B. rupestris, B. splendida).

There are many intermediate situations as well as

species that show either type because they occur in

both habitats (B. uniflord). Several of the Antillean

species that inhabit serpentine soils have thick-co-

riaceous, linear leaves (B. linearis in Cuba, B. den-

sifolia in Puerto Rico, Fig. 4). Other species that

grow in high-altitude cloud forests have developed

thick-coriaceous, very shiny leaves (B. lactea in

Puerto Rico (Fig. 5), B. dwyeri in Panama, B. un-

iflora in Bolivia).

Another specialization in vegetative characters

is the crowded or subverticillate arrangement of

the leaves at the end of the stem (B. hydrangei-

formis, B. mire). This generally appears in con-

junction with a dense or capituliform inflores-

cence and very large leaves. The verticillate ar-

rangement of the leaves in B. rupestris is of a

slightly different nature, with crowded groups of

leaves occurring at intervals along the stem.

An example of convergent evolution has ap-

parently taken place in the leaves and fruits of B.

macrocarpa and B. chocoensis of the Pacific low-

lands of Colombia. These species belong to sec-

tion Franciscea and section Guianenses, respec-

tively. The leaves of these species are very similar

in size, shape, and venation, and both species

have fleshy, yellow fruits. However, the flowers

indicate that they belong to different phyletic

lines, although both are somewhat anomalous in

their sections.

The digression of various species from the trop-

ical forest environment and subsequent adaptation

to more specialized ecological zones, such as

rocky outcrops (B. brasiliensis), sandy shores (B.

latifolia), marshes (B. obovata), campo rupestre

habitat (B. rupestris), or coastal rocks (B. ameri-

cana), are also indicative of evolutionary special-

ization. These aspects are discussed in "Ecology

and Reproductive Biology." Further details are

found in "Taxonomic Treatment" under individ-

ual species entries. The trends of specialization in

the genus are summarized in Table 6.

IX. Geographical Distribution and

Relationships8

The current distribution of species and subspe-

cies, correlated with morphological, cytological,

8 Plowman (1979) provides an in-depth discussion of

the biogeography of Brunfelsia; the material in this sec-

tion is intended as a summary only. — Eds.

PLOWMAN: REVISION OF BRUNFELSIA

29

TABLE 6. Trends in specialization of Brunfelsia.

Less specialized

More specialized

Plants of humid, tropical forests

Leaves scattered on stem

Leaves large, soft-leathery

Inflorescence terminal and subterminal

Inflorescence lax, branched

Flowers few-many, small

Flowers scentless

Calyx shorter than corolla tube

Calyx terete, narrow

Calyx present at base of fruit only

Calyx not accrescent

Corolla tube straight, gradually dilat-

ed, and open at apex

Corolla lobes white or greenish, not

changing color

Corolla lobes narrow, not overlapping,

plane

Fruits dehiscent

Capsule valves leathery or fleshy

Plants of other habitats

Leaves crowded at branch tips or subverticillate

Leaves small, coriaceous

Inflorescence exclusively terminal or axillary

Inflorescence compact, sessile, and unbranched

Flowers solitary or few, large

Flowers fragrant

Calyx equaling corolla tube

Calyx angled, inflated

Calyx partially or totally enclosing fruit

Calyx more or less accrescent

Corolla tube curved and inflated at apex, then slightly constricted

and thickened to form ring

Corolla changing color (white to yellow, purple to white)

Corolla lobes broad, overlapping, undulate to crenate

Fruits indehiscent

Capsule valves thick, fleshy or bony, or thin-walled and dry

and other characters, provides us with the basis

for erecting hypothetical phylogenies that indicate

the possible historical relationships among pres-

ent-day species. Unfortunately, no fossil remains

of Brunfelsia or any related Solanaceae have been

found to corroborate the relationships proposed

here, nor have in-depth phylogenetic analyses

been undertaken.

In Brunfelsia evidence from morphology and

geographical distribution can be interpreted and

evaluated against the background of the geologi-

cal history and biogeographical patterns in other

plants and in animals to infer the history of the

genus. Such a scheme must be conjectural consid-

ering the paucity of data on several fronts, es-

pecially the all-important gaps in distributional

patterns, the result of inadequate collecting. Nev-

ertheless, the available data are full of implica-

tions that demand thoughtful consideration.

The major divisions of Brunfelsia, represented

by three sections, appeared as the result of an ear-

ly divergence in the ancestral, pre-Brunfelsia lin-

eage, which in turn arose from primitive Salpig-

lossoid stock somewhere in South America. These

three lines of divergence are evident today in the

basic differences in floral morphology among the

three groups and are associated with distinct and

mostly allopatric areas of distribution. This sug-

gests that the early differentiation of the ancestral

pre-Brunfelsia was accompanied by geographical

isolation in the three major areas in which they

are found today.

The three sections of Brunfelsia are centered in

three major areas: section Brunfelsia in the West

Indies; section Guianenses in northern South

America, primarily in the Amazon basin; and sec-

tion Franciscea throughout most of tropical South

America, excluding desert regions and the low-

land forest of the Amazon.

All the species occurring in the West Indies be-

long to Brunfelsia sect. Brunfelsia (Fig. 9). This

group, which exhibits both primitive and ad-

vanced characters, apparently diverged very early

from a South American progenitor that was pos-

sibly similar to present-day section Guianenses.

The most likely route of entry of the genus into

the Antilles seems to be via the inner arc of the

Lesser Antilles. These islands, which were formed

in the late Cretaceous or early Tertiary period,

have been volcanically active since the Eocene

epoch to the present (Malfait & Dinkelman, 1972)

[but see references cited in Liebherr, 1988, and

Page & Lydeard, 1994, for competing theories for

the origin of the Antilles — Eds.].

The hypothesis that the Lesser Antilles was the

entry route of Brunfelsia into the Caribbean is

based primarily on the present distribution of spe-

cies. The group most closely related to the Antil-

lean species is section Guianenses, which has its

center of diversity in northern South America.

The only species of Brunfelsia that occurs to-

day in the Lesser Antilles is B. americana, the

least specialized and most widely distributed spe-

cies of the genus. It is also found in Puerto Rico

and the easternmost tip of Hispaniola. This plant

has fleshy, bright orange fruits that may be carried

30

FTELDIANA: BOTANY

JAMAICA HISPANI

6 endemic/6 total 1 endemic/2 total

FIG. 9. Distribution of section Brunfelsia.

by birds from island to island. The agency of

long-distance dispersal by birds cannot be ignored

in considering the migration of Brunfelsia into the

Caribbean, although a path of "island hopping"

seems more logical. Brunfelsia americana occurs

as far south in the Lesser Antilles as St. Lucia;

currently prevailing arid conditions apparently ex-

clude it from the southernmost islands of the

chain. Although it is cultivated in St. Vincent and

Trinidad, B. americana is not known to occur nat-

urally in these islands at the present time.

No species are known from Central America

north of Panama. This greatly detracts from the

PLOWMAN: REVISION OF BRUNFELSIA

possibility that Brunfelsia may have entered the

Antilles via some former land connection with

Central America. Such a connection has been sug-

gested with regard to the Nicaraguan Rise, a shal-

low area in the Caribbean extending from Nica-

ragua toward Jamaica [but see papers on Carib-

bean biogeography cited above — Eds.].

About 20 species of section Brunfelsia are

known and are distributed throughout the major

islands of the Caribbean, excluding the Bahamas,

the eastern "limestone" islands, and the southern-

most Lesser Antilles. There is a high degree of

endemism on the largest islands, with about ten

BJQLOGY LIBRARY

101 BWWLL HALL

31

CCD

1 9 1999

FIG. 10. Distribution of sections Guianenses (dark gray shading) and Franciscea (pale gray shading); areas where

both sections occur are shown in black.

endemic species in Cuba, six in Jamaica, one in

Hispaniola, and three in Puerto Rico. Only B.

americana is found on more than one island.

This radiation of species, typical of island evo-

lution, appears to have arisen from reproductive

and ecological isolation of populations. Many of

the species are restricted to very local areas and

appear to be effectively allopatric. Nothing is

known of the breeding systems or pollinator-

flower relationships that may have played a major

role in reproductive isolation in these plants, es-

pecially among the sympatric species.9 A partic-

ularly fertile area in this regard is the Oriente

Province of Cuba, in which eight endemic species

have been reported. These questions, however, are

9 V. Fuentes, of the Botanic Garden, Havana, is pres-

ently working on this group; his results will be of great

interest in resolving some of the questions posed here. —

Eds.

beyond the scope of the present study, in which

only South American species and species com-

plexes are treated in depth.

Brunfelsia sections Guianenses and Franciscea

are essentially South American. Two species do

extend into Panama but are clearly of South

American origin. For the most part the two sec-

tions displace each other geographically (Fig. 10).

Section Guianenses, with six species, occupies the

lower Amazon basin, extending north into the

Guianas, south to Bahia in Brazil, and west almost

to the Colombian border. One species is isolated

in the Choco region of northeastern Colombia.

Section Franciscea, which includes 15 species, is

much more widespread, occurring over a large

portion of the South American continent in a di-

versity of habitats. These species are found from

the Caribbean coast south to about 30° latitude

and from sea level to about 3300 m altitude.

32

FIELDIANA: BOTANY

Most species of section Franciscea are found

in three major geographical provinces, corre-

sponding to the "regional centers" recognized by

Tryon (1972) for major areas of speciation in

ferns. The most outstanding of these, in terms of

numbers of species, is southeastern Brazil, which

has 1 1 species of Brunfelsia, nine of which are

endemic. The second major area is the Andean

region, which has five species, three of which are

endemic. The third region more or less agrees

with the Guiana center of Tryon and includes

parts of Guyana, Venezuela, and Roraima Terri-

tory in Brazil. Two species are found here; one is

endemic, and the other is widely distributed.

The patterns of distribution of the species dis-

cussed here may reflect to some extent the amount

of collecting in the different areas, especially

where gaps in distributions are found in poorly

collected areas such as the Amazon basin [see

Nelson et al., 1990, for a discussion of this prob-

lem— Eds.] Recent intensive collecting sponsored

by the New York Botanical Garden (New York,

N.Y.) and the Institute Nacional de Pesquisas da

Amazonia (Manaus, Brazil) have added many

new distribution records in formerly unknown ar-

eas. It is certain that the species that occur there

will be shown to be more widely distributed as

these areas become floristically better known.

It is here believed [see Plowman, 1979 — Eds.]

that most of the present distributions of Brunfelsia

in the circum-Amazonian region, and probably in

other parts of South America, are the result of

climatic and geological changes that occurred dur-

ing the Pleistocene epoch or more recently. Vuil-

leumier (1971), in a review article, summarized

these events in view of their effects on the flora

and fauna of the continent, with particular em-

phasis on the role of glaciation.10

Distribution patterns of Brunfelsia species oc-

curring in the Andes correspond surprisingly well

to the refugia of Haffer, Vanzolini, and Williams

[see Plowman, 1979 — Eds.]. It will be some time

before these theories can be tested and retested by

correlation with distribution patterns in other

groups of organisms and with the ever-increasing

body of knowledge of the geological and climatic

history of the South American continent. But for

the moment, the cyclic fragmentation and expan-

sion of forests and their component species pre-

sents a working model for understanding the iso-

10 The refugium hypothesis to which Plowman is al-

luding here and its present status are covered in a variety

of works (e.g.. Nelson et al., 1990; Mallet, 1993).— Eds.

lating barriers that have created the patterns of

differentiation and geographical distribution in

present-day species.

As here interpreted, the distribution and hypo-

thetical relationships of Brunfelsia sect. Francis-

cea suggest a complex picture of speciation, mi-

gration, and extinction of populations. Some spe-

cies are confined to relatively small areas in one

of the three major centers of distribution. Others

are found in more than one of these distant areas,

either as the same species or as vicarious pairs. It

is these plants that offer clues to the history of

their origin and distribution over wide areas.

An outstanding aspect of the distribution in sec-

tion Franciscea is the relationship between the

Andean and southeastern Brazilian centers, which

have one species and two pairs of vicarious spe-

cies in common. In floral and fruit characters

these species are relatively unspecialized with the

exception of B. hydrangeiformis. In terms of the

vegetative and inflorescence characters, B. gran-

diflora and B. bonodora are the least specialized

and are considered to be relatively old species that

may have given rise to others.

The present bicentric distribution of these spe-

cies suggests that the two centers may have been

in contact in the past. Several workers (Good-

speed, 1954; Smith, 1962; Vuilleumier, 1971)

have offered the opinion that the intervening re-

gion (eastern Bolivia, Chaco, Mato Grosso) may

have been wetter and cooler during the past, per-

haps in conjunction with periods of glaciation or

other climatic cycles. A more favorable climate

would have permitted the development of tropical

forests and an avenue for migration and gene ex-

change between the Andean region and south-

eastern Brazil.

Today this region is hot and dry and largely

covered with cerrado vegetation, an arid forma-

tion of shrubs and stunted trees. It is an inhospi-

table habitat for wet tropical forest species, serv-

ing as an effective barrier to gene flow. The final

uplift of the Andes was apparently responsible for

the rain shadow now existing in the Chaco-Mato

Grosso area. According to L. B. Smith (1962), the

southward flow of heat from the Amazon and the

northward encroachment of aridity from the south

via the Chaco depression resulted in a climatic

barrier that prevented further gene exchange be-

tween the two regions, creating many of the pres-

ent-day disjunctions of species between them.

During the hypothetical period of more or less

continuous forest between the Andes and southern

Brazil, species were free to migrate in either di-

PLOWMAN: REVISION OF BRUNFELSIA

33

rection. At this time the primitively Andean spe-

cies migrated into the Brazilian highlands and to

the Serra do Mar, where a secondary center of

speciation developed.

There is evidence that some of these species

eventually migrated back to the Andes during a

period of forest expansion and were later severed

from their parent populations by the unfavorable

climate that ensued. This appears to have occurred

in B. uniflora.

In many groups of plants, the area in which the

greatest diversity of species occurs is also the area

of origin of the group. This idea was particularly

favored by Vavilov (1950), who worked with the

wild progenitors of cultivated plants. Cain (1971),

however, has pointed out that this by no means

holds true for all groups of plants and has pro-

posed the origin of secondary centers of specia-

tion as follows:

It is entirely possible, then, that a phyletic stock that has

had its origin elsewhere, may, through migration, en-

counter a region in which there are numerous available

ecological niches that are unsaturated, that is, in which

competition pressure is low. Such a region may provide

a variety of habitats with at least partial evolutionary

radiation. It is apparent that such a region of polymor-

phism is not necessarily indicative of the original center

of origin or of dispersal, but is a fortuitously derived

center of differentiation. It may constitute a secondary

center of evolution and of dispersal.

To summarize the hypothetical relationship in

Brunfelsia sect. Franciscea, I suggest that this

group first appeared in the Andean sector of South

America and subsequently migrated to southeast-

ern Brazil. Here the species encountered extreme-

ly favorable and diverse environments in which

they became differentiated into several additional

species. Secondary and perhaps repeated contacts

with the Andean populations occurred as the new-

ly evolved species migrated far and wide in pe-

riods of optimal climatic conditions. During un-

favorable periods, large-scale extinction of popu-

lations occurred, creating the disjunct distribu-

tions that are seen today. Section Guianenses

apparently originated in northern South America,

probably in the Guiana-northeast Amazon region,

and spread throughout the lower Amazon basin

and to the Atlantic coast of Bahia. Other popu-

lations became isolated in the Choco region, giv-

ing rise to the now disjunct B. chocoensis.

X. Taxonomic Treatment

Brunfelsia Linnaeus

Brunfelsia L., Sp. PL: 191. 1753, ("Brunsfelsia");

Gen. PL ed. 5:87. 1754, non Braunfelsia Par.

1894. Lectotype species designated by Knapp in

Jarvis et al. (1993): B. americana L. Based on

B. flore albo, fructu croceo, molli. Plumier, PL

Am. 12. 1703; Nov. PL Am. Gen. t. 22. 1703.

Franciscea Pohl, PL Bras. Ic. 1: 1, t. 1-7. 1826. Lec-

totype species: F. uniflora Pohl. (See note under

synonymy of section Franciscea.)

Brunfelsiopsis (Urb.) Kuntze in Post & Kuntze, Lex.

81. Dec. 1903 ("1904"). Type species: Brunfel-

sia densifolia Krug & Urb.

Evergreen shrubs or small trees, 0.3-10 m tall.

Branches few to many, naked or leafy, terete.

Branchlets terete or somewhat angled in cross

section, glabrous or pubescent, the epidermis

splitting lengthwise. Leaves alternate, with 5/13

phyllotaxy, scattered or subverticillate, simple,

exstipulate, short petiolate, blade 3-30 cm long,

0.1-5 cm wide, elliptic, oblong, lanceolate, obo-

vate to oblanceolate, rarely spathulate, ovate, lin-

ear or linear-oblong, apically acuminate, acute,

blunt or rounded, entire at margin, basally nar-

rowed, cuneate or blunt, glabrous to pilose on up-

per surface, glabrous to velutinous on lower sur-

face, firmly membranaceous to thickly coriaceous,

lateral nerves pinnate, 3-12, spreading, arcuately

anastomosing near margin, veinlets finely reticu-

late. Inflorescence a cincinnal cyme or reduced

to a single flower, terminal or axillary, subsessile,

dense and compact, rarely lax and somewhat

branched, with 1-many flowers. Bracts 1-3 per

flower, 1-10 mm long, linear to lanceolate, ca-

ducous. Flowers white turning yellow, reddish

purple, or violet-purple fading to pale violet or

white, with or without fragrance. Calyx gamosep-

alous, 3-35 mm long, tubular to campanulate, gla-

brous or variously pubescent or glandular, light to

dark green or purplish, membranaceous to sub-

coriaceous, 5-dentate, teeth 1-10 mm long, sub-

erect, subequal, valvate in bud, triangular, ovate

or lanceolate, apically acute, acuminate or trun-

cate, minutely glandular-ciliolate; calyx in fruit

persistent, enclosing capsule or present only at

base, scarcely or rarely greatly accrescent, often

becoming coriaceous and dotted with lenticels,

striately nerved. Corolla gamopetalous, hypocra-

teriform, slightly zygomorphic with bilateral sym-

metry; tube 1.5-13 cm long, 1-5 mm in diameter,

1-24 times as long as calyx, cylindric, straight or

34

FIELDIANA: BOTANY

slightly curved, fleshy, gradually dilated or briefly

inflated at apex, open at mouth or slightly con-

stricted and thickened into fleshy white ring, gla-

brous or pubescent; limb 1.5-8 cm in diameter,

5-lobed, with quincuncial-imbricate estivation,

perpendicular to tube or inclined at an angle,

plane, undulate or crispate at margin; lobes sub-

equal, obovate or broadly rounded, lateral margins

overlapping or free, rarely abruptly deflexed. Sta-

mens 4, didynamous, completely included within

corolla tube or upper pair briefly exserted; fila-

ments 2-10 mm long, inserted in uppermost part

of corolla tube; lower pair posterior, shorter; upper

pair anterior, longer, surpassing the pistil, slender,

terete or somewhat flattened, curved posterially at

apex; anthers 1-3 mm long, all fertile, rarely up-

per pair reduced or rudimentary, medifixed, sim-

ple, 1 -celled, entire or slightly bilobate, oblong or

orbicular-reniform, emarginate, dehiscing by a

single longitudinal slit toward back of tube. Ova-

ry superior, 1.5-3 mm long, sessile, subtended by

short, nectariferous disc, bilocular, with 2 axile

placentae; style filamentous, broader, and curved

backward at apex, equaling the filaments; stigma

1-2 mm long, included between anthers, suben-

tire, shallowly bilobate and subcapitate or briefly

bifid in the form of a forceps with gaping lobes.

Ovules numerous, anatropous. Fruit capsular or

subbaccate, 1-5 cm long, subglobose to ovoid, bi-

locular or unilocular by dissepimental breakdown,

valves entire, exocarp thin, membranaceous or

subcoriaceous, smooth or becoming rugose-retic-

ulate, with corky outgrowths, mesocarp thin co-

riaceous, dry at maturity, or thick, leathery or

fleshy, endocarp thin, cartilaginous, drying brittle,

sparingly dehiscent or indehiscent. Seeds few-nu-

merous, 2-13 mm long, 1-7 mm in diameter, ob-

long, concave, angular, reticulate-pitted, reddish

brown, partially immersed in fleshy placenta or

drying free. Embryo straight or slightly curved,

imbedded in endosperm; cotyledons flattened,

ovate-elliptic. Chromosome number 2n = 22, 24?

The genus Brunfelsia, as here circumscribed,

consists of 46 species of tropical American shrubs

or small trees. The genus is divided into three

sections: Brunfelsia, Guianenses, and Franciscea.

Section Brunfelsia consists of 22 species that are

restricted to the West Indian islands. These spe-

cies are not included in the present revision, but

a brief description of the section and a list of the

names of the species and synonyms are given in

Appendix II. Section Guianenses contains six spe-

cies that are found in South America, being par-

ticularly common in the Amazon basin and east-

ern Brazil. Section Franciscea consists of 18 spe-

cies that are found in tropical areas of South

America, two of which also occur in Central

America.

Artificial Key to the Sections, Species, and Infraspecific Taxa of Brunfelsia

1 . Corolla tube 6-24 times as long as calyx, white turning yellow with age, or reddish. West Indies

Section Brunfelsia

1. Corolla tube 1-4 times as long as calyx, purple fading to white or pale violet, or white to greenish

white, not turning yellow. South and Central America

2. Corolla tube open at mouth, not constricted; lobes narrow, deflexed at lateral margins or scarcely

overlapping, white or greenish white Section Guianenses

3. Inflorescence sessile or with a very short peduncle; capsule dry at maturity, thin-walled

4. Pedicels elongate, 13-20 mm long; calyx glabrous 1. B. amazonica

4. Pedicels short, 3-9 mm long; calyx teeth minutely glandular-papillose 4. B. clandestina

3. Inflorescence pedunculate, the penducle occasionally very short but always visible; capsule at

maturity fleshy (or unknown)

5. Inflorescence bracts not ciliolate at margins, glabrous or very occasionally villous

6. Leaves obovate to obovate-elliptic, with 4-5 lateral nerves; calyx 7-10 mm long. Brazil

(Para), Guianas 5. B. guianensis

6. Leaves ovate, with 5-9 lateral nerves; calyx 1 1-20 mm long. Brazil (Rio Tocantins)

2. B. burchellii

5. Inflorescence bracts ciliolate at the margins, otherwise glabrous

7. Petiole 10-15 mm long; leaf blade 9-11 cm wide, with 6-7 lateral nerves; corolla limb

22-25 mm in diameter. NW Colombia, Panama 3. B. chocoensis

1. Petiole 1-4 mm long; leaf blade 4-8 cm wide, with 8-12 lateral nerves; corolla limb

15-22 mm in diameter. E Guyana 6. B. martiana

PLOWMAN: REVISION OF BRUNFELSIA

35

2. Corolla tube constricted at mouth forming thickened ring; lobes broadly rounded, overlapping at

lateral margins, purple or violet fading to pale violet or white Section Franciscea

8. Calyx distinctly 5-angled, inflated, drying plicate

9. Leaves cuneate to short acuminate at apex; pedicel and calyx glandular or villous

pubescent; calyx teeth ovate to lanceolate, 2-7 mm long

10. Leaves oblong-obovate to elliptic-lanceolate, the margins glabrous; calyx teeth

ovate-lanceolate, 3-7 mm long. S Brazil 12. B. cuneifolia

10. Leaves obovate to elliptic-obovate, the margins villous ciliate; calyx teeth

ovate, 2-4 mm long. Bolivia 8. B. boliviano

9. Leaves blunt to subacute at apex; calyx teeth broadly ovate, 1-4 mm long. E Brazil

20. B. obovata

11. Leaves firmly membranaceous to subcoriaceous; flowers (1)2-5; pedicels 1-8

mm long; calyx tubular, 16-22 mm long; corolla tube l-lVi times as long as

calyx 20a. B. obovata var. obovata

11. Leaves thick coriaceous; flowers (1-2)3-10; pedicels 5-12 mm long; calyx

10-16 mm long; corolla tube 1V2-3 times as long as calyx

20b. B. obovata var. coriacea

8. Calyx terete, scarcely inflated, drying flat

12. Leaves usually 16-31 cm long

13. Flowers solitary on branch tips; calyx greatly accrescent in fruit; capsule

fleshy, 4-5 cm in diameter. Pacific coast of Ecuador and Colombia

18. B. macrocarpa

13. Flowers 3 or more per inflorescence; calyx very slightly or not at all accres-

cent; capsule dry, thin-walled at maturity, less than 2 cm in diameter

14. Corolla tube l-Wi times as long as calyx; capsule completely enclosed

in calyx; calyx 1.5-3.5 cm long. SE Brazil ... 15. B. hydrangeiformis

15. Leaves subspathulate to broadly oblanceolate, reticulately rugose

with sulcate nerves; pedicel 2-4 mm long; calyx 3-6 mm in diam-

eter with yellowish brown pubescence; calyx teeth linear to linear-

lanceolate, 4-10 mm long

15a. B. hydrangeiformis subsp. hydrangeiformis

15. Leaves oblong to oblong-lanceolate, smooth or sulcately nerved;

pedicel 4-8 mm long; calyx 5-10 mm in diameter, glabrous or with

sparse glandular trie homes; calyx teeth ovate to ovate-lanceolate, 2-

4 mm long 15b. B. hydrangeiformis subsp. capitata

14. Corolla tube 2-5 times as long as calyx

16. Corolla lobes deflexed at anthesis; flowers few, 4-7(20); leaves 7-12

cm wide. Amazonian Colombia and Ecuador ... II. B. chiricaspi

16. Corolla lobes spreading at anthesis; flowers (3)5-many; leaves less

than 8 cm wide

17. Leaves subverticillate, with 8-13 lateral nerves; inflorescence

dense, capituliform; corolla tube 25-38 mm long. Peru (Cuzco,

Puno), Bolivia, W Brazil 19. B. mire

17. Leaves scattered, with 5-9 lateral nerves; inflorescence lax, of-

ten short-branched; corolla tube 15-40 mm long. W South

America 14. B. grandiflora

18. Corolla tube 30-45 mm long; limb 35-52 mm across; cap-

sule 1.7-2.2 cm long

14a. B. grandiflora subsp. grandiflora

18. Corolla tube 15-30 mm long; limb 20-40 mm across; cap-

sule 1—1.6 cm long 14b. B. grandiflora subsp. schultesii

12. Leaves usually 2-15 cm long

19. Calyx with short or long, nonglandular trichomes

36 FIELDIANA: BOTANY

20. Flowers 8-many per inflorescence; pedicels articulated with short, per-

sistent peduncle; trichomes sparse to dense, yellowish to yellowish

brown; corolla tube 17-22 mm long. SE Brazil . . lOa. B. brasiliensis

subsp. brasiliensis

20. Flowers all 1-3 per inflorescence; pedicels not stalked; trichomes long,

unbranched, pilose; corolla tube 25-32 mm long. S Brazil

22. B. pilosa

19. Calyx glabrous or with sparse glandular hairs

21. Corolla tube 25-40 mm long

22. Flowers solitary on branch tips; leaves thick coriaceous; capsule ru-

gulose, pericarp leathery, 1-4 mm thick. Panama . . 13. B. dwyeri

22. Flowers 1-many per inflorescence; leaves membranaceous to sub-

coriaceous; capsule smooth, pericarp thin- walled, drying brittle

23. Corolla tube 1-2 times as long as calyx; capsule completely

enclosed by calyx

24. Pedicel articulating with short peduncle; peduncle 2-5 mm

long; corolla limb 25-32 mm across; leaves 1.2-2.2 cm

wide. SE Brazil . . lOb. B. brasiliensis subsp. macrocalyx

24. Pedicels sessile or nearly so; corolla limb 40-80 mm across;

leaves 2-6.5 cm wide

25. Leaves oblong to oblong-lanceolate, with 5-11 lateral

nerves; flowers 1-11; pedicel 11-25 mm long; corolla

tube 1-1 Vi times as long as calyx. SE Brazil . . 21. B.

pauciflora

25. Leaves oblong-elliptic to oblong-obovate, with 6-7 lat-

eral nerves; flowers 1-2; pedicel 8-9 mm long; corolla

tube 1^2-2 times as long as calyx. E Venezuela

16. B. imatacana

23. Corolla tube 2-5 times as long as calyx; capsule exceeding the

calyx in fruit

26. Flowers 1-3(4) per inflorescence; leaves elliptic-obovate,

apically blunt to acute. S Brazil, Paraguay, Argentina ....

7. B. australis

26. Flowers (3)5-many per inflorescence; leaves oblong-lanceo-

late to lanceolate, apically acuminate. W South America

14. B. grandiflora

21. Corolla tube 15-25 mm long

27. Flowers all solitary. Widespread 24. B. uniflora

27. Flowers >1 per inflorescence

28. Leaves congested near branch tips, with strongly revolute mar-

gins, 1-4 cm long. Campo rupestre, Brazil . . 23. B. rupestris

28. Leaves scattered along branches, not strongly revolute, usually

>5 cm long. Forests

29. Leaves elliptic-oblong, ovate or obovate, blunt or abruptly

short acuminate. Rio de Janeiro 17. B. latifolia

29. Leaves narrow-elliptic to oblong-lanceolate, apically acu-

minate

30. Pedicels 6-12 mm long; calyx in fruit nearly as long as

capsule. SE Brazil 9. B. bonodora

30. Pedicels 2-6 mm long; calyx in fruit half as long as

capsule or less. W South America . . 14b. B. grandi-

flora subsp. schultesii

PLOWMAN: REVISION OF BRUNFELSIA 37

Brunfelsia sect. Guianenses Plowman in Hawkes,

J. Linn. Soc. (Bot.) 76: 294. 1978.

TYPE SPECIES — Brunfelsia guianensis Benth. in

DC., Prodr., 10: 200. 1846.

Shrubs to 2 m tall. Inflorescence terminal and

axillary. Flowers small, 1-7 per inflorescence.

Corolla white or greenish white, tube 2-3 times

as long as calyx, open at mouth, not constricted,

the lobes narrow, usually with deflexed margins.

Stigma briefly bifid, in the form of a forceps. Cap-

sule dry or fleshy at maturity. South America.

1. Brunfelsia amazonica C.V. Morton, Proc.

Biol. Soc. Wash. 62: 151. 1949. Type: BRA-

ZIL Amazonas: Manaus, Estrada do Raiz, sil-

va secundaria, non inundabilis, manacd, 24

Mar. 1937, Ducke 430 (holotype, us

1693434; isotypes, A, F-902265, K, MO-

1156896, NY, R-75434, s). Figure 11.

Shrub or small tree 1.5-2.5 m tall. Branches

spreading, knobby. Bark buff to reddish brown,

cracked in longitudinal fissures. Brant hk-ts shiny,

glabrous, light, yellowish brown. Leaves scattered

on branchlets, blade 6.5-12 cm long, 2.5-5 cm

wide, elliptic-oblong to lanceolate, abruptly acu-

minate, acumen often oblique, broadly cuneate to

narrowed at base, glabrous, dark green above,

pale green beneath, firmly membranaceous, lateral

nerves 5-8(10), straight; petiole 2-5(8) mm long,

transversely corrugate with cracked epidermis,

glabrous. Inflorescence terminal or axillary, ses-

sile with 1-2(5-6) flowers. Flowers white or

greenish white, fragrant. Bracts 1-2 per flower,

2-6 mm long, linear to lanceolate, glabrous, ca-

ducous. Pedicel curved in bud, becoming erect,

1 3-20 mm long, 1 mm in diameter, slender, thick-

ened toward apex, glabrous. Calyx 8-12 mm

long, 5-8 mm in diameter, tubular-campanulate,

slightly inflated, obovate in bud, glabrous, green,

membranaceous, teeth 2-6 mm long, erect, some-

what unequal, ovate-lanceolate, apically blunt to

acuminate; calyx in fruit, enclosing capsule at

base. Corolla tube 20-25 mm long, 1.5-3.0 mm

in diameter, twice as long as calyx, cylindric,

slightly dilated toward mouth, glabrous, minutely

glandular at mouth; limb 15-25 mm in diameter,

spreading, lobes 6-11 mm long, subequal or the

posterior lobe somewhat larger, oblong to obo-

vate, somewhat convex, blunt or rounded at apex,

the sides of the lobes reflexed after anthesis, nar-

rowed slightly toward base. Stamens inserted at

middle of tube, included; filaments subligulate,

upper pair 5 mm long, lower pair 2-3 mm long;

anthers 1-1.5 mm in diameter, orbicular-reniform.

Ovary 1.5 mm long, 1 mm in diameter, narrowly

conical; style 16-18 mm long; stigma 1 mm long,

bifid, upper lobe somewhat larger. Fruit 10 mm

in diameter, globose, apiculate at apex, green,

smooth, pericarp thin, about 0.2 mm thick, dry at

maturity, sparingly dehiscent. Seeds 8-10, 5-6

mm long, 3 mm in diameter, oblong-reniform,

flattened on 1 or 2 sides, angular, dark brown,

reticulate-pitted with a medial hilum. Embryo 3

mm long, slightly curved; cotyledons about 0.5

mm long, elliptic.

DISTRIBUTION — Brazil (Amazonas). See Figure

15.

SPECIMENS EXAMINED — BRAZIL. Amazonas:

Manaus, Estrada da Raiz, 18 Mar. 1943, Ducke

430 (paratype, MG, us); perto da Estrada da Raiz,

22 Mar. 1932, Ducke sp. nov. (GH, RB); Mata do

Aleixo, 4 Mar. 1945, Froes 20535 (NY, us); Uyp-

iranga, Rio Negro, perto de Manaus, 21 Feb.

1923, J.G. Kuhlmann 943 (GH, RB); road Manaus

(Cacau-Pireira) to Manacapuru, km 25, 3 Jan.

1967, Prance, Pena & Ramos 3880 (F, INPA, K,

MG, P, R, s, us); ad oram meridionalem Rio Negro,

usque ad concursum flum. Solimoes, May 1851,

Spruce 1495 (EM, BR, CGE, F, G, GH, K, LE, MG, NY,

p, RB, w); *Rio Negro, Tapuruquara, Santa Izabel,

Black 48-2905 (IAN).

This species was first collected by Spruce in the

middle of the last century and distributed under

the name B. spruceana to the world's herbaria.

This name was recorded but not validly published

by Schmidt, who placed it in synonymy with B.

maritima Benth.

Brunfelsia amazonica is a very local species,

known only from the environs of Manaus at the

mouth of the Rio Negro in Brazil. Notwithstand-

ing the intensity of collecting in this region, B.

amazonica has not been collected often. It is a

very distinct species of section Guianenses and is

easily separated from B. guianensis and B. cho-

coensis by the longer pedicels, the oblong-lanceo-

late leaves, and dry, thin-walled fruit, and from

B. martiana by the longer pedicels and smaller

leaves.

Brunfelsia amazonica grows in open secondary

forests on the southern bank of the Rio Negro in

white sandy soil. It is known to flower from Jan-

uary to May; fruits have been collected only in

38

FIELDIANA: BOTANY

FIG. 1 1 . Brunfelsia amazonica.

PLOWMAN: REVISION OF BRUNFELSIA

39

May. Like many brunfelsias in Brazil, this species

is locally called manacd.

2. Brunfelsia burchellii Plowman, Fieldiana Bot.

n. s. 8: 9. 1981. Type: BRAZIL. Goias: Porto

Real (now Porto Nacional), ford of Igarape,

1828-1830, Burchell 8527 (holotype, P; iso-

types, K, L).H Figure 12.

Shrub. Mature branchlets 2-3 mm in diame-

ter, spreading, with shiny, yellowish to dark

brownish, longitudinally cracked bark. Leaves

scattered along stem, short petiolate, blade 70-

180 mm long, 30-75 mm wide, broadly to nar-

rowly ovate, apically acuminate with a long,

pointed, often falcate acumen, sometimes abruptly

acuminate, minutely ciliolate at margin, basally

rounded to obtuse, glabrous on both sides, dull or

somewhat shiny, medium green above, dull, paler

green beneath, firmly membranaceous to charta-

ceous, the lateral nerves 5-9, strongly arcuate,

anastomosing near margin, nerves prominent on

lower surface; petiole 1-6 mm long, with a few

scattered glandular hairs, glabrescent. Inflores-

cence terminal on current year's branchlets, very

briefly pedunculate, with 1-3 flowers. Peduncle

14 mm long, persistent, sparsely glandular-pubes-

cent, glabrescent. Bracts 1-3 per flower, 4-20

mm long, linear to lanceolate, apically acuminate,

glabrous or villous at nerves and margin, cadu-

cous. Flower color unknown. Pedicel 4-8 mm

long, 1 mm in diameter, sparsely glandular-pubes-

cent, glabrescent. Calyx 1 1-20 mm long, 5-9 mm

in diameter, tubular or tubular-campanulate, te-

rete, with scattered glandular hairs, glabrescent,

striately nerved, teeth 2-6 mm long, unequal, tri-

angular-ovate, apically acute to acuminate, mi-

nutely gland-tipped. Corolla tube 24-28 mm

long, 1.5-2 mm in diameter, \Vi-2 times as long

as calyx, straight, glabrous or bearing few scat-

tered glandular hairs, orifice 5 mm in diameter;

limb 1 8-30 mm in diameter, spreading, lobes sub-

equal, rounded, overlapping at the lateral margins.

Stamens inserted in upper part of corolla tube;

filaments 1 mm wide, straplike, the anterior pair

3.5-6 mm long, apically slightly incurved or sub-

erect, the posterior pair 3-4 mm long; anthers 1

mm in diameter, orbicular-reniform. Ovary 1.5

mm long, conical-ovoid; style 20-21 mm long,

broadened and incurved at apex; stigma included

between pairs of anthers, 1 mm long, briefly bifid,

11 All Burchell localities are taken from his itinerary,

published by Smith and Smith (1967).

upper lobe slightly larger. Fruit and seed un-

known.

DISTRIBUTION — Basin of the upper Rio Tocan-

tins (Goias State) and adjacent Maranhao State,

Brazil. See Figure 15.

SPECIMENS EXAMINED — BRAZIL. Goias: Porto

Real, Porto Real to Igarape, 1828-1830, Burchell

8415 (L); about village, Burchell 8494 (K); at en-

trance to village, Burchell 8653 (K); "about the

Manga" (cf., Rio dos Mangues, where it meets

the Tocantins), Burchell 8752 (CH). Maranhao:

Island of Sao Luiz, Estrada do Barreto, Feb.-Mar.

1939, Froes 11620 (A, F, NY, s, us).

Brunfelsia burchellii is known from only a few

collections by William J. Burchell in the vicinity

of Porto Real (now called Porto Nacional) on the

upper Rio Tocantins 150 years ago. Unfortunately,

these specimens bear no field data, and the plant

has not been recollected in this region. A modern

collection by Froes from Sao Luiz Island, more

than 1,000 km north of Porto Nacional, is tenta-

tively assigned to B. burchellii but lacks mature

flowers or fruits and differs somewhat in leaf ve-

nation.

Brunfelsia burchellii appears to be most closely

related to B. guianensis, from which it differs by

the ovate, acuminate leaves, the longer tubular ca-

lyx, and the broadly rounded corolla lobes. With

B. guianensis, B. burchellii is provisionally placed

in section Guianenses on the basis of the gradu-

ally dilated and not apically constricted corolla

tube.

Two additional collections with ovate acumi-

nate leaves appear to be related to B. burchellii

but cannot be placed with certainty. One of these,

Ducke s.n. (RE 18141), was collected at Bragan?a

in Para and bears small, white, terminal, solitary

flowers and immature leaves. The other, Sucre &

da Silva 9204 (F, RB), was collected at Buriti dos

Lopes, Piaui, and has rather shiny, thick leaves

and immature fruits completely enclosed in per-

sistent, accrescent calyces.

3. Brunfelsia chocoensis Plowman, Bot. Mus.

Leafl. 23(6): 245, t. 14. 1973. Type: CO-

LOMBIA. Antioquia: Chigorodo, forest just

southeast of Chigorodo, 45 km S of Turbo,

alt. 50 m. Shrub 2 m high. Flowers rather

conspicuous, of a peculiar flat white color,

like unglazed paper, 15 Apr. 1945, Haught

4563 (holotype, uc M 048365; isotypes, NY,

COL 109727). Figure 13.

40

FIELDIANA: BOTANY

BRUNFELSIA

burchellii Plowman

FIG. 1 2. Brunfelsia burchellii. Reproduced courtesy of the Field Museum of Natural History.

PLOWMAN: REVISION OF BRUNFELSIA

41

BRUNFELSIA chocoensis

Plowman

FIG. 13. Brunfelsia chocoensis. Reproduced courtesy of the Botanical Museum of Harvard University.

Shrub 2 m tall. Branches sparse, terete, some- long, 9-11 cm wide, elliptic to oblong, abruptly

what knobby at nodes, naked, gray or grayish to long acuminate at apex, acumen often subfal-

brown. Branchlets 3 mm in diameter, glabrous, cate, obtuse at base, glabrous, dull, dark green

shiny, light grayish brown. Leaves few, usually above, pale green beneath, firmly membrana-

crowded at ends of branchlets, blade 18-28 cm ceous, lateral nerves 6-7, spreading; petiole 10-

42

FIELDIANA: BOTANY

15 mm long, stout, brown. Inflorescence subses-

sile, terminal, 1 -flowered, or, on previous branch-

es, axillary with 1-3 flowers. Flowers white, fra-

grant. Bracts 2-4 mm long, ovate to ovate-lan-

ceolate, concave, ciliolate at margin, caducous.

Pedicel 5-6 mm long, erect, slender, glabrous.

Calyx 8-12 mm long, 3-6 mm in diameter, tu-

bular-campanulate, globose in bud, glabrous, firm-

ly membranaceous, veinlets prominulous, reticu-

late, teeth 2-4 mm long, erect, ovate, blunt to

acute; calyx in fruit to 13 mm long, persistent,

dotted with lenticels. Corolla tube 22-27 mm

long, 2-2.5 mm in diameter, twice as long as ca-

lyx, gradually dilated toward mouth, glabrous;

limb 22-25 mm in diameter, spreading from

rounded mouth 5 mm in diameter, the lobes 8-10

mm long, subequal, rounded at apex, rarely emar-

ginate, slightly overlapping at margin above base.

Stamens included in upper third of tube; filaments

5 mm long, somewhat flattened, lower pair curved

at apex, upper pair suberect, glabrous; anthers 2

mm long, orbicular-reniform, light brown. Ovary

2.5 mm long, ovoid; style 10-20 mm long, slen-

der; stigma 1.5 mm long, bifid, gaping, the upper

lobe somewhat larger. Fruit 2-3 cm long, 2-3 cm

in diameter, globose to ovoid, smooth, yellow

when ripe, mesocarp soft, fleshy, up to 1.5 mm

thick, endocarp thin, cartilaginous. Seeds few, 2-

5, 8-10 mm long, 6-7 mm in diameter, ovoid-

ellipsoid, angular, dark brown, reticulate-pitted.

Embryo 8 mm long, straight; cotyledons 3.5 mm

long, widely elliptic.

DISTRIBUTION — Panama, Colombia. See Figure

15.

SPECIMENS EXAMINED — PANAMA. Darien:

Cerro Pirre, 11 Apr. 1967, Bristan 566 (ECON,

osu), Bristan 569 (ECON, osu), * Bristan 1220

(MO), *Duke 5300 (MO), ^Gentry & Clewell 7138

(MO); *along Rio Armila, WSW of Puerto Obal-

dia, Mori 6828 (MO). COLOMBIA. Choco: Hy-

dro Camp No. 14, Rio Salaqui, 6 days upstream

from Rio Sucio, alt. ca. 200 m, 23 May 1967,

Duke 11351 (ECON); *hills below first rapids on

Rio Truando, Duke 13331 (MO).

Brunfelsia chocoensis is restricted to the north-

ernmost part of the Choco region of Colombia in

the low-lying basin of the Rio Atrato, an area

bound by the Serrania de Baudo in the west and

the Serrania de Abibe in the east. Its range ex-

tends northward to Cerro Pirre across the Pana-

manian border. Although additional collecting

may reveal a larger area for the species, B. cho-

coensis appears to be endemic at this low eleva-

tion (50-200 m) in swampy forest. In 1945 it was

reported as a common species at Chigorodo, the

type locality. I failed to find the plant on a col-

lecting trip there, primarily because of the large-

scale destruction of the forests for agriculture and

pasturage.

Most closely related to B. guianensis of the

Guianas and northern Brazil, this species differs

in having much larger leaves with longer petioles,

longer pedicels, and broader corolla lobes. The

fruit of B. chocoensis is very similar to the fleshy

capsules of both B. guianensis and B. macrocar-

pa. Although B. chocoensis was originally

thought to be related to B. macrocarpa, it now

seems that the resemblance is only superficial.

The flowers of B. macrocarpa are very large and

purple, fading to white with age in the manner of

species of section Franciscea. In addition, the ca-

lyx of B. macrocarpa is accrescent and encloses

the mature fruit, whereas the calyx in B. chocoen-

sis is small and present only at the base of the

fruit.

Data from herbarium labels (Bristan 566 and

569) indicate that the fleshy, yellow fruits of this

species are edible and that the flowers are fra-

grant.

4. Brunfelsia clandestina Plowman, Fieldiana

Bot. n.s. 8: 11. 1981. Type: BRAZIL. Bahia:

Municipio Itapebi, Fazenda Lombardia, BR

101 ao lado leste. Arbusto de 3 m de altura.

Flor branca. Capoeira, 12 Aug. 1971, T. S.

dos Santos 1777 (holotype, CEPEC 7086; iso-

type, F 1849031). Figure 14.

Shrub or treelet to 6 m tall. Trunk to 8 cm in

diameter. Bark on trunk and branches yellowish

brown, cracking longitudinally and transversely,

shedding in thin, irregular flakes. It ranch lets 1 .5-

2 mm in diameter, glabrous, grayish to dark, red-

dish brown, more or less shiny, cracked longitu-

dinally. Leaves short petiolate, the blade 35-120

mm long, 15-50 mm wide, elliptic to oblong-lan-

ceolate, rarely lanceolate or obovate, apically

acute to acuminate, the apex itself obtuse, basally

acute or obtuse, glabrous or rarely sparsely pu-

berulent on the costa beneath, medium green

above, somewhat lighter green beneath, shiny on

both surfaces, chartaceous to subcoriaceous, rare-

ly coriaceous, the lateral nerves 6-8, mostly

straight, forming an angle of 45-70° with midrib,

anastomosing with the arcuate marginal nerve 3-

6 mm from margin; petiole 26 mm long, glabrous

PLOWMAN: REVISION OF BRUNFELSIA

43

BRUNFELSIA clandestiiia Plowman

FIELDIANA: BOTANY

or with scattered glandular hairs. Inflorescence

terminal or subterminal on mature or newly

formed twigs of current season, with or without a

short peduncle, 1- or 2-flowered. Bracts 0-3, 2-

6 mm long, linear or cymbiform, truncate at apex,

sparsely pubescent or glandular-pubescent, cadu-

cous. Flowers white. Pedicel 3-9 mm long, 1 mm

in diameter, glabrous, becoming thicker in fruit,

to 3 mm in diameter, warty-lenticellate. Calyx 8-

16 mm long, 3-7 mm in diameter, tubular or tu-

bular-campanulate, terete, glabrous, light green,

membranaceous, teeth 2-5 mm long, subequal,

triangular-ovate, apically acute or acuminate, the

apex itself blunt and minutely glandular-papillose;

calyx in fruit persistent, 11-16 mm long, shiny,

coriaceous, striately nerved, tightly enclosing ba-

sal half of capsule. Corolla 18-25 mm long, 1-3

mm in diameter, tube 1.5-2 times as long as ca-

lyx, straight, glabrous, rarely with a few glandular

hairs, orifice 3-5 mm across; limb 15-26 mm in

diameter, spreading, inclined, lobes 6-10 mm

long, subequal, expanding somewhat with age,

broadly obovate to rounded. Stamens inserted in

upper part of corolla tube; filaments 0.6-1 mm

wide, strap-shaped, the anterior pair 3-5 mm long,

incurved at apex, included, the posterior pair 1 .5-

3 mm long; anthers 1 mm in diameter, orbicular-

reniform. Ovary 1-1.5 mm long, 0.8-1.2 mm in

diameter, conical-ovoid, with about 10-12 ovules

per locule; style 16-20 mm long, incurved at

apex; stigma 1-1.5 mm long, briefly bifid, upper

lobe slightly larger. Capsule dry at maturity, 13-

20 mm long, 13-15 mm in diameter, ovoid to

subglobose, apiculate, smooth, shiny, dark green

at maturity; pericarp thin, 0.5-1 mm thick, crus-

taceous. Seeds ca 9-15, 5-7 mm long, 2.5-3 mm

in diameter, oblong-reniform, terete, or somewhat

flattened on one side, dark brown, reticulate-pit-

ted. Embryo 3-6 mm long, straight; cotyledons

1-2 mm long, ovate to elliptic, radicle 2-4 mm

long.

DISTRIBUTION — States of Bahia and Espirito

Santo, Brazil. See Figure 15.

SPECIMENS EXAMINED — BRAZIL. Bahia:

"Prov. Jacobina," 1841, Blanchet 3354 (G, LE, w),

1843, Blanchet s.n. (c, G, w); "Igreja Velha,"

1841, Blanchet 3352 (BM, BR, G, P); Serra de Sin-

cora, Brejao a Iracema, 17 Feb. 1943, Froes

20210 (IAN, NY, us); Municipio Jaguaquara, Ja-

guaquara a Apuerama, 4 Oct. 1972, Pinheiro 1980

(CEPEC, F, NY); Municipio Itacare, Itacare-Ubai-

taba, 14 Apr. 1980, T.S. dos Santos 689 (CEPEC,

F); Rodovia BA 654, km 6 ao oeste de Itacare,

approx. 14°18'S, 39°02'W, ca. 60 m, 12 Apr.

1980, Plowman, Mattos Silva & dos Santos 10066

(CEPEC, F), Plowman, Mattos Silva & dos Santos

10087 (CEPEC, F); Municipio Itambe, Itambe, 24

Nov. 1942, Froes 20067 (IAN, NY); Municipio Ita-

ja do Colonia, 12 km da estrada em dire9ao a

Feirinha ao lado oeste, margem esquerda do Rio

Corro, 23 Oct. 1969, T.S. dos Santos 433 (CEPEC,

F); Municipio Belmonte, Estaqao Experimental

Gregorio Bondar, km 58 da rodovia Belmonte/Ita-

pebi, 16 May 1979, Mattos Silva et al. 357 (CE-

PEC); Municipio Santa Cruz de Cabralia, Reserva

Biologica do Pau-Brasil, 18 Sep. 1971, T.S. dos

Santos 1964 (CEPEC, F); cerca de 16 km a oeste

de Porto Seguro, 21 Mar. 1978, Mori et al. 9775

(CEPEC, F); antiga rodovia que liga a Esta9&o Ecol-

ogica de Pau-Brasil a Santa Cruz, 5-7 km ao NE

da Esta?ao, ca. 12 km ao NW de Porto Seguro,

16°23'S, 39°8'W, ca 80-100 m, 5 Jul. 1979, Mori

et al. 12082 (CEPEC, F, us); Municipio Guaratinga,

rodovia Guaratinga/Sao Paulinho, km 25, 2 Apr.

1973, Pinheiro 2086 (CEPEC, F); without locality,

1857, Blanchet s.n. (G, L, LE); without locality or

date, Blanchet 1455 (F), Blanchet s.n. (F, MG, NY);

*Municipio de Cairu; Estrada Cairu-Itabera, 8 km

S of Cairu, Carvalho & Gatti 795 (F). Espirito

Santo: Linhares, Vale do Rio Doce, km 6 da ro-

dovia BR 101, lado sul, 30 Sep. 1971, T.S. dos

Santos 2015 (CEPEC, F, NY).

Brunfelsia clandestina was first collected by the

Swiss collector Blanchet in the state of Bahia

more than 100 years ago. Following earlier au-

thors, I originally assigned Blanchet's several col-

lections of the species to B. uniflora, although I

pointed out certain differences between typical B.

uniflora and the Blanchet material (Plowman,

1973). These early collections lacked adequate

field data and fruiting material [see Plowman,

1981— Eds.}.

Brunfelsia clandestina superficially resembles

several other brunfelsias and may be confused

with them. Complete specimens, including field

data and flower color, are essential for making

positive identifications. Brunfelsia clandestina ap-

pears to be most closely related to B. martiana, a

FIG. 14. Brunfelsia clandestina. Reproduced courtesy of the Field Museum of Natural History.

PLOWMAN: REVISION OF BRUNFELSIA

45

FIG. 15. Distribution of species in section Guianenses: Brunfelsia amazonica (open square), B. burchellii (open

triangle), B. chocoensis (open circle), B. clandestina (solid triangle), B. guianensis (solid square), and B. martiana

(solid circle).

species that also grows in the moist forests of Ba-

hia. Brunfelsia clandestina differs in having much

smaller leaves with fewer lateral nerves and one-

or two-flowered inflorescences that are borne ter-

minally. In dried specimens, B. clandestina may

also be confused with B. uniflora, a species

known from Bah fa but belonging to a different

section of the genus (section Francisced). Brun-

felsia clandestina differs mainly in having dark

brown or reddish brown branchlets, glabrous or

nearly glabrous leaves and twigs, and a tubular-

campanulate rather than narrowly tubular calyx.

The flowers of B. clandestina are white; those of

B. uniflora are violet, fading to white with age.

Brunfelsia clandestina grows in the moist

coastal forests of southern Bah fa and Espirito

Santo. Earlier collectors in Bahia also found this

species further inland in areas formerly covered

with mesophytic forest. Most of these areas have

now been converted to pastures (Mori & Mattos

Silva, 1979). The last collection in the drier, in-

terior part of Bahia was made in 1943.

5. Brunfelsia guianensis Benth. in DC., Prodr.,

10:200. 1846. Lectotype (designated here):

SURINAM. Without locality or date, Host-

mann 1278 (lectotype, K; isolectotypes, G, s,

w). Lectoparatype: FRENCH GUIANA. Cay-

enne, 1820, Perrottet s.n. (G, G-DC, non vidi;

photograph, NY). Figure 16.

Shrub to treelet 1-3 m tall. Branches lax,

spreading. Branchlets terete, becoming knobby at

nodes, glabrous, rarely pubescent, epidermis split-

46

FIELDIANA: BOTANY

BRUNFELSIA guianensis Benth

FIG. 16. Brunfelsia guianensis. Reproduced courtesy of the Linnean Society, London. (First published in Plowman,

1978: Botanical Journal of the Linnean Society, 76: 294-295.)

ting in longitudinal cracks, eventually shedding in glabrous, dull, dark green above, paler green be-

flakes, light to dark brown. Leaves with blade 6- neath, firmly membranaceous, lateral nerves 4-5,

15 cm long, 2-6.5 cm wide, obovate to elliptic, arcuate; petiole 3-8 mm long, slender, glabrous,

rarely lanceolate, abruptly acuminate at apex, acu- rarely pubescent. Inflorescence axillary or ter-

men often subfalcate, cuneate to narrowed at base, minal, with 1-2 flowers. Flowers white, odorless.

PLOWMAN: REVISION OF BRUNFELSIA

47

Bracts 1-2 per flower, 2-10 mm long, linear-lan-

ceolate or squamiform, glabrous, caducous. Ped-

icel 2-5 mm long, slender, short, glabrous. Calyx

7-10 mm long, 4-7 mm in diameter, ovoid-cam-

panulate, inflated, rarely tubular-campanulate,

ovoid in bud, glabrous, green, membranaceous,

sparsely veined, teeth usually unequal, two pos-

terior teeth somewhat shorter, 2-5 mm long,

ovate, acute to acuminate at apex; calyx in fruit

persistent, becoming somewhat coriaceous, split-

ting on 1 or more sides. Corolla tube 20-26 mm

long, 1-1.5 mm in diameter, 2.5-3 times as long

as calyx, slender, straight or curved, gradually di-

lated above middle, glabrous, greenish white to

white, limb 15-22 mm in diameter, spreading,

lobes 6-8 mm long, subequal, oblong-obovate to

spathulate, narrowed at base, rounded or blunt at

apex, lateral margins strongly reflexed, white, oc-

casionally with traces of purple. Stamens includ-

ed within tube; filaments slender, suberect, ligu-

late, anterior pair 4-5 mm long, posterior pair 2-

4 mm long; anthers 1 mm in diameter, orbicular-

reniform. Ovary 1.5-2 mm long, 1 mm in di-

ameter; ovoid-conical, style slender, broadened to-

ward apex; stigma included between anthers, brief-

ly bifid, upper lobe slightly larger. Fruit 2.5-3.2

cm in diameter, 2.5-4 cm long, globose to ovoid,

smooth, yellow when ripe, exocarp thin, rubbery,

mesocarp soft, fleshy, 5-8 mm thick, endocarp

thin, cartilaginous. Seeds 7-10, 10-13 mm long,

5-7 mm in diameter, oblong, flattened on 1 or 2

sides, angular, reddish brown to nearly black, re-

ticulate-pitted. Embryo 8 mm long, straight; cot-

yledons 2 mm long, elliptic.

DISTRIBUTION — Surinam, French Guiana, Bra-

zil. See Figure 15.

SPECIMENS EXAMINED — SURINAM. In monti-

bus Nassau, in forest near km 0.1, 17 Feb. 1949,

Lanjouw & Lindeman 2137 (K, NY); near km 6, 7,

10 Mar. 1949, Lanjouw & Lindeman 2564 (K, NY);

along creek with large rapids, 24 Mar. 1949, Lan-

jouw & Lindeman 2888 (K, NY); Saramacca River,

Jacob Kondre, 19 Jun. 1944, Maguire 23894 (NY);

Jodensavanne-Mapane creek area; Dist. Broko-

pondo, N of Brokopondo, 55°W, 5°N, 14 Feb.

1966, van Donselaar 3114 (NY); ad flum. Paloe-

meu et Tapanahoni confl., 3°20'N, 55°27'W, 10

Apr. 1963, Wessels Boer 1195 (NY, us); without

locality, Sep. 1853, Wullschlagel 1505 (w);

*Brownsberg Nature Park; 90 km S of Paramar-

ibo, Mori, Bolten 8394 (MO); *Lely Mountains;

175 kms SSE of Paramaribo, Mori, Bolten 8575

(MO); *near airport, road Oelemari, Boer 1047 (u);

*Mapanegebied, Elburg 9819 (u); *prope Joden-

savanne, Heyligers 438 (u); *Zuid River near

Kayser Airstrip, Irwin et al. 55978 (u); *Forest

Reserve Zanderij, Lanjouw 345 (GH); *Ribanau E

of Moengo, Lindeman 5889 (u), *Lindeman 6069

(u); *S of Moengo tapoe, Lindeman 6162 (u); *W

slopes of Bakhuis Mountains, Florschutz & Maas

2924 (GH, u), *Maas 3094 (GH, u); Blakawatra,

Schulz 8645 (u). FRENCH GUIANA. Mathoury,

Feb. 1901, Lemme s.n. (p); Karouany, 1855, Sagot

423 (s, w); without locality, 1821, Perrottet s.n.

(G), 1821, Poiteau s.n. (G); *above Taruni Creek,

Forestry Department of British Guiana 7495 (NY);

*Tumuc-Humac, Frontiere Bresil Guyane, Gran-

ville 1378 (CAY); *N de Degrad Claude, Granville

2204 (GH); *Chemin de Smerillons, a 5 km de

Degrad Claude, Granville 2224 (CAY); *Basse

Ouaqui, au niveau de "Degrad Roche," Granville

B-4854 (CAY); *Saul, Monts La Fumee, Mori &

Boom 14726 (F), *Mori & Boom 14834 (F), *Mori

et al. 8767 (MO); *Haute Approuague, au conflu-

ent de la crique Parepou, Oldeman B-1896 (CAY);

*region of Tumuc-Humac, Massif du Mitaraka,

Sastre 1649 (CAY, p); *Haut Oyapock, Peola 20

km, en aval de Trois Sauts, Sastre 4751 (P). BRA-

ZIL. Amapa: Rio Oiapoque, islands of first cach-

oeira on Rio laue, 2°53'N, 52°22'W, 0.5 km E of

confluence with Rio Oiapoque, 28 Aug. 1960, Ir-

win, Pires & Westra 47900 (MG). Para: Parque

Indigena do Tumucumaque, Rio Paru de Oeste,

Missao Tiriyo, Cavalcante 2518 (F); Belem, in

sylvis ad Para, May 1819, Martius s.n. (M); vicin-

ity of Belem, Sep.-Oct. 1961, Pires 51899 (NY,

us); estrada do Cafezaldo IAN, 28 Oct. 1949, N.T.

da Silva 336 (GH, NY, p); Almeirim, Rio Arrayo-

los, 26 Apr. 1903, Ducke s.n. (MG, INPA); Orixi-

mina, 8 Dec. 1906, Ducke s.n. (MG); Rio Cumi-

namirim, 16 Dec. 1906, Ducke s.n. (MG); Obidos,

10 Jan. 1920, Ducke s.n. (RB); Rio Branco, 21

Mar. 1924, J.G. Kuhlmann 1722 (RB); Cachoeira

do Rio Capim, 29 Jun. 1897, Huber 887 (MG);

*Belem, IPEAN, Reserva Mocambo, Pires & Sil-

va 10218 (GH, IAN), * Pires & Silva 11436 (GH);

*Belem, IPEAN, Schubert 2150 (us); *Rio Trom-

betas, planalto Saraca, N.T. da Silva & Santos

4631 (F).

Brunfelsia guianensis was for a long time con-

sidered an anomalous species among the South

American brunfelsias. With the recent discovery

of several related species, it is now known to be-

long to a distinct section of the genus, of which

it is the type species (Plowman, 1978). Brunfelsia

guianensis differs from other species in this group

48

FIELDIANA: BOTANY

in having obovate leaves and a somewhat inflated,

ovoid-campanulate calyx. The fruit is large and

fleshy and yellow at maturity.

A lowland tropical forest species, B. guianen-

sisis is distributed throughout the lower Amazon

basin, chiefly north of the river, reaching well into

the Guianas. It is known to flower from October

to April, and fruits have been collected throughout

the year. In the Brazilian state of Para, this species

is used in local medicine under the name manaca

with properties similar to those of B. uniflora (see

LeCointe, 1947; Plowman, 1977).

6. Brunfelsia martiana Plowman, Bot. Mus.

Leafl. 24(20): 37. 1974. Type: BRAZIL.

Provinciae Rio Negro (Amazonas): habitat in

sylvis ad flum. Japura prope Sao Joao do

Principe, Dec. 1819, Martins 3247 (holotype,

M; isotypes, M). Figure 17.

Shrub to 1 m tall. Branches few, naked below,

somewhat knobby at nodes, glabrous. Bark

cracked longitudinally, dark reddish brown, shiny,

furnished with lenticels. Leaves scattered along

branchlets, subsessile, blade 10-25 cm long, 4-8

cm wide, mostly oblong, sometimes elliptic-ob-

long or oblong-obovate, acuminate at apex, broad-

ly cuneate to blunt at base, glabrous on both sides,

upper surface dark green, nitid or dull, lower sur-

face paler green, sometimes nitid, firmly membra-

naceous to subcoriaceous, midrib dark reddish

brown, lateral nerves 8-12, spreading, straight;

petiole very short, 1-4(12) mm long, glabrous,

dark brown, becoming cracked, rugose. Inflores-

cence terminal or axillary in the upper leaf axils,

glabrous, axis 2-10(15) mm long. Flowers 1-7

per inflorescence, sometimes with short peduncle

that articulates with pedicel, greenish white to

white. Bracts 1-3 per flower, 1-10 mm long, lin-

ear-lanceolate, concave, ciliolate at margin, ca-

ducous. Pedicel 3-6(10) mm long, short, slender,

glabrous. Calyx 8-12 mm long, 3-8 mm in di-

ameter, tubular to tubular-campanulate, glabrous,

rarely striately veined, teeth 2-6 mm long, sub-

equal, triangular-ovate to ovate-lanceolate, acute

to acuminate at apex. Corolla tube 20-24 mm

long, 1-3 mm in diameter, twice as long as calyx,

straight, cylindric, inflated at apex, glabrous; limb

15-22 mm across, spreading, somewhat undulate,

lobes 5-10 mm long, subequal, oblong-obovate,

abruptly reflexed at lateral margins, rounded-sub-

truncate at apex. Stamens included within upper

part of corolla tube; filaments subligulate, anterior

pair 4 mm long, posterior pair 3 mm long; anthers

about 1 mm in diameter, orbiculate-reniform,

slightly unequal, the upper pair somewhat smaller.

Ovary 2 mm long, oblong-ovoid; style about 15

mm long, filamentous, a little broader and curved

at apex; stigma about 1 mm long, briefly bifid,

upper lobe slightly larger. Fruit and seed incom-

pletely known.

DISTRIBUTION — Guyana, Brazil (Amazonas, Ba-

hia, Maranhao, Para). See Figure 15.

SPECIMENS EXAMINED — GUYANA. Essequibo-

Demerara River, Madray-Bubu Trail, Topy Trysil

Forest, 8 Feb. 1944, Forest Department 4422 (K).

BRAZIL. Amazonas: Manaus and vicinity, road

Manaus-Caracarai, km 22, 21 Mar. 1967, Prance

et al. 4704 (wis); Rio Negro, Cachoeira Baixa de

Tarumo, 11-14 Apr. 1973, Schultes & Rodriguez

26132A (ECON), Schultes & Rodrigues 26133A

(ECON); *Mun. de Barcelos; Rio Demeni, Froes

28393 (IAN); *vicinity of Manaus, Prance et al.

11611 (MO). Bahia: Dist. Ilheus, Ferradas, Dec.

1818, Martius s.n. (M); Feira de Santana, Apr.

1850, collector unknown (G). Maranhao: Ilha

Sao Luis; estrada que vai do Rio Anil para Maio-

ba, Froes 25650 (IAN); Mun. Sta. Luzia; Fazenda

Cacique, W of Sta. Inez and E of Entroncameen-

to, Taylor et al. 1084 (F). Para: Faro, 22 Jan.

1910, Ducke s.n. (MG); Bella Vista, Rio Tapajoz,

matta das immediacoes da Campina do Perdido,

12 Feb. 1917, Ducke s.n. (MG); Regiao do Jutah

y de Almerim. Palhal, 16 Apr. 1923, Ducke s.n.

(RB); Gurupa, 25 Feb. 1923, Ducke s.n. (RB);

Para, May 1819, Martius 3300 (M). Without

state: Schott 5361? (w), Schott s.n. (w).

Brunfelsia martiana belongs to section Gui-

anenses. The type specimen collected by Martius

(Schedulae 3247) on the Rio Japura includes a

handwritten description of the plant that Martius

apparently intended to publish as B. acuminata, a

name that was later used by Pohl and Bentham

for a form of B. brasiliensis.

Brunfelsia martiana has been very poorly col-

lected, as is true of many of the species of section

Guianenses. The few collections available were

collected over an extensive geographical area

spanning the entire breadth of Brazil, and nearly

all are preserved in European and Brazilian her-

baria. The species has been collected only twice

in the last 50 years; no collections have been

made in the state of Bahia since 1850. One ap-

parently quite early collection with six duplicates

is preserved in Vienna. The label bears only the

data "Schott 5361, Brasilia." This appears to be

PLOWMAN: REVISION OF BRUNFELSIA

49

BRUNFELSIA martiana Plowman

FIG. 17. Brunfelsia martiana. Reproduced courtesy of the Botanical Museum of Harvard University.

50

FIELDIANA: BOTANY

the wrong label for the specimen because Schott

collected only in the state of Rio de Janeiro and

no other collection of the species has ever been

made there. Mislabeling of the collections of

Schott and Pohl has been recognized since the

time of Martius, who made a special note of it in

Flora Brasiliensis [1(1): 82 (1906)]: "Dolendum

quod, quae specimina a cl. POHL, quae a cl.

SCHOTT collecta sint, ex plagulis herbarii pala-

tini Vindobonensis non semper statui potest."

The origin of this collection remains question-

able, yet it must be taken into consideration be-

cause it represents one of the more complete gath-

erings of the species and includes an immature

fruit.

Brunfelsia martiana is a small shrub with

greenish white flowers. It is related to both B.

amazonica and to B. guianensis and occurs sym-

patrically with both of these species. Brunfelsia

martiana differs from B. amazonica in having

larger leaves (10-25 vs. 6-12 cm) with more lat-

eral nerves (8-12 vs. 5-8) and much shorter ped-

icels (3-6 vs. 13-20 mm).

Brunfelsia martiana is closely allied with B.

guianensis as well, and specimens of the former

were originally considered to be exceptional

forms of B. guianensis. Additional material

proved the distinctiveness of B. martiana. It is

now distinguished from B. guianensis by having

larger (10-25 vs. 6-15 cm), oblong leaves with

more lateral nerves (8-12 vs. 4-5), a calyx that

is usually narrow-tubular instead of ovoid-cam-

panulate, and a corolla tube that is rarely more

than twice as long as the calyx. In B. guianensis

it is frequently 2%-3 times as long.

The flowers of all three of these related species

are remarkably similar in the shape and color of

the corolla, especially in the curious form of the

corolla lobes, the lateral margins of which are de-

flexed, giving them a rectangular appearance. The

fruit of B. guianensis is a thick-walled, fleshy cap-

sule; that of B. amazonica is thin-walled and dry

at maturity. The collection labeled Schott 5361

mentioned above includes a small, immature fruit

enclosed by a coriaceous and striately veined ca-

lyx. The pericarp appears cartilaginous but not

fleshy in this early stage.

A second collection, which for B. martiana has

exceptionally small leaves reminiscent of B. gui-

anensis, was made by Martius near Belem do

Para. This specimen included the remnants of a

fruit that is small, definitely thin-walled, and brit-

tle, perhaps the remains of the endocarp only.

Martius tentatively named this plant B. flexuosa

and wrote a description (Schedulae 3300) that ap-

pears on the same page attached to the type spec-

imen of B. martiana. The nature of the fruit in

this species remains problematical. When, indeed,

it is fully known, it should be very useful in de-

termining the relationship of B. martiana to other

species.

Brunfelsia martiana is one of the most widely

distributed species of section Guianenses, grow-

ing in tropical lowland forests from Guyana south

to Bahia in Brazil and westward to the Colombian

frontier. The curious distribution between the Am-

azon basin and the coastal forests of Bahia has

been observed in other groups of plants (Plow-

man, 1979).

Brunfelsia sect. Franciscea (Pohl) Griseb., Fl.

Brit. W. I.: 432. 1861. Franciscea Pohl, PI.

Bras. Icon. Descr. 1: 1, t. 1-7. 1826. Lecto-

type species (designated here): Franciscea

uniflora Pohl.

TYPE SPECIES — Brunfelsia uniflora (Pohl) D.

Don, Edinburgh New Philos. J. 86. Jul. 1829.

Shrubs or small trees, 0.3-10 m tall. Inflores-

cence terminal, rarely axillary. Flowers large or

small, 1-many per inflorescence. Corolla purple

fading to pale violet or white with age, with white

eye at mouth, tube 1-4 times as long as calyx,

slightly inflated near apex, then slightly constrict-

ed to form thickened ring, the lobes broadly

rounded, overlapping at lateral margins. Stigma

briefly bifid, in the form of a forceps. Anthers

orbicular-reniform in outline. South and Central

America.

Note: Pohl, in establishing his new genus Fran-

ciscea, included seven species. In recognizing this

name as a section of Brunfelsia, it is necessary to

designate a type species. I have chosen B. uniflora

(Pohl) D. Don for the following reasons. It is the

oldest known species of the genus, having ap-

peared in 1 648 in the work of Piso under the name

Manacd, accompanied by a good illustration

showing the flowers and fruits (Fig. 18). The ho-

lotype collected by Pohl and preserved in Vienna

is a reasonably good specimen showing the char-

acters of the species. The species is one of the

most widespread of the genus and is well repre-

sented in herbaria. Lastly, B. uniflora (as B. ho-

peand) was the only species mentioned by Gri-

sebach (1861) in setting up Franciscea as a sec-

tion of Brunfelsia.

PLOWMAN: REVISION OF BRUNFELSIA

51

FIG. 18. Manaca (Brunfelsia uniflora). From Piso, W., De Medicina Brasiliensis (1648).

52

FIELDIANA: BOTANY

Although the date 1827 appears on the title

page of Pohl's work Plantarum Brasiliae Icones

et Descriptiones, it has been pointed out (Stafleu,

1967; Stafleu & Cowan, 1983) that the first part

of the work, including the text and plates of Fran-

ciscea, actually appeared in August 1826.

7. Brunfelsia australis Benth. in DC., Prodr. 10:

200. 1846. Lectotype (designated by Hunt,

1978): BRAZIL. Brasilia meridionalis, Rio

Uruguay, Baird s.n. (K). Lectoparatypes: AR-

GENTINA. Buenos Aires, in hortis Bonarien-

sibus sub nomine Jasmini Paraguayensis cul-

ta, Tweedie s.n. (K); Estrada de Minas, Mar-

tius s.n. (?). No collection of Martius with

this data has been located. Query was made

by Bentham. Sellow 1573, also cited by Ben-

tham, is excluded as a type; it belongs to B.

pilosa Plowman. Figure 19.

Franciscea australis (Benth.) Miers, Ann. Mag. Nat.

Hist. ser. 2, 5: 250. 1850.

Brunfelsia hopeana var. australis (Benth.) J.A.

Schmidt in Mart., Fl. Bras. 8(1): 262. 1864.

Brunfelsia paraguayensis Chodat, Bull. Herb. Bois-

sier, ser. 2, 1: 406. 1901. Type: PARAGUAY. In

silvis pr. Rio Apa, May, without year, Hassler

725a (holotype, G).

Brunfelsia uniflora f. typica Hassl., Feddes Repert.

Spec. Nov. Regni Veg. 15: 243. 1919. Non sensu

D. Don. Type: PARAGUAY. In silvis pr. Rio

Apa, May, without year, Hassler 725a (holotype,

G).

Brunfelsia uniflora f. obovatifolia Hassl., Feddes Re-

pert. Spec. Nov. Regni Veg. 15:243. 1919. Type:

PARAGUAY. Cordillera de Altos, in silvis, fru-

tex vel arbor parva 3-5 m, corolla albo-coeru-

lescens vel violacea in eadem planta, Sep. 1898-

1 899, Hassler 3246 (holotype, o; isotypes, A, G,

NY, P, w).

Brunfelsia uniflora f. intermedia Hassl., Feddes Re-

pert. Spec. Nov. Regni Veg. 15: 243. 1919. Lec-

totype: PARAGUAY. In regione lacus Ypacaray,

Cordillera do Altos, arbusto 2-2.5 m, fl. lila-

morado, monies y orillas, Cordilleras y bajos,

azucena, Sep. 1913, Hassler 12257 (lectotype,

G; isolectotypes, c, GH, K, L, MO, NY). Lectopar-

atypes: PARAGUAY. In regione lacus Ypacaray.

Picada a Bernalcue, Mar. 1913, Hassler 12147

(A, G); San Bernardino, orillas monies, fl. morado

despues blanca, arbusto 3-4 m, Aug. 1915,

Hassler (cited as Rojas) 1333 (G).

Shrub to small tree 2-4 m tall. Trunk often

branched from near base, to 8 cm in diameter.

Bark rugose, yellowish brown. Branches ascend-

ing and spreading, leafy, glabrous, green with yel-

lowish brown cracks, furnished with lenticels.

Branchlets glabrous, green. Leaves scattered

along branchlets, blade 4.0-12.5 cm long, 2.5-6.0

cm wide, broadly elliptic to obovate, rarely orbic-

ular or rhombic, blunt to rounded at apex, occa-

sionally acute, cuneate at base, often briefly pu-

bescent-ciliate when young, becoming glabrous,

or sparsely glandular beneath at costa, dull dark

green above, pale green beneath, membranaceous

to subcoriaceous, lateral nerves 5-7, prominulous

beneath, straight or somewhat arching, veinlets

often obscure; petiole 3-8 mm long, glabrous or

sparsely pubescent. Inflorescence terminal at the

tips of the branchlets, sessile, with 1-4 flowers.

Bracts present or absent, to 2 mm long, lanceo-

late, concave, ciliolate, early caducous. Flowers

showy, violet, fading to pure white with age, fra-

grant. Pedicel stout, 4-7 mm long, glabrous, be-

coming thicker and corky-verrucose in fruit. Ca-

lyx 7-12 mm long, 4-10 mm diameter, campan-

ulate, more or less inflated, glabrous, rarely pa-

pillose-punctate, pale green, firmly membrana-

ceous, reticulately veined, teeth 3-6 mm long,

ovate-lanceolate, acute; calyx in fruit coriaceous,

partially enclosing fruit, smooth. Corolla tube

25-32 mm long, 1.5-2.0 mm in diameter, 2-3

times as long as calyx, glabrous, often sparsely

glandular in bud, glabrescent; limb 30-40 mm in

diameter, plane, spreading, thickening at mouth

prominent, round, white or yellowish, lobes equal,

broadly rounded, overlapping laterally, abruptly

narrowed at base. Stamens included; filaments

slender, upper pair 2-3 mm long, lower pair 3-5

mm long; anthers about 1 mm in diameter, reni-

form-orbicular. Ovary 2.5 mm long, 1 mm in di-

ameter, ovoid-conical; style 20-28 mm long; stig-

ma about 1 mm long. Capsule 10-22 mm long,

10-20 mm in diameter, subglobose to ovoid, apic-

ulate at apex, smooth, glabrous, dark green,

pericarp to 0.5 mm thick, dry at maturity, inde-

hiscent or tardily dehiscent. Seeds 20-40, 4-6

mm long, 3-4 mm in diameter, oblong-ellipsoid,

angular, dark brown. Embryo 3-4 mm long,

slightly curved; cotyledons ovate.

DISTRIBUTION — Brazil (Parana, Rio Grande do

Sul, Sao Paulo), Paraguay, Argentina (Chaco,

Corrientes, Formosa, Misiones, Santa Fe), and

Uruguay; cultivated in some areas, especially in

Central America. See Figure 20.

SPECIMENS EXAMINED — BRAZIL. Parana:

*Municipio Foz do Igua9u; Vila Postes, Buttura

157 (F); *Ceu Azul, Hatschbach 43192 (F). Rio

Grande do Sul: Neu Wurttemberg, 450 m, 15

Oct. 1904, Bornmuller 219 (A, G, GH, M, w); Sao

Leopoldo, 16 Oct. 1934, Rambo 321 (C); Porto

Alegre, 12 Sep. 1945, Rambo 37819 (A, c, w);

PLOWMAN: REVISION OF BRUNFELSIA

53

Morra da Gloria, Rambo 29147 (BR, MO). *Mun-

icipio Itaqui, Dobereiner & Tokarnia 841 (RB).

Sao Paulo: *Parque Estadual do Turvo, Tenente

Portela, Paiva, Stehmann s.n. (F). PARAGUAY.

Cordillera de Altos: Cerros de Tobati, 18 Aug.

1902, Fiebrig 49 (A, F, G, L, M, PR); Hassler 725

(G); Sep. 1913, Hassler 12559 (A); Villarrica, Sep.

1931, Jorgensen 3661 (A, c, F, MA, MO, NY, PH, s,

us); Dep. Cantera, Zona Grande Bosques, Herrera

Vegas, 220 m, 11 Aug. 1939, Monies 4703 (BAB);

Dep. San Pedro, Colonia Primavera, 11 Sep.

1955, Woolston 571 (c, GH, NY, s, uc); without

locality, 26 Aug. 1874, Balansa 2240 (G, GOET, K,

LD, LE, PE, s); 4 Sep. 1854, Palmer 155 (us); *Par-

aguari, Cerro San Jose, a civitate Ybicui, Bernardi

18138 (F); *Neembucu, Curupaity ad Laureles,

Bernardi 20495 (F). ARGENTINA. Chaco: Dep.

Resistencia, Margarita Belen, 8 Sep. 1946, Agui-

lar 841 (s), 18 Sep. 1948, Aguilar 1302 (w); Dep.

Primer de Mayo, Colonia Benitez, Oct.-Nov.

1928, Schulz 85 (BAB), 26 Nov. 1906, Stuckert

16307 (G); Las Palmas, Jorgensen 2098 (GH, MO,

uc); Rio de Oro, km 140 de Resistencia, 29 Aug.

1944, Rojas 12133 (uc). Corrientes: Capital, Feb.

1821, Bonpland 528 (P); 1908, Llamas s.n. (BAB);

Quinta La Eloisa, 77 m, 6 May 1969, Plowman

2713 (ECON, GH). Dep. Conception, Tabay, 1 Nov.

1965, Krapovickas & Cristobal 11706 (uc); 7

June 1969, Krapovickas & Cristobal 15385 (c,

MO); 7 May 1969, Plowman 2723 (ECON, GH);

Dep. Mburucuya: Estancia "Santa Teresa," 6 Sep.

1949, Pedersen 418 (BR, c, GH, p, s, us); Estancia

"Santa Maria," Campo "Dios Gracia," 18 Dec.

1949, Pedersen 418 (BR, c, NY, p, s); Estancia

Santa Ana, 21 Oct. 1944, Schwarz 93 (A); Canada

Paso Clarios, 11 Nov. 1949, Schwarz 8661 (BH,

G, w). Dep. Bella Vista, Bella Vista, Oct. 1904,

N. Rojas Acosta 13004 (BAB); Dep. Saladas: La-

guna Salada, 28 Sep. 1944, Schwarz 36 (A); La-

guna Soto, 28 Sep. 1944, Schwarz 48 (A); Paso

del Deseo, 1 Nov. 1944, Schwarz 139 (A, uc); 2

Dec. 1949, Schwarz 8972 (BH, G, w). Dep. Itu-

zaingo, Salto Apipe, 6 Oct. 1949, Schwarz 8114

(BR); Dep. General Paz, Laguna Rincon, 2 Nov.

1949, Schwarz 8547 (LD); Dep. San Roque, San

Roque cercanias, 20 Apr. 1945, Ybarrola 2963

(uc, w); Dep. San Luiz del Palmar, Colonia J. M.

Llano, 19 Sep. 1945, Ybarrola 3270 (B, w); Cer-

rudo Cue, 5 Sep. 1946, Ybarrola 3430 (G, p); vi-

cinity of Goya, 15-30 m, 2-10 Nov. 1913, Curran

s.n. (us); *Dep. Santo Tome, Garruchos, Krapov-

ickas et al. 26092 (MO); *Dep. San Martin, Carlos

Pellegrini, Krapovickas 29261 (MO); *Dep. San

Martin, Sussini, Schinini et al. 18580 (F). For-

mosa: Dep. Pilcomayo, Estancia Riacho Negro, 3

Oct. 1947, Morel 3783 (BR, MO); Salvation, 17

Nov. 1947, Morel 4144 (c); S.O. a 12 km de Fi-

lipina, Est. Salaberry, 21 Nov. 1949, Morel 8921

(BH, c, G, w); Formosa, 29 Oct. 1900, Kermes

1363 (BAB); Fontana, Aug. 1933, Meyer 166 (GH).

Misiones: Dep. Iguazii, Salto Iguazu, 2 Oct. 1947,

Pierroti 6563 (G); 6 Oct. 1910, Rodriguez 484 (A,

us); Dep. Candelaria, Martires, 24 Oct. 1917, Ber-

toni 3434 (w); Loreto, 26 Sep. 1932, Muniez 99

(BAB); Dep. San Javier, Acaragua, 16 Sep. 1946,

Bertoni 2939 (B); Dep. G. M. Belgraw, "El Ca-

burei," 450 m, 22 Sep. 1957, Monies 27471 (M,

NY); without locality, Fox s.n. (CGE). Santa Fe:

Dep. General Obligado, Florencia, 30 Oct. 1923,

Chiar s.n. (BAB); Villa Guillermina, 14 Feb. 1939,

Ragonese 3646, (F); Indio Muerto, 6 Mar. 1939,

Meyer 3076 (A). URUGUAY. Banda Oriental,

1816-1821, St. Hilaire C-2197 (p).

CULTIVATED SPECIMENS OF INTEREST — BRA-

ZIL. Rio Grande do Sul: Porto Alegre, 20 Mar.

1893, C. Lindman A597a (s); Santa Maria, Sil-

vacultura, 14 Oct. 1955, Michel 8 (B). Rio de Ja-

neiro: Rio de Janeiro, Quinta de Boa Vista, 13

Oct. 1879, Glaziou 12110 (BR, c, G, LE, p, R); Rio

de Janeiro, Oct. 1915, Vincent s.n. (L). Santa Ca-

tarina: Brusque, 4 Oct. 1961, Reitz & Klein

11250 (us). ARGENTINA. Tucuman: Capital,

Institute Lillo, 9 Sep. 1947, Meyer 12621 (B);

Dep. Trancaas, San Pedro de Colalao, 1,200 m,

Apr. 1926, Venturi 4408 (A).

Bentham described B. australis from three

specimens, two of which are included in the pres-

ent concept. Baird s.n. was chosen (by Hunt,

1978) as the lectotype. One of the other speci-

mens, collected by Tweedie, was made from hor-

ticultural material in Buenos Aires. The third col-

lection cited by Bentham, Sellow 1573, is here

referred to B. pilosa, a closely allied species. A

fourth collection by Martius (with a query by

Bentham) cannot be located. The locality "Estra-

da de Minas" indicates another species was in-

volved because B. australis does not occur north

of Rio Grande do Sul in Brazil.

FIG. 19. Brunfelsia australis. Reproduced courtesy of Curtis 's Botanical Magazine and Christabel King.

54

FIELDIANA: BOTANY

Brunfelsia australis

CHRISTABEL KING del.

PLOWMAN: REVISION OF BRUNFELSIA

55

FIG. 20.

circle).

Distribution of Brunfelsia australis (solid triangle), B. boliviano (solid square), and B. bonodora (solid

Schmidt in Martius (1864) considered B. aus-

tralis to be a variety of B. uniflora (B. hopeana}.

He cited the Sellow collection mentioned above

and two additional specimens. One of these, Rie-

del s.n., collected near Macae, Rio de Janeiro,

agrees with B. bonodora. The other collection,

Martius s.n., from Para, is referred to B. guianen-

sis, although most of this material is sterile.

Brunfelsia australis is closely related to B. un-

iflora and B. pilosa and by some authors (Schmidt

1864; Hassler, 1919) has been considered synon-

ymous with B. uniflora. However, these workers

suffered from the disadvantage of working exclu-

sively with dried herbarium specimens (Schmidt)

or with local populations without understanding

the entire morphological range of the species in-

volved (Hassler). Studies of both B. australis and

B. pilosa in the field and of all three entities in

cultivation and from herbarium specimens reveal

that these plants are morphologically, ecological-

ly, and geographically distinct and merit the rank

of species. Experimental crosses suggest that re-

productive isolation exists between B. pilosa and

B. australis.

Brunfelsia australis is distinguished from B.

uniflora and B. pilosa by the broadly elliptic to

obovate leaves, the variable number (one to four)

of flowers per inflorescence and the uniformly

glabrous, campanulate calyx.

Even though their ranges overlap, almost no in-

termediates have been found between B. australis

and B. pilosa. One collection from Rio Grande do

Sul, Bornmuller 219, has oblong leaves and a

more deeply incised calyx and may represent

56

FIELDIANA: BOTANY

some introgression from B. pilosa, which also oc-

curs in this region.

Commonly growing as a tall shrub or treelet,

B. australis is found primarily in the relatively

recent drainage systems of the Rio Parana, Rio

Uruguay, and Rio Jacui (Rio Grande do Sul),

growing in the understory of low-elevation forests

near the rivers and in adjacent uplands. It occurs

from sea level to 450 m altitude. In the low-lying,

often swampy river basin of the Rio Parana, B.

australis is found in the scattered woodlands that

appear on higher and drier ground.

Two reports of widely disjunct populations are

now known to represent cultivated plants. Venturi

4408 from Tucuman, Argentina, was distributed

to some herbaria without the notation "culta."

Specimens of Glaziou 12210 were distributed

with the fictitious provenance "Rio de Janeiro,

Serra da Palmital." However, a duplicate speci-

men at the Museu Nacional de Rio de Janeiro

bears the correct locality, "Quinta da Boa Vista

(Rio), Arbusto cult." This erroneous data may be

the result of direct falsification by Glaziou, which

was pointed out in the Melastomataceae by Wur-

dack (1970).

Brunfelsia australis flowers from August to

November; mature fruits appear from April to

June. Frequently the entire plant is covered with

fragrant, showy purple and white flowers at the

same time, making this species a valuable orna-

mental. It is widely cultivated in gardens between

latitudes 35°N and 35°S (San Francisco, Califor-

nia, to Buenos Aires, Argentina).

8. Brunfelsia boliviano Plowman, Fieldiana, Bot.

n.s. 8:1. 1981. Type: BOLIVIA. Santa Cruz:

Prov. de la Cordillera, region of Lagunillas,

Cordillera of Incahuasi, 900 m, "at dry and

sandy slopes; shrub 1-2 m; flowers white or

light violet; when cattle eat the leaves of this

plant [they] die." N. v. bella union. Aug.

1934, Cardenas 2813 (holotype, F 756420).

Figure 21.

Shrub 1-2 m tall (fide Cardenas). Bark on

branches thin, yellowish to grayish brown, lon-

gitudinally rugose, not exfoliating. Branchlets 2-

3 mm in diameter, more or less villous, becoming

glabrous. Leaves scattered along branchlets or

crowded at tips of lateral short shoots, 40-135

mm long, 23-58 mm wide, blade obovate, rarely

elliptic-obovate, apically rounded with a short

acumen 5-10 mm long, the acumen itself blunt to

acute, somewhat revolute at margin, basally atten-

uate, glabrescent on both surfaces except at the

midrib, which bears villous and glandular hairs,

ciliate-villous on the adaxial surface of the revo-

lute margin, dull green above, paler, yellowish

green beneath, firmly membranaceous or some-

what coriaceous, the lateral nerves 4-9, straight,

arcuately anastomosing toward the margin, prom-

inulous beneath; petiole short, 2-6 mm long, more

or less villous, more densely so on upper side.

Inflorescence corymbiform, terminal on last sea-

son's branchlets, short pedunculate, branched,

with 6-15 flowers. Peduncle 5-12 mm long, more

or less villous. Bracts small, 14 mm long, lan-

ceolate or cymbiform, sparsely to densely villous

especially at margin, caducous. Flowers light vi-

olet fading to white (fide Cardenas). Pedicel short,

2-6 mm long, 1 mm in diameter, scarcely thick-

ened at apex, glabrous or with sparse glandular

hairs. Calyx 9-15 mm long, 3-4 mm in diameter,

tubular, truncate at base, 5-angled in cross section,

appearing plicate in pressed specimens, glabrous

or with scattered glandular hairs, teeth 2-4 mm

long, subequal, ovate, apically short acuminate,

the acumen itself blunt and glandular-papillose;

fruiting calyx persistent, withering. Corolla tube

21-25 mm long, 2-3 mm in diameter, about twice

as long as calyx, straight; limb 18-30 mm in di-

ameter, spreading, lobes 8-14 mm long, subequal,

rounded. Stamens inserted in upper part of co-

rolla tube; filaments strap-shaped, the anterior pair

3 mm long, suberect, briefly exserted from mouth

of corolla tube, the posterior pair 2 mm long, in-

cluded; anthers 1-1.2 mm in diameter, orbicular-

reniform, subequal or the upper pair slightly larg-

er. Ovary 2-3 mm long, 1.8-2.2 mm in diameter,

narrowly ovoid, with about 30 ovules; style 19

mm long, incurved at apex; stigma 2 mm long,

briefly bifid. Capsules 3-6 per infructescence,

20-25 mm long, 20-25 mm in diameter, dry at

maturity, subglobose, smooth, green; pericarp

thin, 1-1.5 mm thick, endocarp thin, crustaceous,

with 22-28 seeds per capsule. Seeds 5-8 mm

long, 2-3 mm in diameter, oblong to subreniform,

subterete, dark brown, reticulate-pitted. Embryo

4 mm long, straight; cotyledons 1.5 mm long,

ovate; radicle 2.5 mm long.

DISTRIBUTION — The eastern Andes of southern

Bolivia. See Figure 20.

SPECIMENS EXAMINED — BOLIVIA. Chuquisa-

PLOWMAN: REVISION OF BRUNFELSIA

57

BRUNFELSIA boliviana Plowman

FIG. 21. Brunfelsia boliviana. Reproduced courtesy of the Field Museum of Natural History.

58

FIELDIANA: BOTANY

ca: mountain above Bartolo, on road from Mon-

teagudo to Sucre, 20°00'S, 64°45'W, 5000 ft, 20

Sep. 1949, Brooke 5653 (BM). Santa Cruz: Prov.

Cordillera, Nov.-Dec. 1845, Weddell 3621 (p, 2

sheets); 5 km N of Yatarenda, 63°32'W, 19°12'S,

17 Apr. 1977, Krapovickas & Schinini 31476

(CTES, F); *Prov. Luis Calvo: Monteagudo 26 km

hacia Camiri, Beck & Liberman 9379 (F).

Brunfelsia boliviano is known from only five col-

lection^,', all from a relatively small area in the

foothills/ of the Andes in southern Bolivia. The

early collection of Weddell in 1845 is labeled

merely "Provincia de la Cordillera," referring to

the large province in the department of Santa

Cruz. However, Urban (1906) mentions that in the

months given, Weddell traveled and collected

from Santa Cruz da la Sierra in Santa Cruz south

to Sauces (now Monteagudo) in the department of

Chuquisaca. His collection was probably made

near the border between the two departments.

Brunfelsia boliviana is most closely related to

B. cuneifolia, which occurs in southern Brazil

from the state of Parana south to Rio Grande do

Sul. Like the present species, it is known from

only a few specimens. Brunfelsia boliviana differs

in both leaf and inflorescence characters as sum-

marized in Plowman (1981).

The geographical distribution of these two

closely related species deserves some comment

because it exemplifies further a pattern observed

in other brunfelsias in which vicarious species

pairs are found in southeastern Brazil and the

eastern Bolivian Andes. These include B. bono-

dora-B. grandiflora subsp. schultesii and B. hy-

drangeiformis-B. mire. In addition, disjunct and

somewhat distinct populations of B. uniflora, a

species primarily of eastern Brazil, have also been

found in the Bolivian Andes. In each of these

cases, vicarious species or populations are now

separated by more than 2,000 km in which no

intervening populations are known to occur. This

phytogeographical pattern has been discussed by

Smith (1962) and by Plowman (1979).

Interestingly, the leaves of B. boliviana are re-

puted to be lethal to cattle, a feature that has been

reported for other Brunfelsia species (Plowman,

1977).

9. Brunfelsia bonodora (Veil.) J. F. Macbr., Field

Mus. Nat. Hist., Bot. Ser. 13, 5-B(l): 152.

1962. Figure 22.

Besleria bonodora Veil., Fl. Flumin. 261. 1829

("1825"). Ic. 6, t. 80. 1831 ("1827"). Lectotype

(designated here): Vellozo, J.M. da C., Fl. Flu-

min. Ic. 6, t. 80. 1831 ("1827"). Representative

collection: BRAZIL. Rio de Janeiro, Serra de Es-

trella, Mar. 1833, Riedel 84 (LE).

Shrub 1-2 m tall. Branches ascending, spread-

ing, naked, glabrous. Bark thin, light brown.

Branchlets glabrous or sparsely puberulent, gray-

ish brown. Leaves scattered along branchlets,

blade 5-14.5 cm long; 2-5 cm wide, narrowly el-

liptic to lanceolate, acuminate at apex, blunt to

cuneate at base, glabrous, at times puberulent at

midrib on lower surface, shiny, dark green on up-

per surface, pale green beneath, firmly membra-

naceous, lateral nerves 5-8, more or less arcuate;

petiole 2-6 mm long, slender, glabrous. Inflores-

cence terminal, compact, unbranched, axis 3-8

mm long. Flowers 3-15, rarely reduced to 1 or 2;

blue fading to white with age, fragrant. Bracts 2-

12 mm long, lanceolate, glabrous or puberulent,

caducous. Pedicel 6-12 mm long, slender, thick-

ened slightly at apex, glabrous. Calyx 11-14 mm

long, about 5 mm in diameter, tubular-campanu-

late to campanulate, ellipsoid to obovoid in bud,

rarely somewhat inflated, glabrous, firmly mem-

branaceous, striately veined, teeth 3-4 mm long,

triangular to ovate, acute to blunt at apex, rarely

apiculate, calyx to 15 mm long in fruit, subcor-

iaceous, conspicuously veined, enclosing fruit.

Corolla tube 17-25 mm long, 2 mm in diameter,

twice as long as calyx, glabrous or sparsely glan-

dular; limb 20-35 mm in diameter, plane spread-

ing, lobes rounded to oblong, subtruncate or

rounded at apex, rarely emarginate, overlapping

at sides, narrowed abruptly at base. Stamens in-

cluded in upper part of corolla tube; filaments

about 4 mm long, subligulate; anthers 1.5 mm

long, reniform. Ovary 1.5 mm long, ovoid-coni-

cal; style slender, equaling the filaments; stigma 1

mm long, briefly bifid, upper lobe slightly larger,

obtuse. Capsule 8 mm long, 6-8 mm in diameter,

partially enclosed by calyx, ovoid-turbinate, apic-

ulate at apex, smooth, pericarp thin-walled, drying

crustaceous, sparingly dehiscent. Seeds about 5,

incompletely known.

DISTRIBUTION — Brazil (Rio de Janeiro, Sao Pau-

lo). See Figure 20.

SPECIMENS EXAMINED — BRAZIL. Rio de Ja-

neiro: Serra de Estrella, 700 m, 29 Oct. 1946,

Brade 18643 (RB), 12 Nov. 1823, Riedel 18 (A,

LE, mixed collection; NY, mixed collection; us), 14

Dec. 1823, Riedel 59 (A, LE, mixed collection);

PLOWMAN: REVISION OF BRUNFELSIA

59

( Tal)

60

FIELDIANA: BOTANY

Macahe, May 1832, Riedel sp. nov. (LE); Barra

Mansa, Fazenda Paraiso, 3 Dec. 1960, Duarte

5477 (RB, us). Sao Paulo: Taubate, Dec. 1817,

Martius 541 (M). Without state: Riedel 112 (us).

Brunfelsia bonodora was first described in the ge-

nus Besleria by Vellozo in Flora Fluminensis, a

work that appeared in 1829 although 1825 ap-

pears on the title page (Carauta, 1973). Vellozo's

scanty description was accompanied by a some-

what stylized illustration. This drawing must now

be considered the lectotype of the species since

no specimens of Vellozo exist (Fig. 22).

In his monograph of Brunfelsia in 1846, Ben-

tham erroneously placed Vellozo's species in syn-

onymy with his B. latifolia (based on Franciscea

latifolia Pohl). This concept also included mate-

rial now recognized as B. grandiflora subsp.

schultesii of the western Amazon and Andes. This

problem was discussed by Monachino (1953) and

Macbride (1962), but neither of these workers

were able to resolve the issue, primarily because

they saw neither the type materials nor represen-

tative collections from southeastern Brazil.

It is now clear that the name B. latifolia and its

synonym, B. maritima Benth., apply to a distinct

species known only from the sand restingas

around the city of Rio de Janeiro. Brunfelsia bon-

odora is represented by several collections from

the mountains of the state of Rio de Janeiro that

agree with Vellozo's description and plate. This

concept is, however, not identical with Macbride's

interpretation of the species since he included ma-

terial here referred to as B. grandiflora subsp.

schultesii.

Nevertheless, B. bonodora is related to both B.

latifolia and B. grandiflora subsp. schultesii,

which is in part responsible for the confusion of

these concepts. Brunfelsia bonodora is distin-

guished from B. latifolia in having a more robust,

upright habit; longer, lanceolate, and apically acu-

minate leaves; and longer pedicels. In addition, B.

bonodora differs ecologically, growing in the

montane rain forests of the Serra do Mar and not

in open, coastal formations (restingas). There

seems to be some overlap in these two species,

but the status of possible hybrids cannot be de-

termined with so few specimens at hand, most of

which lack adequate collection data. In fact, only

two modern collections of B. bonodora are

known.

Macbride (1962) considered B. bondora to be

conspecific with the plant here treated as B. gran-

diflora subsp. schultesii. On close examination,

however, there are several points of difference be-

tween them. Brunfelsia bonodora has longer ped-

icels (6-12 vs. 2-5 mm), a more compact inflo-

rescence, larger bracts, and the capsule almost

completely enclosed by the calyx in fruit. Ac-

cording to the specific epithet, B. bonodora is fra-

grant; B. grandiflora almost never is.

Geographically, B. bonodora and B. grandiflo-

ra are separated by several thousand miles. They

exhibit, however, a distribution similar to those of

B. hydrangeiformis and B. mire and may be con-

sidered to be vicarious species that arose from a

common ancestor. Brunfelsia bonodora is known

only from the coastal montane rain forests in the

state of Rio de Janeiro, at elevations up to 700 m.

It flowers from October to December and fruits

during May and June. Additional collections of

this plant with complete field data will greatly aid

in understanding its specific characters and rela-

tionships.

10. Brunfelsia brasiliensis (Spreng.) L. B. Sm. &

Downs in Reitz, Fl. Ilustr. Catar., pt. 1, Sola:

303. 1966.

Gerardia brasiliensis Sprang., Syst. Veg. 2: 806.

1825. Type: BRAZIL. Brasilia meridionalis,

without locality, Sellow s.n. (type destroyed at

Berlin; representative collection and possible

isotype, M).12

Shrub 0.3-2 m tall. Bark roughish, reddish to

light brown. Branches numerous, arising from

near base, leafy, yellowish brown, pilose, es-

pecially so when younger, or glabrous. Branch-

lets of current year leafy, slender, erect, densely

pubescent with yellowish brown, nonglandular

hairs. Leaves scattered along branchlets, some-

times more crowded toward apex, blade 3-13 cm

long, 1-4.5 cm wide, lanceolate to oblong-lanceo-

12 In the manuscript Plowman gave no obvious indi-

cation that he wished to lectotypify this species as cited.

We have thus left the lectotypification of this species to

later Brunfelsia specialists. — Eds.

FIG. 22. Lectotype of Brunfelsia bonodora (originally as Besleria bonodora). From Vellozo, J. M. C, Flora

Fluminensis (1881).

PLOWMAN: REVISION OF BRUNFELSIA

61

late, more rarely obovate-oblong, acute to acu-

minate at apex, occasionally blunt, cuneate to nar-

rowed at base, the upper surface sparingly pubes-

cent, dark to light green, rarely nitid, subscabrous

or glabrate, the lower surface variously pubescent,

often villous at midrib, rarely glabrous, paler

green, firmly membranaceous to coriaceous, lat-

eral nerves 6-10, mostly straight; petiole 1-5 mm

long, puberulent to villous. Inflorescence terminal

and subterminal, corymbiform, compact and near-

ly sessile or lax-elongate, with 5-many flowers,

rarely reduced to 1-2, the axis 1-2 cm long, pu-

bescent-villous. Bracts small, 2-15 mm long, lan-

ceolate, ciliolate at margin, pubescent, caducous.

Flowers often showy, violet fading to pale lav-

ender or white with age, odorless, short pedun-

culate. Peduncles 2-8 mm long, thickened at apex

and articulating with pedicel, persistent, villous or

glandular-pubescent, bearing 1-3 bracts. Bracts

minute, 1-2 mm long, linear-lanceolate. Pedicel

2-15 mm long, erect, pubescence yellowish to

yellowish brown, sparsely pilose to villose, rarely

glandular, thickening in fruit. Calyx 8-22 mm

long, 3-8 mm in diameter, tubular to tubular-cam-

panulate, with variable yellow to yellowish brown

pubescence, sparsely to densely villous, rarely pi-

lose or glandular, light green to purple, firmly

membranaceous, teeth 1-5 mm long, erect, some-

times recurved in fruit, triangular to ovate, acute

at apex; calyx in fruit campanulate to urceolate,

becoming coriaceous, striate-nerved. Corolla tube

17-25 mm long, 1-3 mm in diameter, 1-2 times

as long as calyx, sparsely pilose or glabrous, pale

violet below, white toward apex, limb 18-32 mm

in diameter, plane spreading, lobes subequal,

broadly rounded to truncate at apex, lateral mar-

gins. Stamens included in upper part of tube; fil-

aments slender, slightly broader at base, glabrous,

white, lower pair 2-4 mm long, upper pair 4-5

mm long; anthers 1 mm long, oblong-reniform.

Ovary 1-5 mm long, ovoid-conical, green; style

17-18 mm long, filiform, bright green; stigma 1-

2 mm long, briefly bifid, the upper lobe somewhat

larger, bright green. Capsule 8-13 mm long, 8-

10 mm in diameter, enclosed by calyx, ovoid to

subglobose, apically apiculate, smooth, shiny,

light to dark green, pericarp thin, 0.5 mm thick,

cartilaginous to crustaceous, dry at maturity, spar-

ingly dehiscent. Seeds (2)7-10, 4-6 mm long, 3

mm in diameter, oblong-ovoid, more or less an-

gular, dark reddish brown, reticulate-pitted. Em-

bryo 4 mm long, very slightly curved; cotyledons

1.5 mm long, ovate-elliptic.

DISTRIBUTION — Brazil. See Figure 23.

Key to the Subspecies of Brunfelsia brasiliensis

(Spreng.) L. B. Sm. & Downs

1 . Inflorescence terminal, nearly sessile, compact,

with 1-8 flowers; peduncles shorter than ped-

icels; pedicels 5-15 mm long; calyx 15-22 mm

long; leaves often glabrescent

lOa. subsp. macrocalyx

1. Inflorescence terminal and subterminal, lax or

dense, often branched, with 8-many flowers;

peduncles equaling or longer than pedicels;

pedicels 2-5 mm long; calyx 7-17 mm long;

leaves usually pubescent-villous beneath

lOb. subsp. brasiliensis

lOa. Brunfelsia brasiliensis subsp. brasiliensis.

Figure 24.

Franciscea acuminata Pohl, PI. Bras. Icon. Descr. 1:

4, t. 3. 1826. Type: BRAZIL. Rio de Janeiro: in

umbrosis inter frutices ad Mandioca, Schott 6163

(holotype, BR).

Brunfelsia acuminata (Pohl) Benth. in DC., Prodr.

10: 199. 1846.

Brunfelsia brasiliensis var. acuminata (Pohl) L. B.

Sm. & Downs in Reitz, Fl. Ilustr. Catar, pt. 1,

Sola: 305. 1966.

Franciscea ramosissima Pohl, PI. Bras. Icon. Descr.

1: 5, t. 4. 1826. Type: BRAZIL. Rio de Janeiro:

ad Serra Tingua, Schott 3720 (holotype, w; iso-

types, GH, w).

Brunfelsia ramosissima (Pohl) Benth. in DC.,

Prodr., 10: 199. 1846.

Franciscea confertiflora Pohl, PI. Bras. Icon. Descr.

1: 6, t. 5. 1826. Type: BRAZIL. Minas Gerais:

inter frutices locis umbrosis in via de Villa Fan-

ado vel Bom Successo dicta ad Villa do Principe,

Oct.-Nov. 1820, Pohl 3478 (holotype, w; iso-

type, w).

Brunfelsia confertiflora (Pohl) Benth. in DC.,

Prodr., 10: 199. 1846.

Brunfelsia ramosissima var. confertiflora (Pohl) J.

A. Schmidt in Mart., Fl. Bras. 8(1): 260. 1864.

Franciscea divaricata Pohl, PI. Bras. Icon. Descr.

1: 6, t. 6. 1826. Type: BRAZIL. Minas Gerais:

inter frutices, locis umbrosis in via de Sao Joao

d'El Rey, ad Villa Paracatudo Principe, floret ab

Octubri usque ad Martium, 1818, Pohl 593, 6165

(holotype, w; isotypes, F, GH LE, w).

Brunfelsia ramosissima var. laxiflora J. A. Schmidt

in Mart., Fl. Bras. 8(1):260. 1864. Type: BRA-

ZIL. Brasilia meridionalis, without locality, Sel-

low s.n. (holotype, M).

Brunfelsia ramosissima var. parcifolia J. A.

Schmidt in Mart., Fl. Bras. 8(1):260. 1864. Type:

BRAZIL. Minas Gerais: without locality, Claus-

sen 171 (holotype, BR).

62

FIELDIANA: BOTANY

FIG. 23. Distribution of Brunfelsia brasiliensis (solid square) and B. chiricaspi (solid circle).

Leaves 3-13 cm long, 1-4.5 cm wide, puber-

ulent or scabrous above, often glabrescent, vari-

ably villous beneath, often densely so at midrib,

firmly membranaceous to subcoriaceous. Inflo-

rescence variable, dense to lax, occasionally elon-

gate and well-developed, with 10-many flowers,

rarely reduced to about 5. Peduncles equaling or

longer than pedicels, 2-10 mm long, rufo- villous,

not glandular. Pedicels 2-5 mm long, villous. Ca-

lyx 7-15 mm long, variably villous with dense,

yellowish brown hairs, rarely pilose or glabres-

cent, eglandular. Corolla tube 17-22 mm long,

1.5-2 times as long as calyx; limb 16-30 mm in

diameter.

DISTRIBUTION — Brazil (Distrito Federal, Goias,

Minas Gerais, Parana, Rio de Janeiro, Sao Paulo).

SPECIMENS EXAMINED — BRAZIL. Distrito

Federal: gallery forest ca. 10 km NW of Planal-

tina, 950 m, 27 Feb. 1966, Irwin et al. 13194 (us);

gallery forest ca. 10 km E of Lagoa Paranoa, near

DF-6, 1000 m, 25 Feb. 1970, Irwin et al. 26556

(wis); *Brasilia; Bacia do Rio Sao Bartolomeu,

Corrego Forquilha, Heringer et al. 6148 (F).

Goias: *25 km NW of Brasilia, Irwin et al. 15799

(F). Minas Gerais: Mun. Santo Luzia, Fazenda

da Chicaca, 1 100 m, 20 Nov. 1945, Assis 102 (GH,

R, us), Assis 137 (GH), Assis 142 (GH), 13 Dec.

1945, Assis 181 (GH, uc, us); Lagoa Santa, Oct.-

Nov. 1863, Engler 1006 (c), 14 Jan. 1864, Engler

1009 (c), 30 Oct. 1863, Engler sp. nov. (c), 25

Mar. 1933, Mello Barreto 7786 (F), Warming 103

(NY, s); Serra do Cipo, km 120, 2 Sep. 1933, Mel-

lo Barreto 7788; Mun. Belo Horizonte, Morro do

Candido, Jan. 1934, Atamp. 6560 (R); Matta da

Caixa d'Areia, 15 Dec. 1918, Gehrt 3276 (SP);

PLOWMAN: REVISION OF BRUNFELSIA

63

FIG. 24. Brunfehia brasiliensis subsp. brasiliensis (originally as Franciscea confertiflora). From Pohl, J. E.

Plantarum Brasiliae Icones et Descriptiones (1826).

64

FIELDIANA: BOTANY

estrada para Lagoa Santa, 15 Jan. 1951, Joly 1133

(SP); Villa Paraiso, 20 Nov. 1922, Mello Barreto

7789 (F); Barreiro, 21 Jan. 1933, Mello Barreto

7791 (F); Serra do Taquaril, 18 Jan. 1933, Mello

Barreto 7791 (F), 29 Mar. 1933, Mello Barreto

7793 (F); Jardim Botanico, 17 Nov. 1933, Mello

Barreto 7794 (R); Serra da Mutuca, 27 Jan. 1945,

Moreira 5775 (GH); Fazenda da Baleia, 4 Mar.

1940, Oliveira 21 (us); Morro das Pedras, 3 km

S of Belo Horizonte, 1000 m, 12 Feb. 1945, Wil-

liams & Assis (GH); near Lagoa Pampulha, 1000

m, 8 Mar. 1945, Williams & Assis 6088 (F, GH,

MO, NY); Mun. Carandai, Carandai, km 416, 25

Nov. 1946, Duarte 631 (GH, RB); Hermilo Alves,

10 Jan. 1965, Duarte 8741 (GH, RB); Mun. Ouro

Preto, Itacolomi, 1400 m, 31 Dec. 1950, Macedo

2829 (MO, NY, s); Mun. Araxa, Barreiro, 10 Feb.

1951, Macedo 3131 (s, SP, us); Mun. Santa Bar-

bara, Serra do Cara9a, 14 Apr. 1933, Mello Bar-

reto 7784 (F); Mun. Tombos, Fazenda das Antil-

has, 21 Jan. 1936, Mello Barreto 7767 (F, R);

Mun. Diamantina, Serra dos Crystais, 6 Nov.

1937, Mello Barreto 9603 (F, R); Po de Araca, 17

Nov. 1937, Mello Barreto 9833 (F, R); Mun. Ja-

boticatubas, Lagoa de Dona Ignacia, 6 Jan. 1940,

Mello Barreto 10530 (R); Mun. Vi£osa, state ag-

ricultural school, 17 Dec. 1958, Irwin 2269 (NY,

R, us); Fazenda de Dr. Christioma, 700 m, 30 Dec.

1930, Mexia 5467 (BM, BH, F, G, GH, MICH, MO, NY,

PH, s, uc); Mun. Serro, 3 km from Serro, 12 May

1945, Williams & Assis 6859 (GH); Aguas Virtu-

osas, 15 Jan. 1919, A. Amaral s.n. (SP); So^ego, Jan.

1921, A. Amaral s.n. (SP); Carmo do Rio Cairo,

Fazenda Corrego Bonito, 5 Sep. 1961, Andrade

1024, Emmerich 985 (us); Campanha, Nov. 1896,

Brandao 2066 (R); Serra do Cipo, km 131-132,

Palacio, 4 Dec. 1949, Duarte 2029 (GH, RB); Ar-

raial do Pinheiro, Oct. 1840, G. Gardner 5063

(BM, CGE, F, G, GH, K, NY, p, us, w); Serra do Pal-

mital, 24 Jun. 1884, Glaziou 15311 (c, K, R);

Santa Barbara do Mato Dentro, 12 Jan. 1921,

Hoehne 4920 (SP, us); Caeti, Nov. 1915, Hoehne

6129 (R); Sabara, Jan. 1916, Hoehne 6846 (us);

Serra de Grao Mogol, 1,100 m, 12 Nov. 1938,

Markgraf 3495 (RB); prope Pitangui, Cueillie a

Sabana, 11 Mar. 1862, Netto s.n. (R); Barbacena,

21-24 Jun. 1879, Netto et al. s.n. (R); BR-4, km

938, 40 km depois de Medina, 16 Jan. 1965,

Pabst 8340, Pereira 9451 (HB, RB); Serra do Cipo

(Concei9ao), 1,200 m, 22 Dec. 1948, Palacios,

Balegno & Cuezzo 3467 (B, w); Caldas, 3 Feb.

1857, Regnell 1-375 (BR, c, M, p, R, s, us); Itabura,

1816-1821, St. Hilaire D-80 (MO, NY); Penha, St.

Hilaire s.n. (us); Sitio, 20 Nov. 1905, Sampaio

68, 74 (R, us); Serra de Itabira do Campo, 1 1 Sep.

1887, Schwacke s.n. (R, us), 20 Dec. 1888,

Schwacke s.n. (R); Congonhas do Campos, 1843,

Stephan s.n. (BR); Serra do Frio, 1833, Vauthier

541 (GH, K, L, w); Serra da Mantiqueira, Passa

Quatra, Esta9ao Florestal da Mantiqueira, 31 Jun.

1949, Vidal s.n. (R); without locality, Capanema

s.n. (GH, RB), 1842-1843, Claussen 146 (BM, CGE),

Claussen 147 (BM, CGE), Claussen 157 (w), 1840,

Claussen 334 (BR), Claussen s.n. (BM, BR, GH, K,

w), Weddell 1648 (p); *Serra do Espinha9o, SW

of Rio Jequiti and Mendanha on road to Diaman-

tina, Anderson 8949 (F, MO); *Municipio Tombas;

Fazenda das Antilhas, Mello Barreto 7787 (F, R);

*Belo Horizonte; Serra do Taquaril, Hab. Canga,

Barreto 7790 (F); *ca. 17 km SW of Minas Ger-

ais-Bahia state border, Davidse et al. 11611 (MO);

*13 km W of Lavras along Highway 265 to Bar-

bacena, Davidse & Ramamoorthy 10709A (F);

*Municipio Pato de Minas; Corrego Barreiras,

Hatschbach 42939 (F); *Municipio Itamonte; Ser-

ra da Mantiqueira, Hatschbach & Kummrow

45569 (F); *Municipio Diamantina; Biri-Biri,

Hatschbach & Pelanda 27993 (GH); *Serra do Es-

pinha9o, Serra do Itabirito ca. 45 km S of Belo

Horizonte, Invin et al. 19549 (MO); *Serra do

Espinha9o, Serra do Cipo, Irwin et al. 20150

(MO); *Serra do Espinha9o, Irwin et al. 29189

(MO); *Serra de Espinha9o, S of Ouro Preto, Irwin

et al. 29407 (MO); *Serra do Espinha9o, ca 34 km

E of Belo Horizonte, Irwin et al. 30628 (MO);

*Municipio Lima Duarte; Conceicao do Ibitipoca,

Parque Florestal Estadual de Ibitipoca, Souza et

al. 521 (F). Parana: *Municipio Senges, Rio Ca-

juru, Hatschbach 43398 (F); *Ponte Nova-Minas

Gerais, Hatschbach & Ahumada 31333 (MBM).

Rio de Janeiro: Monte Corcovado, 300-700 m,

6 Mar. 1924, L.H. & E.Z. Bailey 758 (BH); by way

of Laranjeiras, 31 Dec. 1825, Burchell 1372 (p),

1841, Claussen s.n. (w), 26 Dec. 1920, Ducke &

J.G. Kuhlmann s.n. (RB); summit, 1836, Gardner

248 (BM, CGE, G, K, w), 10 Jan. 1865, Glaziou 812

(BR, c, R), May 1838, Guillemin 825 (P), Langs-

dorff s.n. (us), Nov.-Dec. 1832, Luschnath s.n.

(BR), 1834, Luschnath s.n. (BR), 700 m, 22 May

1969, Plowman 2786, Sucre 5086, (GH, ECON, PH,

RB), Feb. 1832, Riedel 467 (L, GOET, LE), Nov.-

Dec. 1832, Riedel 1190 (L, LE), 12 Jul. 1915, Rose

& Russel 20221 (GH, us); Pico da Tijuca, 18 Jan.

1931, Brade 10569 (R), Brade 10583 (GH, R),

1838, Guillemin 41 (P); Excelsior, 3 Feb. 1930,

J.G. Kuhlmann s.n. (GH, RB), Dec. 1824, Riedel

1249 (LE), 23 Aug. 1896, Ule s.n. (R); mata da

Tijuca, 10 Nov. 1896, Ule s.n. (R), 1836, Vauthier

PLOWMAN: REVISION OF BRUNFELSIA

65

42 (F, c, P). Novo Friburgo, 9 Dec. 1918, Curran

619 (GH, s, SP), Curran 623 (F, GH); Morro Pos-

sole, Sitio Dr. Dungs, 1300 m, 28 Jan. 1968, Pabst

9115 (HB), 23 Dec. 1887, Glaziou 17170 (BR, c,

LE, P, us); Petropolis, 1883, Glaziou 14171 (BR, c,

c); pres de la roche de Juraracacu, 25 Mar. 1879,

Glaziou 11394 (c, P); Carangola, Sep. 1943, Goes

& Constantino 543 (RB); Retire, Nov. 1943, Goes

& Constantino 710 (RB); Caetetu, Dec. 1943,

Goes & Constantino 883 (RB); Estrada de Con-

torno, 650 m, 2 Feb. 1968, Sucre 2360, Braga 203

(GH, RB); Mata do Judeo, 700 m, 7 Dec. 1968,

Sucre 4213, Braga 1169 (RB), 1858-1860, Wawra

& Maly 426 (LE, w); Terexopolis, Parque Nacional

da Serra dos Orgaos, 16 Dec. 1942, Duarte de

Barros 1195 (RB); Posso, Morro das Antenas de

Televisao, 12 Feb. 1968, Sucre 2339, Braga 182

(GH, RB); Serra dos Orgaos, 3000 ft, 7 Jan. 1837,

Gardner 564 (BM, CGE, G, GH, K, NY, us, w), 1885,

Glaziou 15132 (c, G, K, LE), May 1839, Guillemin

954 (P); Imbuhy Lane, Jan. 1838, Miers 4495 (us),

900 m, Dec. 1952, Vidal 11-5784 (R); Santa Ana,

1822-1823, Beyrich s.n. (P, s); Montagnes pres

Sumidoro, Laland s.n. (P); Cantagallo, Peckolt

181 (w); ad confines Rio de Janeiro-Minas Ger-

ais, in graminosis partis superioris montis Itatiaia,

2000-2500 m, Sep. 1901, Wettstein & Schiffner

s.n. (w); without locality, 1834, Gaudichaud 442

(G, P), Raben 703 (BR), 1840, Regnell 173 (s),

1832, Riedel s.n. (GH, GOET, L, LE, NY, P, s, us, w),

St. Hilaire A-746 (P), 1838-1842, U.S. South Pa-

cific Exploring Expedition s.n. (NY), Oct. 1916,

Vincent s.n. (L), Widgren 486 (s); *Morro Quei-

mado, Almeida 1286 (F); *Floresta da Tijuca; Ped-

ra do Archer, Angeli 250 (F); *Morro Queimado,

Brade 18785 (F); *Rio de Janeiro, Pedra da Gav-

ea, base da Chamine, Carauta 2420 (F); *Gua-

nabara; Pedra da Gavea, Mesa, Carauta and de

Oliveira 1523 (F); *Guanabara; Tijuca, Castellan-

os 23973 (F); *Municipio de Macae, Pico de Fra-

de, Farney et al. 584 (F); *Municipio de Macae,

Pico de Frade, Farney et al. 597 (F, RB); *Rio de

Janeiro; Alto de Boa Vista, Estrada da Vista Chi-

nesa, mata atras do Decam, Landrum et al. 4150

(F); *Tijuca, Bom Retire, Pabst 7411 (F); *Estrada

da Vista Chinesa, Strang 172 (F); *Pedra da Gav-

ea, Sucre 7446 (F); *Morro Queimado, Sucre

8002 (F); *Serra da Carioca, no topo do Morro

Queimado, Vianna et al. 1574 (F). Sao Paulo:

Jundiahy, region of Morro da Mursa, 1000 m, 4

Apr. 1915, Brade 7020 (SP); Sao Paulo, Vila Ema,

Dec. 1932, Brade 12119 (R), Jan. 1952, Brade

21218 (RB); Campos do Jordao, 10 Sep. 1937,

Campos Porto 3381 (R), Campos Porto s.n. (GH,

RB), 22 Oct. 1938, Hashimoto 1 (RB), 27 Jul. 1898,

Lofgren 3986 (SP); Atibaia, Nov. 1931, Constan-

tino s.n. (RB); Guaratingueta, 27 Jan. 1920, Gehrt

3667 (SP); Mogi das Cruzes, 18 Apr. 1921, Hoeh-

ne 5481 (SP, us), Oct. 1833, Riedel 1466 (G, LE,

NY, us); Sao Jose dos Campos, 17 Dec. 1908, Lof-

gren 86 (RB, s); ca. 7.7 km SW em linha reta da

pra9a principal de Sao Jose dos Campos, 300 m,

NW da Via Dutra, 600 m, 25 Oct. 1961, Mimura

53 (us); Ipiranga, Dec. 1912, Luderwaldt 1086

(RB, SP), Jun. 1907, Usteri 23/306 (NY); Taubate,

Lund 754, Lund 756 (c); inter Lorena et Taubate,

Apr. 1833, Riedel 1465 (K, LE, NY); Serra da Bo-

caina, Jan. 1925, Lutz s.n. (R); in nemorosis editis

retro Bananal in via publica versus Sao Paulo,

Dec. 1817, Martius s.n. (M); Morro Formozo,

Dec. 1817, Martius s.n. (M); Serra da Mantiqueira,

Fazenda de Cruzeiro do Dr. Major Novaes, 3-12

Jan. 1884, Saldanha 8587 (R); *Municipio Sao

Jose de Barreiro; Parque Nacional da Serra da Bo-

caine, Fazenda das Garrafas, Lima, H.C. de, &

Ramamoorthy 1224 (F); *Municipio Registro;

Rod. Registro-Sete B arras, Hatschbach & Gui-

mardes 45004 (F); *Serra da Bocaina, W. Hoehne

6154 (F), *W. Hoehne 6155 (F); *Salesopolis; Es-

ta£ao Biologica de Boracea, perto do Rio Coruja,

Mattos, J., & Mattos, N., 14284 (F); * Serra da

Bocaina; Parque Nacional, Occhioni 9174 (F).

Without state: Bowie & Cunningham 228 (BM,

s), 1842, Dupre s.n. (BR, NY, P), Germain 171 (F),

Guillemin 171 (G), Lobb s.n. (K), Martius 1292

(G, L, M, NY), Pohl 263 (PR), 265 (PR), 266 (PR),

Schiich s.n. (BR, w), Sellow s.n. (BM, BR, CGE, F,

K, L, M, s), Weir 23 (CGE, K).

lOb. Brunfelsia brasiliensis subsp. macrocalyx

(Dusen) Plowman comb, et stat. nov.

Brunfelsia hopeana van macrocalyx Dusen, Archiv.

Mus. Nac. Rio de Janeiro 13: 94. 1905. Type:

BRAZIL, Rio de Janeiro, na mata da encosta da

Serra do Itatiaia, in silva, 1600 m, 5 Jan. 1896,

Ule 636 (holotype, R 66551).

Leaves 3-8 cm long, 1.2-2.2 cm wide, gla-

brous and often nitid above, glabrescent beneath

or with pubescent midrib and nerves, subcoria-

ceous to coriaceous. Inflorescence short, subses-

sile, terminal, with 2-8 flowers, rarely reduced to

1 . Peduncles 2-5 mm long, usually much shorter

than pedicel, sparsely long-pilose or densely glan-

dular-pubescent. Pedicel 5-15 mm long, villous

or glandular-pubescent. Calyx 15-22 mm long,

pilose or glandular-pubescent or glabrescent. Co-

rolla tube 1 8-25 mm long, scarcely exserted from

66

FIELDIANA: BOTANY

calyx or up to H/2 times as long; limb 25-32 mm

in diameter.

DISTRIBUTION — Brazil (Minas Gerais, Parana,

Rio de Janeiro, Santa Catarina, Sao Paulo).

SPECIMENS EXAMINED — BRAZIL. Minas Ger-

ais: Mun. Diamantina, Guinda, 5 Nov. 1937, Mel-

lo Barreto 9479 (F, R). Parana: Ipiranga, Monte

Alegre, 1100 m, Dusen 3388 (R); Bauhado, 13

Dec. 1911, Dusen 13562 (GH, NY, s); Guaratuba,

21 Feb. 1952, Reitz 4375 (us); *Palmeira, Hatsch-

bach 11878 (MBM); *Barra Rio Papagaios (Balsa

Nova), Hatschbach 18747 (MBM); *Municipio

Guaratuba; Rio Sai, Hatschbach 19447 (MBM);

*Municipio Jaquariaiva; Fazenda Chapada S. An-

tonio, Hatschbach 20997 (GH); *Municipio Pir-

quara; Serra do Emboque, Hatschbach 26532

(MBM); *Municipio Pirai do Sul; Joaquim Murtin-

ho, Hatschbach 39209 (GH, MBM); *Municipio

Palmeira, Rio Tibagi, Kummrow 970 (GH); *Bran-

dalize, Lindeman and Haas 1174 (u). Rio de Ja-

neiro: Alto Macahe de Nova Friburgo, 22 Jan.

1874, Glaziou 6908 (BR, c, s), Glaziou 13478 (c,

F, LE); 1882, Glaziou 14172 (c, G, LE); Petropolis

a Itamaraty, 25 Mar. 1879, Glaziou 11395 (BR, c,

G, LE, R); Serra dos Orgaos, 18 Nov. 1887, Gla-

ziou 17169 (BR, c, p, R), 2,200 m, Jun. 1915, Liitz-

elburg 6187 (M); Morro Assu, 2400 m, Jan. 1916,

Lutzelburg 6559 (M, NY), Feb. 1838, Miers s.n.

(BM); Serra do Itatiaia: 1900 m, Sep. 1913, Brade

& Tamandare 6576 (SP), 1915, Campos Porto 173

(R, RB), 2100 m, Jun. 1902, Dusen 569 (R), 18

Oct. 1903, Dusen 2023 (K, R, s, us), 1750 m, 28

Oct. 1927, Ginzberger 93 (F), Moreira s.n. (R);

Parque Nacional de Itatiaia, km 15, 17 Aug. 1948,

Occhioni 1131 (GH, RB); prope Registro, 1800 m,

14 Nov. 1954, Pabst & Brade 10283 (us); Ma-

cieras, 1900 m, Sep. 1934, Brade 14074 (B, R, RB),

1800-2000 m, 23 Oct. 1927, Zerny s.n. (w); entre

Maromba e Macieiras, 1820 m, 29 May 1969,

Plowman 2891, Sucre 5191 (ECON, GH, MICH, PH,

RB, uc, YEN), 1,700 m, Plowman 2906, Sucre 5206

(GH, MICH, PH, RB), 1200-1900 m, 22°27'S,

44°39'W, 10 Jan. 1929, L.B. Smith 1788 (GH, s,

us), 26 Feb. 1965, Vogel 677 (us); *Parque Na-

cional de Itatiaia, Lindeman & Haas 5157 (u);

*Parque Na9ional de Itatiaia; entre Maromba e

Macieras, Sucre 5806 (F). Santa Catarina: Mun.

Salete, Taio, Ribeiro Grande, 16 Dec. 1950, Reitz

3985 (G, NY, uc, us); Mun. Rio do Sul, Alto Mat-

ador, 800 m, Reitz & Klein 7601 (us). Sao Paulo:

Alto da Serra, 2 Feb. 1913, Brade 6037 (s, SP),

Brade 7625 (us), 2 Dec. 1920, Hoehne 4595 (SP,

us), 22 Dec. 1919, Hoehne 8250 (SP, us), 27 Nov.

1902, Puttemans 5890 (SP); Serra da Bocaina,

1650 m, 12 May 1951, Brade 20953 (RB); Fazen-

da Campo Grande, Linha Ferrea Ingleza, 17 Nov.

1892, Edwall 1981 p.p. (B, us); Esta?ao Biologica,

800-900 m, lat. 23°46'S, long. 46°20'W, 23 Feb.

1929, Smith 1989 (GH, us); Campos do Jordao,

Umuarama, 1,750 m, 22 Nov. 1949, M. Kuhlmann

2338 (SP), 22 Nov. 1957, M. Kuhlmann 4309 (SP);

proximo do Rio Guaratuba, 18 Mar. 1958, M.

Kuhlmann 4359 (SP).

A name that has been in the synonymy of this

species, Gerardia brasiliensis Spreng., was first

published in 1825. Although it was not mentioned

by Pohl, both Bentham and Schmidt placed it in

synonymy with B. ramosissima. This specific ep-

ithet was only recently recognized by Smith and

Downs (1966) as the first legitimate name and

basionym for the species, which now must be

called B. brasiliensis (Spreng.) L. B. Sm. &

Downs.

Sprengel described G. brasiliensis from a spec-

imen collected by Sellow in Brazil, and his de-

scription, albeit brief, more or less agrees with the

present concept of the species. There is, however,

one point of difference in Sprengel's description,

where he writes "pedunculis axillaribus solitar-

iis." Usually this species has terminal, branched

inflorescences, although there may be subterminal

axes produced in the uppermost leaf axils.

Unfortunately, Sprengel's herbarium was dis-

mantled and sold after his death by his son. Some

of the specimens survived and were eventually

deposited at the Berlin herbarium, where they

were destroyed in the Second World War (Stafleu,

1967; Stafleu & Cowan, 1985). Sellow apparently

made several collections of this species with nu-

merous duplicates, none of which bears collection

numbers or localities. It is thus impossible to de-

termine which specimens might be isotypes of the

original specimen that Sprengel described. As a

representative collection, I selected another spec-

imen of Sellow that was sent to Munich from Ber-

lin in exchange [this is not to be taken as neotyp-

ification of the name (see footnote 11) — Eds.}.

In 1826 Pohl published seven new species of

his new genus Franciscea, including F. ramosis-

sima, F. acuminata, F. confertiflora, and F. di-

varicata. He stated at that time that these "spe-

cies" were all closely related but believed that

their differences were great enough for them to

be considered separate entities. Pohl's specific

concepts were generally followed by Bentham,

who in 1846 transferred all of Pohl's species of

PLOWMAN: REVISION OF BRUNFELSIA

67

Franciscea to the genus Brunfelsia, omitting only

F. divaricata, which he placed in synonymy with

B. confertiflora. According to Bentham, B. con-

fertiflora differed from B. ramosissima in having

a many-flowered, crowded inflorescence, an in-

flated calyx, and the corolla tube only W2 times

as long as the calyx.

Brunfelsia acuminata as maintained by Ben-

tham was merely a glabrous form of B. ramosis-

sima. The type specimen from which Pohl's orig-

inal drawing was made is preserved in Brussels

(BR). A specimen at Vienna (w), collected and ap-

parently annotated by Schott as F. acuminata, has

a well-developed indument and does not conform

to the description of this species. As mentioned

above in the discussion of B. martiana, there has

been considerable confusion of collection data

and numbers in the Brazilian collections of Pohl

and Schott. For example, the same collection in

different herbaria may bear the same number or

numbers with either or both names. For this rea-

son the numbers of these two collectors should be

used with some caution.

Schmidt (1864), in his account of Brunfelsia for

Flora Brasiliensis, treated this group of "species"

as a single variable species, B. ramosissima, with

three varieties. Variety laxiflora included B. ra-

mosissima and B. acuminata of Bentham, and va-

riety confertiflora conformed more or less with

his B. confertiflora. Schmidt described a third va-

riety, parcifolia, for a form with small, coriaceous

leaves. Annotations in Schmidt's handwriting on

the type and other specimens read "parvifolia,"

indicating either a typographical error in publi-

cation or that he changed his mind. The former is

more likely since these plants have quite small,

but not few, leaves. This form is now considered

only a minor variant.

Brunfelsia brasiliensis is here treated as a sin-

gle polymorphic species, widely distributed in

southern Brazil, with one subspecies, macrocalyx,

occurring at higher elevations from the Serra do

Mar and Serra da Mantiqueira southward in the

mountains to Santa Catarina. Brunfelsia bras-

iliensis exhibits a wide range of morphological

variation throughout its area of distribution, which

consists mostly of mountainous, ecologically di-

versified terrain. In this region, primarily in the

states of Minas Gerais, Rio de Janeiro, and Sao

Paulo, there exist ample opportunities for geo-

graphical and ecological isolation of populations

with subsequent morphological differentiation.

Environmental factors also play an important

part in the phenotypic variability seen in this spe-

cies, and different forms may be interpreted as

ecotypes. For example, in exposed, well-drained

places such as the crest of a peak or rocky out-

crop, the plants are frequently low in stature with

horizontal branches, small, often coriaceous

leaves, and fewer and smaller flowers. Lower

down the same slope, in shaded, more protected,

and probably more fertile places, the plants may

attain a height of up to 3 m with well-developed

inflorescences and larger flowers and leaves.

The great morphological variation in B. bras-

iliensis occurs chiefly in the presence, type, and

location of indumentum, in the size and form of

the inflorescence, and to a lesser extent in the size

and form of the flowers and leaves.

An important diagnostic feature of this species

is the presence of a persistent peduncle, a short

stalk that articulates with the pedicel of each flow-

er. This structure is scarcely, if ever, developed in

any other species of Brunfelsia. Some populations

of B. uniflora, however, may have a similar stalk

in its solitary flowers. This "peduncle" is perhaps

more accurately described as an ultimate branch

of the inflorescence that immediately precedes a

flower. It commonly bears 1-3 small bracts near

the base, the axillary buds of which may produce

additional flowers, thereby proliferating the inflo-

rescence. The peduncle may be shorter or longer

than the pedicel or, more rarely, obsolete. The dis-

tal end is dilated into a short, flat torus that be-

comes conspicuous when the flower falls.

The inflorescence arises at the tips of the bran-

chlets and is usually much-branched and many-

flowered. It may be strictly terminal and compact

to lax and open with several subterminal branches

also developing into flowering axes. Alternatively,

the flowers may be reduced to a few, often sub-

sessile ones and rarely to a single flower in sub-

species macrocalyx. The calyx in B. brasiliensis

may be tubular to nearly campanulate in form and

varies greatly in size. The corolla tube may be

scarcely exserted from, or up to twice as long as,

the calyx.

The presence of a well-developed indument,

particularly in the inflorescence, is also diagnostic

in B. brasiliensis. The trichomes are nonglandular,

uniseriate, long or short, and straight or curved.

In dried material, they are frequently yellowish or

yellowish brown. Often the indument may be

densely villous to tomentose. More rarely it ap-

pears quite sparse and pilose, or the plants may

be glabrescent. The upper surface of the leaves is

frequently glabrescent, with short, subscabrous

hairs that break off near the base. The lower sur-

68

FIELDIANA: BOTANY

face of the leaves is usually pubescent, especially

at the nerves and midrib.

Brunfelsia brasiliensis is easily distinguished

from most other species of the genus by the reg-

ularly oblong-lanceolate leaves, the distinctive

yellowish brown pubescence, and the presence of

the short peduncle articulating with the pedicel.

Rare glabrous forms of B. brasiliensis (= B. ac-

uminata (Pohl) Benth.) resemble B. bonodora but

are distinguished by their stalked pedicels. Brun-

felsia brasiliensis appears to be most closely re-

lated to the B. bonodora-B. grandiflora complex

by the nature of the more or less branched inflo-

rescence and floral and fruit morphology.

Brunfelsia brasiliensis subsp. brasiliensis is

distributed in southeastern Brazil from Minas

Gerais south to Sao Paulo, between 700 and 1400

m elevation. The species shows a fairly wide eco-

logical tolerance but prefers well-drained sites. It

is frequently encountered in the mares de morros,

"the seas of hills," which occur in southeastern

Brazil, particularly in Minas Gerais. Here the

plant grows on granitic peaks and rocky outcrops,

often in dry, exposed habitats. It seems to occur

less commonly in the humid forests on the slopes

of the Serra do Mar. It also grows in association

with the campos, or open grassland communities,

and in capaes, small patches of woodland found

scattered in the campos. Brunfelsia brasiliensis is

also encountered in capoeiras, a general term for

cut-over, secondary woods, and is one of the few

brunfelsias that survive cutting of the primary for-

ests. Around Brasilia, it has been reported from

gallery forests. Brunfelsia brasiliensis is a freely

flowering plant with the flowers appearing pri-

marily from September to March. The fruits have

been collected from October to March. This spe-

cies has been cultivated in European conservato-

ries since its discovery, although it is much rarer

in cultivation than other species. To my knowl-

edge, it is not planted in the American tropics

outside of Brazil. Peckolt (1909) says that it is a

common garden plant in that country. Brunfelsia

brasiliensis is known as manacd, manacd da ser-

ra, erva de macaco, and manacd assu (Peckolt,

1909). From herbarium labels (Occhioni s.n., V.

Assis 102, 142) we know that it is considered poi-

sonous.

At higher elevations in the mountains of Rio de

Janeiro, Sao Paulo, and south to Santa Catarina,

there occurs a distinct subspecies of B. bras-

iliensis. This was first described by Dusen in 1905

as B. hopeana var. macrocalyx from a specimen

collected on Mount Itatiaia. This type superficial-

ly resembles B. uniflora (=B. hopeana) in having

the inflorescence reduced to a single flower, but

this is somewhat exceptional for the subspecies as

now circumscribed.

Brunfelsia brasiliensis subsp. macrocalyx dif-

fers from B. brasiliensis subsp. brasiliensis in

having fewer flowers per inflorescence (one to

eight), a very short peduncle in which the flowers

may appear sessile, a longer pedicel (5-15 mm)

that always exceeds the peduncle, and a usually

longer calyx (15-22 mm). The corolla tube of

subspecies macrocalyx is short relative to the ca-

lyx and is just barely exserted. The indumentum

is frequently less well-developed, especially in the

leaves, which may be glabrescent and bright

green. In addition, the trichomes may be glandu-

lar, especially in the calyx and pedicels, a char-

acter not normally encountered in subspecies

brasiliensis.

Morphologically intermediate populations are

known to exist between subspecies brasiliensis

and subspecies macrocalyx. In collections from

the Serra dos Orgaos and Alto Macahe in the state

of Rio de Janeiro, subspecies brasiliensis shows

a tendency to develop a large calyx (to 17 mm)

similar to that of subspecies macrocalyx. How-

ever, because these plants are, in general, larger

in all parts, they may be considered exceptionally

vigorous individuals perhaps growing in a partic-

ularly favorable place (cf. Glaziou 6908, 13478,

17169). Other collections (cf. Occhioni 1131)

from Mount Itatiaia are referable to subspecies

macrocalyx, although they exhibit a multiflorous

inflorescence. These plants may indeed suggest

putative hybridization between the two subspe-

cies; however, this cannot be proved at this stage

of knowledge of the group. Several collections

from the states of Parana and Santa Catarina have

been referred to B. brasiliensis subsp. macrocalyx

(e.g., Dusen 13562, Reitz 3985). These popula-

tions are somewhat heterogeneous and in some

characters resemble B. cuneifolia (see discussion

under that species), such as the indument and the

reduction in number of flowers, often to one. Fur-

thermore, the calyx tends to be smaller than in

more northern populations of this subspecies,

somewhat resembling that of subspecies bras-

iliensis.

Because of the complex pattern of variation

that is encountered in this region, I believe this

problem cannot be resolved completely without

additional materials and a knowledge of the plants

in the field.

Brunfelsia brasiliensis subsp. macrocalyx in the

PLOWMAN: REVISION OF BRUNFELSIA

69

northern part of its range occurs at altitudes be-

tween 1650 and 2400 m on relatively few moun-

tain peaks and high campos: Mount Itatiaia, Alto

Macahe, and Serra dos Orgaos in Rio de Janeiro;

and Alto da Serra, Serra de Bocaina, and Campos

do Jordao in Saao Paulo. In Sao Paulo (Campo

Grande) and southward, subspecies macrocalyx

occurs at somewhat lower elevations (800-900

m).

Ecologically, it grows in wet, montane cloud

forests in association with Podocarpus and also

occurs in moist, high-elevation campos. In the

Parque Nacional de Itatiaia, plants have been fre-

quently collected between Maromba and Maci-

eiras at 1600-1900 m elevation and are repre-

sented by two distinct forms. One of these (cf.

Plowman 2891, Sucre 5191) is sparsely branched

with dark leaves and pubescent calyx with erect

lobes. The other (cf. Plowman 2906, Sucre 5206),

which grows sympatrically a few hundred yards

away, is profusely branched with bright green,

glabrous leaves and a glabrous calyx with strong-

ly recurved lobes. Though strikingly different to

the casual observer, these plants must be regarded

as minor phenotypic variants of a single subspe-

cies. Analysis in the field of situations such as this

gives a measure of the kind of variability that is

present in a population and a better basis for un-

derstanding comparable problems where field

studies are not possible.

Efforts were made to rear these plants in the

greenhouse from seed to study the nature of the

characters and their inheritance. However, the

seedlings proved to be exceedingly sensitive to

fluctuations in temperature and humidity and soon

died.

11. Brunfelsia chiricaspi Plowman, Bot. Mus.

Leafl. 23(6): 255, t. 17. 1973. Type: CO-

LOMBIA. Putumayo: Umbria, 0°54'N,

76°10'W, 325 m, forest, shrub 1.5 m, flowers

sky blue, zanango, medicinal, Oct. -Nov.

1930, Klug 1810 (holotype, A; isotypes, F, s,

us 1456539). Figure 25.

Shrub or treelet 1-3 m tall. Trunk to 5 cm in

diameter near base. Bark thin, roughish, grayish

brown. Branches few, lax, spreading, naked.

Branchlets glabrous, light brown to ochraceous,

shiny, outer bark thin. Leaves scattered along

branchlets or somewhat crowded, blade 20-30 cm

long, 7-12 cm wide, elliptic to lanceolate, some-

times obovate, apically obtuse with short, subfal-

cate acumen or acuminate, cuneate to blunt at

base, glabrous, dull, dark green above, pale green

beneath, smooth, subcoriaceous, lateral nerves 8-

10, straight; petiole 5-10 mm long, short, stout,

glabrous, dark brown, roughish. Inflorescence

corymbiform, terminal or axillary, usually few-

flowered with 4-7(20) flowers, puberulent or gla-

brous. Bracts 1-2 mm long, lanceolate, concave,

glabrous. Flowers sky blue to violet, fading

white, with 5-angled white spot at mouth. Pedi-

cels 6-13 mm long, slender, glabrous. Calyx 10-

13 mm long, 4-6 mm in diameter, tubular-cam-

panulate, slightly inflated, glabrous, purplish to

pale green, teeth short, broadly triangular, acute

to blunt with short glandular acumen; calyx in

fruit coriaceous, striately nerved, dotted with len-

ticels. Corolla tube 22-25 mm long, 3 mm in di-

ameter, twice as long as calyx, straight, glabrous;

limb 25-30 mm across, thickening at mouth

prominent, fleshy, 5-angled, lobes subequal,

strongly deflexed at anthesis, the uppermost

slightly larger, rounded. Stamens included in up-

per part of corolla tube; filaments subligulate,

lower inner pair 2.5 mm long, upper outer pair

3.5 mm long, reaching the mouth; anthers to 1.5

mm long, orbicular-reniform, light brown. Ovary

2 mm long, ovoid-conical; style filamentous,

equaling filaments; stigma briefly bifid, obtuse,

the upper lobe somewhat larger. Capsule about

10 mm long, 8 mm in diameter, dry at maturity,

subglobose. Seeds few, 6 mm long, 2.5 mm in

diameter, ellipsoid-reniform, reticulate-pitted.

DISTRIBUTION — Colombia, Ecuador. See Figure

23.

SPECIMENS EXAMINED — COLOMBIA. Putu-

mayo: Orito, near Texaco drilling site, 11 Feb.

1972, Kennedy 1386 (GH); alrededores de Puerto

Limon, 300-400 m, 15 Feb. 1953, Mora 1040

(COL); El Whiskey at Finca "Santa Marta," 13 km

S. of Umbria near road, 400-500 m, 27 Nov.

1968, Plowman 2080 (ECON, GH); forest about 2

km SW of San Pedro, N of Puerto Asis, 400-500

m, 28 Nov. 1968, Plowman 2081 (ECON, GH); Rio

Guamues, San Antonio, 5 Sep. 1966, Pinkley 420

(ECON); Santa Rosa, 28 Nov. 1966, Pinkley 563

(ECON); trail between Santa Rosa and road to Hor-

miga, 2 Dec. 1968, Plowman 2092 (ECON, GH);

*Rio Putumayo, Puerto Ospina, Schultes 3422

(GH). ECUADOR. Napo: * Along Rio Aguarico

upstream from Santa Cecilia, Duke 16047 (MO);

*Coca (Puerto Francisco de Orellana), S side of

Rio Napo, Marling & Andersson 11771 (GB);

*Coca (Puerto Francisco de Orellana), Marling &

Andersson 11804 (GB, GH); *Hacienda Cotapino

70

FIELDIANA: BOTANY

BRUNFELSIA chiricaspi Plowman

FIG. 25. Brunfelsia chiricaspi. Reproduced courtesy of the Botanical Museum of Harvard University.

(Conception), Marling et al. 7021 (GB, GH); *Path

Cotapino (Conception), Rio Bueno, Marling et al.

7151 (GB, GH); *Coca de Orellana, lower Rio Pay-

amino, Marling et al. 7772 (GB, GH); *Hacienda

San Carlos at Rio Napo, ca. 15 km E of Coca

(Puerto Francisco de Orellana), Lugo 2724 (GB,

GH); *Tierra Colorada, ca. 3 km N of Coca (Puerto

Francisco de Orellana), Lugo 2942 (GB, GH); *Ca-

non de los Monos, road Coca (Puerto Francisco

de Orellana), Lago Agrio, ca. 12 km N of Coca,

PLOWMAN: REVISION OF BRUNFELSIA

71

Lugo 2981 (GB, GH); *road from Lago Agrio-E 1

Chaco, Lugo 3476 (GB, GH); *7 km N of Coca;

Estacion Experimental de INIAP-Payamino, Zar-

uma et al. 56 (F).

Brunfelsia chiricaspi was originally known

only from a few collections from a small area of

southwestern Colombia on the eastern flank of the

Cordillera Occidental. [Plowman predicted that

the distribution would prove to be much greater,

and this species is now also known from many

recent collections in adjacent Ecuador (see

above). — Eds.] It grows as an understory shrub in

humid, primary forests at elevations of 325-500

m. The fruit is as yet imperfectly known because

of a lack of material.

Brunfelsia chiricaspi is related to B. grandiflora,

which is widespread in the western Amazon, and

to B. mire of Bolivia and adjacent Peru and Brazil.

Brunfelsia chiricaspi differs from B. grandiflora in

having larger leaves, longer pedicels (6-13 vs. 2-

6 mm) and usually fewer flowers. It differs from

B. mire in the basally obtuse, elliptic to lanceolate

leaves; the few-flowered cyme; and smaller flowers

(corolla tube 22-25 vs. 25-38 mm long).

Known generally as chiricaspi in the Colom-

bian Putumayo, this species is used as a strong

medicine and narcotic by the Kofan Indians (see

Plowman, 1977).

12. Brunfelsia cuneifolia J. A. Schmidt in Mart.,

Fl. Bras. 8(1): 259. 1864. Type: BRAZIL. In

Brasilia australiore, Sellow 4016 (lectotype,

designated here, destroyed at Berlin; photo-

graph of lectotype, F 621824, NY, us; isolec-

totypes, F, M).13 Riedel 1467, also cited by

Schmidt, is excluded as a type. It belongs to

B. obovata Benth. var. obovata. Figure 26.

Shrub to about 1 m tall, much branched.

Branches terete, naked. Bark grayish to yellow-

ish brown. Branchlets short, leafy, sparsely to

densely villous, rarely glabrescent, hairs often yel-

lowish to reddish brown, sometimes glandular.

13 It is clear that Plowman was intending to lectotypify

this species using the destroyed B sheet as good photo-

graphs of it exist. He appears to specifically not consider

the unnumbered Sellow sheets at F and M as good can-

didates for lectotypes (in the strict sense). — Eds.

Leaves scattered or crowded near the tips of the

branchlets, blade 3.5-10 cm long, 1.8-3 cm wide,

oblong-obovate to elliptic-lanceolate, cuneate to

blunt at apex, rarely acuminate, narrowed to

broadly cuneate at base, the upper surface gla-

brescent or with long, scattered hairs, the lower

surface velutinous, often densely so at midrib and

nerves, rarely glandular or glabrescent, somewhat

shiny above, dark green, pale green beneath, firm-

ly membranaceous to subcoriaceous, lateral

nerves 4-7; petiole 2-4 mm long, velutinous or

glandular. Inflorescence terminal on tip of

branchlets, sessile, with 1-3 flowers. Bracts 2-8

mm long, small, lanceolate, glandular, ciliolate,

caducous. Flowers violet fading to white with

age. Pedicel 3-5 mm long, slender, thickened at

apex, velutinous or glandular pubescent. Calyx

12-18 mm long, 4-6 mm in diameter, tubular, in-

flated, 5-angled, drying plicate, pubescent with

short glandular hairs, especially near base, or vil-

lous, pale green, membranaceous, teeth 3-7 mm

long, ovate-lanceolate, acute to acuminate at apex.

Corolla tube 18-28 mm long, 1.5-2 mm in di-

ameter, \Vi times as long as calyx, pentagonal at

base, pilose or minutely glandular in upper part;

limb 25-37 mm in diameter, spreading, lobes 10-

15 mm long, undulate, broadly rounded. Stamens

inserted in upper part of tube; upper pair of fila-

ments 3-4 mm long, suberect, exserted slightly

from mouth; lower pair 2-3 mm long, included,

curved at apex; anthers 1.5 mm long, orbicular-

reniform. Ovary 2 mm long, conical-ovoid, style

about 23 mm long, filamentous; stigma 1 mm

long, briefly bifid. Fruit and seed unknown.

DISTRIBUTION — Brazil (Parana, Rio Grande do

Sul, Santa Catarina, Sao Paulo). See Figure 28.

SPECIMENS EXAMINED — BRAZIL. Parana:

Mun. Mangeirinha, Lageado San Antonio, 20 Oct.

1966, Hatschbach 15159 (US); *Municipio de

Guarapuara; Guarapuara, Frenzel 666 (MBM); *In-

acio Martins, Hatschbach & Guimardes 30302

(MBM); *Municipio Campo Largo; Serra do Pu-

runa, Kummrow 2432 (F). Rio Grande do Sul:

Farroupilha, 15 Oct. 1957, Camargo 2123 (B), 3

Oct. 1957, Camargo s.n. (s); *Municipio Cacequi,

Dobereiner & Tokarnia 732 (RB); *Esmeralda;

Estacion Ecologica de Aracuri, Jarenkow 132 (F).

Santa Catarina: Mun. Lajes, Passo do Socorro,

FlG. 26. Brunfelsia cuneifolia.

72

FIELDIANA: BOTANY

I

\

N? 1995963

FIELD MUSEUM

OF

NATURAL HISTORY

~t»

PREFEITURA MUNICIPAL DE CURITIBA

MUSEU BOTANICO MUNICIPAL

i io li -a ZT- A-

icuaa Sta. Clara d

50573

Arbuato 2 B| flor lllas. Interior aata te gal«ria.

PMC • Off 004

PLOWMAN: REVISION OF BRUNFELSIA

73

900 m, 31 Oct. 1963, Klein 4343 (us); Mun. Novo

Horizonte, Lauro Miiller, 400 m, 24 Oct. 1958,

Reitz & Klein 4076 (NY, us); Campos Novos, 12

Sep. 1963, Reitz & Klein 16158 (GH, us); Mun.

Campos Alegre, lower fazenda of Ernesto Schei-

de, Campo Alegre, 900 m, 11 Dec. 1956, Smith

& Klein 8567 (us); Mun. Chapeco, Guatambu, ca.

27°06'S, 52°45'W, 350-400 m, Smith & Reitz

12541 (GH, R, NY, us); Mun. Irani, Campo de Irani,

ca. 26°57'S, 51°50'W, 700-900 m, 8 Nov. 1964,

Smith & Klein 13044 (GH, MO, p, R, uc, us). Sao

Paulo: Ilha do Cardoso, perto de Cananeia,

Prance et al 6933 (F).

In his original description of B. cuneifolia in

Martius's Flora Brasiliensis (1864), Schmidt cited

two syntypes. One of these, Riedel 1467, is clear-

ly B. obovata of Bentham. The other, which now

serves as the type, was collected by Sellow in

southern Brazil, but no specific locality was giv-

en. The specimen cited by Schmidt (Sellow 4016)

was destroyed at the Berlin herbarium; two un-

numbered Sellow duplicates, at Chicago (F) and

Munich (M), which conform in every way with

the photograph of the type, may be considered

isolectotypes.

Known from only a few collections, the avail-

able material of B. cuneifolia shows a consider-

able amount of variation in the form of the leaves

and type and amount of indument. Brunfelsia cu-

neifolia is distinguished from other species of

southern Brazil by the apically cuneate leaves and

a five-angled tubular calyx that appears plicate

when pressed and dried. The leaves and calyces

are always pubescent with either of two types of

trichomes present. The type collection and mate-

rial from Rio Grande do Sul have short glandular

hairs. Most specimens from Santa Catarina have

a dense villous to velutinous indument consisting

of longer, nonglandular hairs. The form of the

leaves and calyx associated with both types is

fairly uniform.

Brunfelsia cuneifolia is not easily distinguished

from other species of southernmost Brazil. It su-

perficially resembles B. pilosa, which is sympatric

with B. cuneifolia. Brunfelsia pilosa, however,

usually has a more pronounced acumen in the leaf

and a terete, tubular calyx that is purple and bears

only a few, long pilose hairs. In addition, the

flowers of B. pilosa are somewhat larger than

those of B. cuneifolia.

The villous pubescence of B. cuneifolia is very

reminiscent of the yellowish indument character-

istic of B. brasiliensis. Populations interpreted

here as B. brasiliensis subsp. macrocalyx extend

as far south as Santa Catarina and overlap with

the range of B. cuneifolia. Collections of these

two species in this region are morphologically

very similar and may be easily confused. Brun-

felsia brasiliensis subsp. macrocalyx may be sep-

arated by the presence of the persistent peduncle,

which articulates with the pedicel of the flowers,

and by the terete, campanulate or tubular-campan-

ulate calyces. This species usually has longer ped-

icels and shorter calyx teeth than B. cuneifolia.

Nevertheless, the intermediate nature of these

populations indicates that some hybridization may

have occurred between these two species and pos-

sibly involved B. pilosa as well.

The true affinities of B. cuneifolia appear to lie

with B. obovata var. obovata, which occurs far-

ther north in Sao Paulo and Minas Gerais [but see

also B. boliviana. — Eds.]. This plant resembles B.

cuneifolia in the tubular pentagonal calyx, which

also nearly equals the corolla tube in length, and

in the obovate leaves. In both species the upper-

most pair of anthers may be briefly exserted from

the corolla tube, a character that is unusual in sec-

tion Franciscea. Brunfelsia cuneifolia differs in

having the leaves cuneate at the apex, fewer flow-

ers per inflorescence, and a pubescent calyx that

is much more slender than in B. obovata (4-6 vs.

6-10 mm). These two species also differ striking-

ly in their ecological preference. Brunfelsia obov-

ata is a plant of marshes and flooded areas,

whereas B. cuneifolia occurs in uplands.

Brunfelsia cuneifolia is a relatively low shrub

that is found in habitats apparently similar to

those of B. pilosa. It is specifically reported from

Araucaria forests (arau-carieto) and also from

gallery forest and capoeiras, or secondary forest,

and between 400 and 900 m altitude. It flowers in

October and November and is known by the com-

mon name flor de Trovoada, or "thunderstorm

flower."

13. Brunfelsia dwyeri D'Arcy, Ann. Missouri

Bot. Gard. 57:259. 1971 ("1970"). Type:

PANAMA. Panama: Cerro Jefe in "Cleusia"

FIG. 27. Brunfelsia dwyeri.

74

FIELDIANA: BOTANY

1709790

FIELD MUSEUM

OF

NATURAL HISTORY

PLANTS OF PANAMA

Brunfelsia dwyeri D'Arcy

det. by R.L. Wilbur, 1971.

Provincia de Panama': slopes of Cerro Jefe

beyond Cerro Azul between 4-8 miles in

mostly heavily wooded slopes.

Common shrub 2-5 m. tall. Corolla deep

violet blue becoming paler with age.

R.L. Wilbur

R.E. Weaver

#11358

25 January 1970

PLOWMAN: REVISION OF BRUNFELSIA

15

(sic) forest, 2700-3000 ft, shrub, flowers pur-

ple to whitish purple, 27 Jan. 1966, Tyson,

Dwyer & Blum 3312 (holotype, MO 1820936;

isotype, MO 1917560). Duke & Dwyer 13933

(MO) is excluded from the cited paratypes.

Figure 27.

Shrub or small tree 2-10 m tall. Trunk to 6-

8 cm in diameter, branched from near the base.

Bark gray to grayish brown. Branches terete or

somewhat angular, naked. Branchlets leafy.

Leaves with blade 5-11 cm long, 2.5-5 cm wide,

obovate, elliptic or oblong-lanceolate, acuminate

at apex, slightly revolute at margin, cuneate or

narrowed at base, glabrous, shiny and light to

very dark green above, pale green beneath, cori-

aceous, lateral nerves 5-7, veinlets reticulate,

prominulous, especially on lower surface; petiole

3-10 mm long, becoming transversely corrugate-

rugose. Flowers solitary, terminal or axillary in

the upper leaf axils, purple, fading to white with

age, with irregular white spot at mouth. Peduncle

1-5 mm long, sometimes obsolete. Bracts 1-3

mm long, linear to ovate, puberulent, ciliate-mar-

gined, caducous. Pedicel 5-15 mm long, about 2

mm in diameter, glabrous. Calyx 12-22 mm long,

5-10 mm in diameter, tubular to tubular-campan-

ulate, inflated or not so, glabrous, subcoriaceous,

reticulately veined, yellowish green or purplish,

teeth incumbent or erect, triangular to ovate-lan-

ceolate, at apex acute or blunt; calyx in fruit co-

riaceous, often becoming dotted with lenticels,

verrucose near base, splitting completely along

one or two sinuses. Corolla tube 25-40 mm long,

2-4 mm in diameter, 2-3 times as long as calyx,

glabrous; limb 30-80 mm in diameter, spreading,

lobes 10-20 mm long, broadly rounded, the upper

lobe slightly larger. Stamens inserted in upper

part of corolla tube; filaments slender, terete, up-

per pair about 3 mm long, lower pair about 5 mm

long; anthers included within corolla tube, 2 mm

in diameter, orbicular-reniform. Ovary about 2

mm long, conical, ovoid, weakly 4-angled; style

filiform; stigma briefly bifid, the lobes subequal,

blunt. Capsule 1.5-2.5 cm long, 1-1.5 cm in di-

ameter, globose to ovoid, usually 2-celled, at ma-

turity often appearing 4-lobed with 4 prominent

cruciform grooves, pericarp to 4 mm thick, tough

and leathery, drying hard, surface smooth and

green at first, becoming rugulose-verrucose, light

brown. Seeds 3-15, 8-10 mm long, 3-4 mm in

diameter, oblong-ovoid, prismatic, dark reddish

brown, reticulate-pitted. Embryo straight, coty-

ledons ovate.

DISTRIBUTION — Panama, known only from Cer-

ro Jefe, Provincia de Panama. See Figure 28.

SPECIMENS EXAMINED — PANAMA. Panama:

Cerro Jefe, beyond Goofy Lake, 16 Nov. 1967,

Correa & Dressier 471 (MO); along main road

before turnoff to summit, 6 Jan. 1971, Croat

13136 (GH, MO); Cerro Jefe, 8 Apr. 1970, D'Arcy,

Dressier & Correa 3948 (MO), D'Arcy 3959 (MO),

21 Jan. 1967, Duke 9429 (MO), 2 Apr. 1969,

Dwyer, Durkee & Castillon 5051 (MO), 12 Mar.

1968, Dwyer 7280 (GH, MO), ca. 2,900 ft, 29 Jul.

1967, Dwyer & Ganger 7349 (GH, MO); along road

from Cerro Jefe to La Eneida (Cerro Azul), Croat

13072 (GH, MO), 12 Feb. 1971, Escobar 359 (COL,

ECON, GH), 26 Jul. 1969, Kennedy 316 (ECON), 12

Apr. 1972, Plowman 3160 (COL, ECON, GH, MICH,

uc), 9 Jul. 1966, Tyson, Dwyer & Blum 4335

(MO), 2300-2700 ft, 19 Jan. 1969, Tyson 5315

(MO), 25 Jan. 1970, Wilbur & Weaver 11358 (GH).

Brunfelsia dwyeri was discovered and de-

scribed during studies on the flora of Panama. So

far as is known, it occurs only on Cerro Jefe in

central Panama between 700 and 1,000 m eleva-

tion, an area now famous for its high floristic en-

demism (Lewis, 1971). The collection Duke &

Dwyer 13933 (det. W. G. D'Arcy, 1973) from

Cerro Pilon in Provincia Cocle, which was cited

in the original description as B. dwyeri, is actually

Schultesianthus megalandrus (Dunal) Hunz.

The thick-walled fruit and shiny, coriaceous

leaves are distinctive features of B. dwyeri. It ap-

pears to be most closely related to B. macrocarpa

of coastal Colombia and Ecuador, the flowers of

which are similarly large and solitary and which

also possesses a thick-walled fruit. The two spe-

cies are unique in section Franciscea in possess-

ing a fleshy pericarp. The resemblance of B.

dwyeri to B. lactea of Puerto Rico, pointed out by

D'Arcy (1971), is indeed "superficial" because

these species are placed in separate sections of the

genus and have no close affinities with one an-

other.

Brunfelsia dwyeri is a beautiful shrub with at-

tractive, shiny foliage and large, showy flowers.

It is locally common in the cloud forests on the

upper slopes of Cerro Jefe, occurring in secondary

woods with seasonal rainfall. This locality and

species represent the northernmost extension of

Brunfelsia sect. Franciscea. As pointed out by

Lewis (1971), the entire forest of Cerro Jefe and

surrounding areas is rapidly being destroyed for

raising poultry and cattle. Brunfelsia dwyeri and

76

FIELDIANA: BOTANY

FIG. 28.

square).

Distribution of Brunfelsia cuneifolia (solid triangle), B. dwyeri (solid circle), and B. grandiflora (solid

many other endemic species of plants are in grave

danger of extinction.

This species flowers and fruits intermittently

throughout the year, although I failed to find flow-

ers in mid-April. Local people reported that the

bark and roots of this plant are used for a "rem-

edy."

14. Brunfelsia grandiflora D. Don, Edinburgh

New Philos. J.: 86. Jul. 1829. Type: PERU.

San Martin: Uchiza, 1798, Tafalla s.n. (ho-

lotype, BM; isotype, MA). Figure 29.

Franciscea grandiflora (D. Don) Miers, Ann. Mag.

Nat. Hist., ser. 2, 5:250. 1850.

Shrub or small tree 1-6(10) m tall. Trunk to

7 cm in diameter near base, much branched. Bark

thin, roughish, light to dark brown. Branches

slender, ascending or spreading, often subvirgate

and arching, leafy, glabrous. Branchlets glabrous,

rarely pubescent, green. Leaves more or less two-

ranked, scattered along branchlets, blade 6-23 cm

long, 2-8 cm wide, lanceolate to oblong, often

arched with ascending laminas, apically long to

short acuminate, the acumen subfalcate, cuneate

to narrowed at base, glabrous or sparingly pubes-

cent beneath at midrib, dull, dark green above,

paler green beneath, rarely glaucescent or nitid,

firmly membranaceous to subcoriaceous; petiole

2-6 mm long, glabrous or minutely pubescent, be-

coming corky, transversely corrugate with age.

Inflorescence terminal and subterminal, simple or

branched, dense or lax, the axis 5-45 mm long.

Flowers 5-many, showy, scentless, violet fading

to white with age, with rounded, white ring at

PLOWMAN: REVISION OF BRUNFELSIA

77

mouth. Bracts 1-3 per flower, 1-5(10) mm long,

lanceolate to ovate, ciliolate, pubescent or glabra-

te, caducous. Pedicel 2-5(10) mm long, glabrous

or with few sparse glandular hairs, becoming

thicker and corky- verrucose in fruit. Calyx 5-13

mm long, 3-7 mm in diameter, tubular-campanu-

late, somewhat narrowed toward base, globose to

obovate in bud, somewhat inflated or not so, gla-

brous, rarely punctate or sparsely glandular with-

in, smooth or striate-nerved, light yellow-green to

gray-green, firmly membranaceous to subcoria-

ceous, teeth 1-5 mm long, triangular to ovate,

blunt to short acuminate, erect or incumbent, re-

curved slightly with age; calyx in fruit persistent,

coriaceous, becoming corky-verrucose especially

near base, often splitting on one or more sides.

Corolla tube 15-40 mm long, 1.5-3 mm in di-

ameter, 2-5 times as long as calyx, glabrous, pale

violet to white; limb 20-50 mm in diameter,

spreading, violet fading to white with age, lobes

7-15 mm long, subequal or uppermost somewhat

larger, rounded or emarginate at apex, overlapping

at sides, abruptly narrowed at base. Stamens

completely included in upper part of corolla

tube; filaments thin, upper pair 4 mm long, lower

pair 3 mm long, white; anthers 1-1.5 mm long,

orbicular-reniform, light brown. Ovary 1.5-2

mm long, sessile, conical to ovoid, pale yellow;

style slender, slightly dilated at apex; stigma

about 1 mm long, briefly bifid, unequal, the up-

per lobe somewhat larger, obtuse, green. Fruit

8-20 mm long, 8-20 mm in diameter, ovoid to

subglobose, obtuse or apiculate at apex, smooth,

nitid, dark green turning brownish, with corky-

punctate or -verrucose outgrowths, pericarp thin,

0.3 mm thick, crustaceous, drying brittle, tardily

dehiscent. Seeds 10-20, 5-7 mm long, 2-3 mm

in diameter, variable in shape, ellipsoid to ob-

long, angular, dark reddish brown, reticulate-pit-

ted. Embryo about 4 mm long; slightly curved,

cotyledons 1.5 mm long, ovate-elliptic.

DISTRIBUTION — Colombia, Venezuela, Ecuador,

Peru, Brazil, Bolivia. See Figure 28.

Key to the Subspecies of Brunfelsia grandiflora

1. Corolla tube 30-45 mm long, 2-3.5 mm in

diameter, opening at mouth linear-oblong; limb

35-42 mm in diameter, capsule 17-22 mm

long 14a. subsp. grandiflora

1. Corolla tube 15-30 mm long, 1-2 mm in di-

ameter, opening at mouth elliptic-obovate;

limb 20-40 mm in diameter; capsule 11-16

mm long 14b. subsp. schultesii

14a. Brunfelsia grandiflora subsp. grandiflora

Brunfelsia calycina f. grandiflora (D. Don) Voss in

Vilm., Blumengartner, ed. 3, 1: 743. 1896.

Brunfelsia tastevinii Benoist, Bull. Soc. Bot. France

75: 295. 1928. Type: BRAZIL. Acre: region de

Rio Jordao, Tastevin s.n. (holotype, p; isotypes,

F, P).

Leaves 10-23 cm long, 3-8 cm wide, glabrous

or sparingly pubescent at midrib; petiole 3-12

mm long. Pedicel 2-10 mm long. Calyx 9-13

mm long, teeth 2-5 mm long, ovate-lanceolate.

Corolla tube 30-45 mm long, 2.5-3.5 mm in di-

ameter, the mouth 6-9 mm long; limb 35-52 mm

in diameter, spreading. Capsule 17-22 mm long.

DISTRIBUTION — Colombia, Ecuador, Peru, Bra-

zil, Bolivia.

SPECIMENS EXAMINED — COLOMBIA. Meta:

*25 km S from Granada toward San Juan, Luteyn

el al. 4769 (F, MO). ECUADOR. Loja: *City of

Loja, Zarucchi & Andrade 2316 (F). Napo: Mera,

4 Mar. 1940, Lugo 23 (s); Canton Napo, near

Achidona, 650 m, 19 Apr. 1935, Mexia 7275 (F,

s, uc, us); *Costado S del Rio Napo, E de Puerto

Napo, por carretera nueva hacia Rio Arajuno,

Neill 6546 (F); *Rio Napo 8 km abajo de Puerto

Misahualli; Reserva Biologica Jatun Sacha, Neill

& Palacios 7101 (F). Zamora-Chinchipe: Reg.

Oriental, Huamboya, 1500-1600 m, 15 Feb.

1944, Acosta-Solis 7576 (F); Camino Gualaquiza

al Rio Zamora (Las Jibarias), 850 m, 25 Dec.

1948, Scolnik 1495 (us); *Zamora-Zumba (along

Rio Jamboe), Marling & Andersson 13906 (GB,

GH); *Guadalupe-San Jose de Yacuambi (28 de

Mayo), along Rio Yacuambi, Harling & Anders-

son 13934 (GB, GH). PERU. Cusco. *Prov. Pau-

cartambo; Callanga, Woytkowksi 514 (F). Huan-

uco: woods above Pozuzo, 6000 ft, Aug. 1863,

Pearce 290 (K); Monzon, 950 m, 17 Jun. 1958,

Woytkowski 5026 (F, LE, P, s); Divisoria, 1700 m,

16 Sep. 1946, Woytkowski 34536 (F, MO, uc);

*Prov. Tingo Maria; Rio Huallaga, Boeke 1224

(F). Loreto: Pongo de Manseriche, Colon, 1924,

Tessmann 5541 (o, s); Rio Ucayali, Canchahuayo,

20 Nov. 1898, Huber 1490 (MG); Santa Catalina,

11 Dec. 1898, Huber 1562 (MG); Prov. Coronel

Portillo, Plantacion Margarita, cerca a la Diviso-

ria, 1500-1600 m, 14 Aug. 1946, Ferreyra 985

(B, M, us). San Martin: Rio Huallaga, Rio Chi-

purana, Tarapoto, 6 Dec. 1898, Huber 1556 (MG);

78

FIELDIANA: BOTANY

BRUNFELSIA grandiflora D.Don

subsp.

grandiflora ;

subsp. sell ul te si i

FIG. 29. Brunfelsia grandiflora subsp. grandiflora and subsp. schultesii. Reproduced courtesy of the Botanical

Museum of Harvard University.

PLOWMAN: REVISION OF BRUNFELSIA

79

*Prov. Mariscal Caceres; Dist. Campanilla, Rio

Sion, Caserio de Sion, Schunke, V. 3480 (NY);

*Prov. Mariscal Caceres; Dist. Tocache Nuevo,

Quebrada de Huaquisha, Schunke, V. 4051 (NY);

*Prov. Mariscal Caceres; Dist. Tocache Nuevo,

Quebrada de Huaquisha, Schunke, V. 4034 (NY).

Without department: E of the Andes, 4000-

6000 ft, Jul. 1884, Pearce 295 (K); woods near

Rio Bambusi, one of the tributaries of Maranon,

Mar. 1868, Pearce s.n. (K); in flum. Maranon ripis

inundatis, Apr. 1855, Spruce 3988 (K); 1909-

1914, Weberbauer 6151 (F). BRAZIL. Acre: ba-

sin of Rio Purus near mouth of Rio Macauhan

(tributary of Rio Yacu), lat. 9°20' S, long. 69°W,

3 Aug. 1933, Krukoff 5284 (F, G, GH, K, LE, M,

MICH, MO, NY, RB, s, sp, uc, us); Rio Moa, 10 km

above and below Maita, 16 Apr. 1971, Prance et

al. 11998 (wis); *Municipio Rio Branco; Rio

Branco-Porto Velho Highway, Albuquerque et al.

1269 (F); *Municipio Rio Branco; Seringal Cor-

redeira, Colocac. ao Terra Alta, H. C. de Lima et al.

2122 (F); *Cruzeiro do Sul, Rio Jurua and Rio

Moa, Maas et al. 12963 (MO); *Cruzeiro do Sul,

Rosa 622 (MO); *Rio Branco, Rod. Rio Branco-

Porto Velho, J.U. Santos et al. 48 (F). Amazonas:

Rio Purus, Monte Verde, 18 Apr. 1904, Huber

4575 (MG), 29 Apr. 1904, Huber 4589 (MG); Rio

Purus-Rio Ituxi, opposite Labrea, 23 Jun. 1971,

Prance et al. 13471 (wis); Serra da Moa near

school, 1 May 1971, Maas et al. P 12709 (wis);

*region of Rio Jordao, Tastevin s.n. (F, P). With-

out locality: 1840-1848, Lobb 65 (K). BOLIVIA.

Beni: Tumi Chucua 30 km S of Riberalta along

Rio Beni, Solomon 7609 (F); *5 km SW of Ri-

beralta, Solomon 7905 (F). La Paz: *Prov. Nor

Yungas: 12 km hacia el Literal, W del Rio Beni,

Beck 13335 (F).

CULTIVATED SPECIMENS OF INTEREST — ECUA-

DOR. Napo-Pastaza: Rio Aguarico, Dureno, 27

Dec. 1965, Pinkley 43 (ECON), 24 May 1966,

Pinkley 202 (ECON); Tena, 28 Sep. 1966, Pinkley

460 (ECON). PERU. Loreto: Maynas, Iquitos,

1924, Tessmann 3541 (G, s). San Martin: Mon-

son River, 6 km from Tingo Maria, 625-1100 m,

30 Oct. 1949-19 Feb. 1950, Allard 21176 (F, us).

BRAZIL. Para: Belem, Jardim Botanico do Mu-

seu Goeldi, Apr. 1908, BAKER 61 (B, c, G, GH, K,

LE, M, NY, us). Rio de Janeiro: *Municipio Rio

de Janeiro; Jardim Botanico do Rio de Janeiro,

Ferreira & Costa 16 (F).

14b. Brunfelsia grandiflora subsp. schultesii

Plowman, Bot. Mus. Lean*. 23(6):259, t. 18.

1973. Type: COLOMBIA. Amazonas: Rio

Amazonas, near mouth of Rio Loretoyacu

and Puerto Narino, bush 12 ft, flowers white

or purple, poisonous to cattle, sanango, very

positive for alkaloid, 13-15 Sep. 1966, Schul-

tes, Raffauf & Soejarto 24108 (holotype, GH;

isotype, COL).

Leaves 6-20 cm long, 2-8 cm wide, lanceolate,

glabrous; petiole 2-6 mm long. Inflorescence

variable, compact or lax. Pedicel 2-6 mm long.

Calyx 5-10 mm long, teeth 1-3 mm long, trian-

gular to triangular-ovate. Corolla tube 15-30 mm

long, 1-2 mm in diameter, curved toward apex;

limb 20-40 mm in diameter, spreading, mouth 3-

5 mm long. Capsule 11-16 mm long, 10-16 mm

in diameter.

DISTRIBUTION — Colombia, Venezuela, Ecuador,

Peru, Bolivia, Brazil.

SPECIMENS EXAMINED — COLOMBIA. Amazo-

nas: Leticia, about 1 km NE of town on road to

brick factory, 30 Jan. 1969, Plowman & Kennedy

2310 (ECON); Rio Amazonas, forest along river

about 2 km downriver from Puerto Narino near

mouth of Rio Loretoyacu, 5 Feb. 1969, Plowman

et al. 2404 (GH, INPA, uc); *Rio Amazonas, down-

river from Puerto Narino near mouth of Rio Lor-

etoyacu, Plowman et al. 2407 (F). Caqueta: Rio

Orteguaza, La Maria (Hacienda de Ramon

Achiardy), 23 km SE of Venecia, 400 m, 7 Mar.

1944, Hermann 11237 (NY); La Rastra, 22 Apr.

1956, Vogel 23 (us); Caqueta, Dec. 1930, Uribe

Uribe P. s.n. (us); Florencia, 420 m, Feb. 1930,

Perez-Arbelaez 688 (COL, us). Meta: llanos of

San Martin, 30 Sep. 1916, Da\ve 241 (K, us); San

Martin, 400 m, 2 Aug. 1946, Uribe Uribe 1353

(COL). Putumayo: selva higrofila del Rio Putu-

mayo, Puerto Porvenir, arriba de Puerto Ospina,

hacia la Loma, 230-250 m, 17 Nov. 1940, Cua-

trecasas 10611 (COL); Puerto Ospina, 230 m, 25

Nov. 1940, Cuatrecasas 10795 (COL, us); Mocoa,

bosque higrofilo en la Quebrada del Rio Mulato,

570-600 m, 26 Dec. 1940, Cuatrecasas 11275

(COL, F); camino entre Mocoa y Puerto Asis, 500

m, 21 Jun. 1935, Garcia Barriga 4586 (COL, us),

500-700 m, 28 Aug. 1963, Juajibioy Chindoy 277

(ECON); Pepino, 20 Nov. 1968, Plowman 2039

(ECON), 1800-2400 ft, 16 Mar. 1953, Schultes &

Cabrera 19115 (GH); Rio Caqueta, Floresta, about

15 km downriver from Puerto Limon, 20 Dec.

1968, Plowman 2183 (ECON); Uchupayaco, en la

planada entre Urcusique en las orillas del Rio

Uchupayaco, 300 m, 22 Feb. 1942, Schultes 3340

(GH); San Antonio del Rio Guamues, 7 Sep. 1966,

80

FIELDIANA: BOTANY

Pinkley 444 (ECON); along trail between San An-

tonio and Santa Rosa, 1 Dec. 1968, Plowman

2090 (ECON). VENEZUELA. Barinas: Barinitas,

Mar. 1953, Aristeguieta 1669 (NY, YEN); Parque

Barinitas, 600 m, 22 Jun. 1956, Bernardi 3310

(G). ECUADOR. Napo: Rocafuerte, high forest

of Hacienda Arcadia, Rio Curaray, 200 m, 7 July

1933, Heinrichs 496 (F, G), Rio Aguarico, Dureno,

8 Dec. 1966, Pinkley 575 (ECON); Confluence of

Quiwado and Tiwaneo Rivers, Davis & Yost 958

(F); *Limoncocha on Rio Napo, Drummond 7330

(MO); *Nuevo Rocafuerte, Rio Yasuni, Marling et

al. 7474 (GB, GH); *Rfo Aguarico, Dureno (30 km

E of Santa Cecilia), Marling et al. 7700 (GB, GH);

*Canon de Los Monos, ca. 12 km N of Coca

(Puerto Francisco de Orellana), Marling & An-

dersson 11724 (GB); *environs of Limoncocha,

Madison et al 5390 (F); *Archidona, Pinkley 457

(F). Pasta/a: *basin of Rio Pastaza, Pacayacu-

Sarayacu region, Gill 45 (NY). PERU. Cuzco:

Prov. Paucartambo, Villa Carmen, 720 m, 1 May

1965, Vargas 16301 (us). Huanuco: *Prov. Leon-

cio Prado; Tingo Maria, Castillo, Plowman 5825

(F); *Prov. Pachitea; Dist. Honoria, Bosque Na-

cional de Iparia, Schunke V. 1480 (NY). Loreto:

Prov. Maynas, Iquitos region, ca. al Cementerio

de Iquitos, 120 m, 18 Sep. 1964, Dodson 2822

(MO); Mishuyacu near Iquitos, 100 m, Jan. 1930,

Klug 741 (us); Rancho Indiana, 110 m, 24 Jan.

1932, Mexia 6444 (BM, F, G, GH, K, MO, NY, PH, s,

uc, us); Rio Nanay, upland forest near Samito, 19

Feb. 1969, Plowman 2494 (ECON, GH); Rio Mazan,

Gamitanacocha, 100-125 m, 12 Jan. 1935, Schun-

ke 8 (A, F, NY, uc, us); Masisea, 275 m, Killip &

Smith 26844 (us); Yurimaguas, lower Rio Hual-

laga, 135 m, Aug.-Sep. 1929, Killip & Smith

27667 (us), 25056 (NY), 281 78 (us), Jan. 1831,

Poeppig 2206 (w); Balsapuerto (lower Huallaga

Basin), 150-350 m, 28-30 Aug. 1929, Killip &

Smith 28644 (us), 220 m, Feb. 1933, Klug 2880

(A, F, G, GH, MO, NY, s); Pucallpa, 200 m, 4 Aug.

1946, Soukup 3041 (F); Pongo de Manseriche,

1924, Tessmann 4921, Rio Maranon, Tessman

5005 (G); Prov. Ucayali, lower Pisque, Nueva Be-

liza, 1923, Tessmann 3243 (G, NY, s); *Prov. May-

nas; Moena Cano between Iquitos and Rio Itaya,

Gentry et al. 15707 (MO); *IVITA, km 59 Pucall-

pa-Tingo Maria road, Gentry et al. 18573 (MO);

*Bosque Nacional von Humboldt, km 86 Pucall-

pa-Tingo Maria road, Gentry et al. 18664 (MO);

*Prov. Maynas; Negro Urcu, Rio Napo, McDaniel

15328 (MO); *Prov. Maynas; Pebas, mouth of Rio

Ampiyacu, Plowman et al. 6455 (F); *Prov. May-

nas; Pebas, Plowman et al. 6737 (F); *Prov. May-

nas; Pebas, Plowman et al. 6939 (F); *Prov. May-

nas; Dist. Iquitos, Caserio Momoncillo, Revilla

304 (MO); *Prov. Maynas; Dist. Iquitos, cerca a la

boca del Rio Nanay, Revilla 1724 (MO); *lower

Rio Nanay, Williams, LI., 584 (F). Madre de

Dios: *ca. 5 km from Puerto Maldonado near Rio

Tambopata, Gentry 16249 (MO). San Martin: Tar-

apoto and vicinity, W side of Rio Huallaga, S of

Shapaja 1-4 km, 900 ft, 28-30 Jul. 1937, Belshaw

3130 (NY, uc, us); Pucaloma, 330-400 m, 12 Jul.

1950, Ferreyra 7778 (us), 1855-1856, Spruce

3973 (BM, BR, CGE, G, MG, NY, w), 350 m, Oct.

1902, Ule 6481 (G, MG, L), 750 m, Dec. 1929, LI.

Williams 5482 (F), 830 m, 6 Feb. 1947, Woyt-

kowski 35008 (F, G, MO, uc); entre Bellavista y

Banos 200-300 m, 5 Sep. 1948, Ferreyra 4760

(us); Zepelacio near Moyobamba, 1100-1200 m,

Oct.-Nov. 1933, Klug 3326 (A, F, G, GH, MO, NY,

s); Roque, 19 Jul. 1925, Melin 186 (s); Saposoa,

400 m, 14 Oct. 1959, Woytkowski 5525 (F, LE, p,

us); Rioja, 800 m, 9 Jan. 1961, Woytkowski 6170

(GH); Boqueron Pass, 92 km from Tingo Maria on

highway to Pucallpa, ca. 400 m, 16 Dec. 1949-5

Jan. 1950, Allard 22096 (us); *Prov. San Martin;

Tarapoto, above Hotel Turista, Plowman 5984 (F).

Without state: Peru, 1835, Matthews 1320 (CGE,

K). BRAZIL. Acre: Rio Jurua-Mirim, vicinity of

Porangaba, 21 May 1971, Maas et al. PI 3228

(wis); Rio Acre, Xapury, Jan. 1911, Ule 9746 (K,

MG); Paraguassu, Seringal, Nov. 1911, Ule 9749

(K, MG); Rio Branco, 26 Feb. 1962, Vasconcelos

& Coelho s.n. (INPA). Amapa: Rio Oiapoque in

shade of virgin forest, 5 km SE of Clevelandia,

3°48'N, 51°53'W, 5 Aug. 1960, Irwin et al. 47359

(MG, NY); lower slopes and base of Mount Tipac,

0-100 m, 3°36'N, 51°19'W, 16 Oct. 1968, Irwin

48753 (NY). Amazonas: Esperan9a at mouth of

Rio Javary, 26 Jan. 1952, Ducke 865 (F, GH, MG,

MO, R, us); Camatian, 30 Jan. 1949, Froes 24035

(RB); Rio Purus, Cachoeira, 21 Mar. 1904, Huber

4204 (MG); 28 Apr. 1904, Huber 4377 (MG); Rio

Purus, Rio Ituxi, Rio Curuquete, Cachoeira Re-

publica, 24 Jul. 1971, Prance et al. 14551 (wis);

upper Amazon, in forest near Tabatinga, 30 Nov.

1874, Traill 599 (K, p); Tocantins, 24 Nov. 1874,

Traill 600 (K, p); Rio Jurua, Bocca do Paranan,

Apr. 1901, Ule 5453 (G, K, MG, L); *Camatian,

Froes 23974 (GH). Rondonia: basin of Rio Ma-

deira, 1 km S of Ribeirao, road Abuna to Guaja-

ra-Mirim, 27 Jul. 1968, Prance et al. 6570 (wis);

*30 km NE de Guajara-Mirim, Krapovickas &

Schinini 35155 (F). BOLIVIA. Beni: Rio Iboy,

Beni, 800 ft, 14 Feb. 1922, White 2386 (NY, us);

Beni River, Jul. 1886, Rusby 2122 (NY). La Paz:

PLOWMAN: REVISION OF BRUNFELSIA

81

Guanai-Tipuani, Bolivian Plateau, Apr.-Jun.

1892, Bang 1398 (A, BM, F, c, GH, K, M, MICH, MO,

NY, PH, R, us, w); Mapiri, 550 m, 9 Apr. 1927,

Buchtien 1299 (us); Mapiri region, San Carlos at

Sarampiuni, 600 m, 28 Feb. 1927, Buchtien 1300

(us); Charopampa, 570 m, Nov. 1907, Buchtien

2029 (us), 5,000 ft, Apr. 1886, Rusby 621 (BM, G,

GH, K, LE, MO, NY, PH, us, w); Tumu-pasa, 1800 ft,

7 Jan. 1902, White 1840 (BM, NY, us). Pando:

bank of Rio Madeira, 6 km above Abuna, 1 1 July

1968, Prance et al. 5854 (wis); Rio Acre, Cobija,

Dec. 1911, Ule 9747 (K, MG), Jan. 1912, We 9748

(G, K, MG). Santa Cruz: Prov. Gutierrez, Canton

Buenavista, Dolores, 450 m, Steinbach 1805 (A,

uc); Buenavista, 500 m, 14 Dec. 1920, Steinbach

5487 (GH), 450 m, 2 Jan. 1926, Steinbach 7399

(A, K); *Buena Vista, Bella Union, Steinbach 7936

(A, BM, F, K).

Brunfelsia grandiflora was first collected dur-

ing the Ruiz and Pavon expedition in 1798. Ben-

tham (1846) considered it to be very similar to

his B. latifolia and perhaps also to its variety.

Most material that Bentham referred to B. latifolia

is now considered B. grandiflora subsp. schultesii,

indicating that he recognized the affinities of the

plants concerned, although he misinterpreted B.

latifolia, which is endemic in Rio de Janeiro (Bra-

zil).

Brunfelsia grandiflora consists of two subspe-

cies: subspecies grandiflora and subspecies schul-

tesii. Subspecies grandiflora is much more limited

in its distribution than is the second subspecies. It

occurs in the drainage systems of the Ribs Mar-

anon, Huallaga, and Ucayali in Peru; at higher

elevations in eastern Ecuador; and in the upper-

most tributaries of the Rios Purus and Jurua in

Brazil. It grows primarily at elevations of 650-

2000 m in the region known as the montana, hu-

mid, montane rain forest that stretches from Co-

lombia to northeasternmost Argentina on the east-

ern slopes of the Andes.

Brunfelsia grandiflora subsp. grandiflora is

characterized by its large flowers (to 50 mm

across) and fruits (to 22 mm in diameter), as well

as by its higher altitudinal preference. Subspecies

schultesii looks very similar but is generally

smaller in all its parts (see key to subspecies).

From the relatively few collections available, sub-

species grandiflora appears to be variable and ex-

hibits some exceptional forms. For example, col-

lections from the northern part of the range (Ec-

uador and northern Peru) often have divaricating,

not spreading, lateral nerves in the leaves and

longer pedicels (Pearce 290, Lugo 23). The latter

feature also appears in high-elevation populations

of B. uniflora in Bolivia.

In the 1920s, a Catholic missionary named Tas-

tevin collected this subspecies in the Rio Tarauaca

in the state of Acre, Brazil. He stated that it was

a high-climbing liana (Tastevin, 1922) and con-

fused the plant with the psychotropic Banisteriop-

sis. Nevertheless, other collectors have indicated

that the plants may reach 10 m in height, an ex-

ceptionally large size. Whether B. grandiflora

subsp. grandiflora may actually become lianoid in

habit has yet to be confirmed by field observa-

tions.

Brunfelsia grandiflora subsp. grandiflora is

well-known in cultivation, especially in the New

World tropics, as an ornamental. In the Amazon

region, it is also cultivated for its narcotic and

medicinal properties [see Plowman, 1977; Schul-

tes & Raffauf, 1990, 1991— Eds.] and is known

by several common names: sanango, chici-qui-

sanango, chiricsanango, and sananguillo (Peru);

and manacd grande (Brazil).

Brunfelsia grandiflora subsp. schultesii has

long been recognized in the literature under sev-

eral misapplied names, including B. bonodora, B.

latifolia, and B. maritima. This was primarily the

result of "follow-the-leader" taxonomy in which

no botanist consulted the type material on which

the original descriptions were based. It is now

clear that these other names all apply to species

that are found only in southeastern Brazil.

Brunfelsia grandiflora subsp. schultesii is wide-

ranging and polymorphic, occurring in western

South America from Venezuela southward to Bo-

livia. In addition, disjunct populations are known

from the Territory of Amapa in northern Brazil.

It grows commonly as an understory shrub in pri-

mary and secondary forests between 100 and 900

m elevation. Because this subspecies is widely

cultivated in South America for medicinal and or-

namental purposes, some activity of man has un-

doubtedly influenced its present distribution.

The two subspecies of B. grandiflora differ pri-

marily in the size and form of the flowers and in

the size of the fruits. Subspecies schultesii also

tends to be smaller in habit and leaf size, but there

is considerable overlap between the subspecies in

these characters.

The two subspecies are also ecologically and,

to some extent, geographically isolated. The typ-

ical subspecies is usually found between 900 and

2000 m elevation in the inter-Andean valleys of

Peru. Brunfelsia grandiflora subsp. schultesii nor-

82

FIELDIANA: BOTANY

mally occurs at much lower elevations, between

100 and 900 m, throughout the upper Amazon and

its tributaries, although isolated populations in

Bolivia may reach 1800 m. From morphological

and geographical evidence, it appears that subspe-

cies grandiflora has arisen from its much more

widespread and variable counterpart through iso-

lation in the intermontane valleys.

Intermediates between B. grandiflora subsp.

grandiflora and B. grandiflora subsp. schultesii

are known from areas where the two coexist. This

occurs in the Brazilian state of Acre and in the

lower Maranon valley of Peru near Pongo de

Manseriche (cf. Tessmann 4921, Krukoff 5284,

Vie 5453). These specimens are intermediate in

the size of their flowers and fruits, which suggests

hybridization between the two subspecies in these

areas. No intermediate forms have been found in

the higher montane valleys or in the tropical low-

lands.

As would be expected from such a wide-rang-

ing plant, B. grandiflora subsp. schultesii shows

considerable morphological variation. Some of

this variation may be due to environmental fac-

tors such as those that exist between the llanos

of Venezuela and the cloud forests of Bolivia,

but it seems that some populations have become

genetically differentiated as well since morpho-

logically distinct groups can be readily and con-

sistently separated. These variants, however, are

relatively slight and in my judgment should not

receive formal names. I will instead discuss

some of the more prevalent variations that occur.

The most common form of subspecies schultesii,

represented by the type collection (Schultes et al.

24108), is found from the Colombian Putumayo

south to Tarapoto in Peru and eastward to Leticia

in the Colombian Amazon. These plants have

large, more or less oblong-lanceolate leaves and

many flowers in a compact inflorescence. It is

this form that occurs disjunct in the north Bra-

zilian state of Amapa (Invin 47359, 48753). This

is the only known collection east of Leticia, a

disjunction of some 2300 km. The species is ap-

parently absent from the lower Amazon, al-

though additional collecting in this region may

prove otherwise.

In the northern part of its range, in the llanos

of Venezuela and Colombia and south to the Co-

lombian Putumayo, another form of B. grandiflo-

ra subsp. schultesii occurs. These populations

have large, oblong, subcoriaceous leaves and

more lax, branched inflorescences (cf. Ariste-

guieta 1669, Perez-Arbelaez 688). The leaves in

some specimens approach those of B. chiricaspi

in their large size and ochraceous coloring on

drying. These plants apparently survive in scat-

tered patches of woodland in the llanos and in

gallery forests along the rivers.

A third form of B. grandiflora subsp. schul-

tesii may be distinguished among populations

occurring on the eastern slopes of the Bolivian

plateau (cf. R. S. Williams 425) in wet, montane

forests up to 1600 m altitude. These plants have

rather small, lanceolate leaves and few flowers

per inflorescence, rarely reduced to three. These

populations show strong similarities to B. bono-

dora of the coastal forests of Rio de Janeiro, a

species still very poorly known. Brunfelsia gran-

diflora subsp. schultesii differs primarily in hav-

ing shorter pedicels and a smaller calyx, al-

though additional characters may be found when

more material becomes available. The floristic

affinities between the Yungas regions of Bolivia

and Rio de Janeiro have been pointed out else-

where (Plowman, 1979).

These differing forms of B. grandiflora subsp.

schultesii intergrade completely with adjacent

populations and should not be considered taxo-

nomically significant. They are mentioned here

to point out the degree of variation that exists in

this subspecies, which may, in some cases, lead

to confusion with other species.

Brunfelsia grandiflora subsp. schultesii grows

usually as a shrub 1-2 m tall, at times becoming

a treelet to 5-6 m. It lives in the understory of

shady, moist forests; at the margins of forest

clearings; and along rivers. It commonly persists

after the forest has been cut for lumber or shift-

ing agriculture, sprouting easily from the roots.

Considered a valuable ornamental and medicinal

plant, subspecies schultesii is much planted in

Indian houseyards and fields and often persists

after the inhabitants move elsewhere. In this

way, the plant has probably increased its distri-

bution under the influence of man. There is some

indication that its occurrence as far east as Le-

ticia in the Colombian Amazon may be due to

man's activities. Populations in this area have so

far been located only near the major rivers in

places most likely inhabited at some time in the

past.

Brunfelsia grandiflora subsp. schultesii flow-

ers and fruits sporadically throughout the year

over most of its range, with two main periods of

flowering: December to February and July to

August. The flowers are usually unscented, al-

though they may occasionally contain the odor

PLOWMAN: REVISION OF BRUNFELSIA

83

of jasmine (Plowman 2019, Ducke 865). This

plant is known by a number of vernacular names

that are indicative of its usefulness in native

medicine [these are given, along with an account

of the medicinal and narcotic uses, in Plowman,

1977— Eds.].

15. Brunfelsia hydrangeiformis (Pohl) Benth. in

DC., Prodr., 10: 198. 1846.

Franciscea hydrangeiformis Pohl, PI. Bras. Ic. 1: 7,

t. 7. 1826. Type: BRAZIL. Rio de Janeiro: ad

Olaria et Sumido rio in via ad Rio Parahybuna,

habitat inter frutices locis umbrosis, floret Sep-

tembri et Octubri, 1818, Pohl (115) (holotype,

w; isotypes, F-875036, M, w).

Shrub 1-2 m tall. Stem erect, sparsely

branched at intervals. Bark grayish brown.

Branches few, appearing at intervals, 4-6 mm in

diameter, stout, naked, glabrous, the epidermis

splitting longitudinally. Leaves crowded at ends

of branches, alternate, subverticellate with 5-10

together, blade 13-32 cm long, 3-15 cm wide,

oblanceolate to subspathulate, rarely lanceolate to

oblong, abruptly to long acuminate with straight

or subfalcate acumen, rarely acute, narrowed-cu-

neate at base, the upper surface glabrous, often

nitid, dark green, the lower surface sparsely pu-

bescent mostly at nerves, or glabrous, pale green,

membranaceous to subcoriaceous, reticulately ru-

gose, with sulcate nerves, lateral nerves 7-11,

spreading or diverging, depressed in sulci; petiole

2-13 mm long, glabrous. Inflorescence terminal,

forming in uppermost whorl of leaves, densely

capituliform to corymbiform, rarely lax, hemi-

spheric, sessile, the axis 1-3 cm long, persistent.

Flowers 5-50, purple with white spot at mouth,

fading to nearly white with age. Bracts paired, 2-

10 mm long, lanceolate to linear-lanceolate, pu-

bescent or glabrous, ciliolate at margin. Pedicel

2-8 mm long, glabrous or rarely pubescent. Calyx

15-35 mm long, 3-10 mm in diameter, tubular to

tubular-campanulate, oblong in bud, slightly in-

flated, variably pubescent, densely pilose, glan-

dular, or glabrous, light green to ferrugineous, thin

to firmly membranaceous, teeth 2-10 mm long,

erect, ovate to linear-lanceolate, acute to acumi-

nate at apex; calyx in fruit, thin, completely en-

closing fruit. Corolla tube 20-35(45) mm long,

1-2 mm in diameter, 1-1 Vi times as long as calyx,

rarely longer, slender, glabrous or glandular pu-

bescent; limb 18-32 mm across, spreading, lobes

7-12 mm long, rounded, overlapping at lateral

margins. Stamens included in upper part of co-

rolla tube; filaments slender, upper pair 5 mm

long, lower pair 3 mm long; anthers 1.5-2 mm

long, reniform. Ovary 2-2.5 mm long, ovoid;

style slender, equaling the filaments; stigma brief-

ly bifid, lobes unequal, about 1 mm long. Fruit

completely enclosed by calyx, 12-20 mm long,

9-15 mm in diameter, ovoid-subglobose, apicu-

late at apex, smooth, greenish brown, pericarp

thin-walled, dry at maturity, crustaceous, sparing-

ly dehiscent. Seeds ca. 5-10, 5 mm long, 2 mm

in diameter, oblong, irregularly angular, dark

brown, deeply reticulate-pitted. Embryo un-

known.

DISTRIBUTION — Brazil. See Figure 33.

Note: According to the International Code of

Botanical Nomenclature (Article 60.8, Recom-

mendation 60G) (Greuter et al., 1994), it is rec-

ommended that in compound names of Latin or-

igin, the final vowel of the first part be reduced

to i before a consonant. Thus, correctly, the orig-

inal orthography of hydrangeaeformis is correct-

able to hydrangeiformis.

Key to the Subspecies of Brunfelsia hydrangei-

formis (Pohl) Benth.

1. Leaves subspathulate to broadly oblanceolate,

reticulately rugose with sulcate nerves; pedicel

2-4 mm long; calyx 3-6 mm in diameter with

yellowish brown pubescence; calyx teeth 4-10

mm long, linear to linear-lanceolate

15a. subsp. hydrangeiformis

1 . Leaves oblong to oblong-lanceolate, smooth or

sulcately nerved; pedicel 4-8 mm long; calyx

5-10 mm in diameter, glabrous or with sparse

glandular hairs; calyx teeth 2-4 mm long,

ovate to ovate-lanceolate

15b. subsp. capitata

15a. Brunfelsia hydrangeiformis subsp. hydran-

geiformis. Figure 30 and frontispiece.

FIG. 30. Brunfelsia hydrangeiformis subsp. hydrangeiformis (originally as Franciscea hydrangeiformis). From

Pohl, J. E., Plantarum Brasiliae Icones et Descriptiones (1826).

84

FIELDIANA: BOTANY

PLOWMAN: REVISION OF BRUNFELSIA

85

Franciscea macrophylla Cham. & Schltdl., Linnaea

2: 601. 1827. Type: BRAZIL. Sellow sp. nov.

(holotype destroyed at Berlin; fragment of ho-

lotype, F-643247; photograph of holotype,

F-621825).

Brunfelsia macrophylla (Cham. & Schltdl.) Benth. in

DC., Prodr. 10: 198. 1846.

Leaves with blade 18-32 cm long, 6-15 cm

wide, subspathulate to broadly oblanceolate,

abruptly acuminate, glabrous on both sides, retic-

ulately rugose with sulcate nerves depressed from

above. Pedicel 2-4 mm long. Calyx 18-35 mm

long, 3-6 mm in diameter, tubular, scarcely in-

flated, sparsely pubescent to velutinous with yel-

lowish hairs, otherwise purple to brownish, dry-

ing ferrugineous, teeth 4-10 mm long, linear to

linear-lanceolate. Corolla tube briefly exserted

from calyx.

DISTRIBUTION — Brazil (Bahia, Minas Gerais,

Rio de Janeiro, Sao Paulo).

SPECIMENS EXAMINED — BRAZIL. Bahia: 1857,

Blanchet s.n. (G). Minas Gerais: Pirange, Oct.

1840, Gardner 5065 (BM, CGE, F, G, GH, K, NY, p,

us, w); Sao Francisco de Paula, Rio do Peixe,

Aug. 1896, Magalhaes 1643 (R); Mun. Serro, Bo-

cada Mata, 1400 m, 6 Oct. 1945, Williams & As-

sis 7928 (GH, R, uc, us); Sabara, Nov. 1839,

Claussen s.n. (BM, BR, G, p); without locality,

Claussen s.n. (CGE), Claussen 201 (BR), Langs-

dorffs.n. (us), Sep. 1836, Raben 14 (c), Schiich

591 (CGE, w), 1858, Weddell 878 (G); *km 30-32

on the road from Batinga to Jequitinhonha, Bun-

ting et al. 10351 (F). Rio de Janeiro: Floresta da

Tijuca, 7 Aug. 1879, Glaziou 11393 (K, p), Nov.

1894, Ule s.n. (R, us); Serra de Estrella, 1841,

Claussen s.n. (w); estrada entre a Estrella e a Es-

trada Rio-Petropolis, 24 Oct. 1932, J.G. Kuhl-

mann s.n. (RB); Nova Friburgo, Oct. 1947, Leite

4188 (NY); Serra de Friburgo, Sitio do Gaucho,

Oct. 1964, Duarte 8445 (HB); Cachoeira de Ma-

cacu, 22 Oct. 1964, Duarte & Pereira s.n. (GH,

RB); Serra Tingua, Rio San Antonio, 24 Aug.

1879, Glaziou 11392 (c, NY, p); Rio Parahyba,

1887, Moura s.n. (R); Ipiahos, Aug. 1934, O.

Peckolt s.n. (R); Magdalena, Marreiras, Oct. 1933,

Santos Lima 200 (RB); without locality, Langs-

dorff43 (LE), 1816-1821, St. Hilaire A-555 (p);

*Nova Friburgo, Parada do Gaucho, Pereira 9241

(F); *Municipio Santa Maria Madalena; Parque do

Desengano, Santa Clara, Araujo 954 (F); *near

Mury, Maas et al. 3326 (u). Sao Paulo: *Muni-

cipio de Banana]; Parque Nacional da Serra Bo-

caina, Martinelli 1115 (F). Without state: 1814-

1817, Bowie & Cunningham 100 (BM), Sellow s.n.

(F, M, MO, s).

15b. Brunfelsia hydrangeiformis subsp. capitata

(Benth.) Plowman comb, et slat. nov. Figure

31.

Brunfelsia capitata Benth. in DC., Prodr., 10: 198.

1846. Type: BRAZIL. Rio de Janeiro: Serra dos

Orgaos, shrub about 4 ft high growing in rather

moist places in virgin forests reaching an ele-

vation of about 4500 ft, manaca dobrado, Feb.

1837, Gardner 563 (holotype, K; isotypes, BM, F

678955, G, NY, P, w).

Franciscea capitata (Benth.) Miers, Ann. Mag. Nat.

Hist. Sen 2, 5:250. 1850.

Brunfelsia capitata var. angustifolia Benth. in DC.,

Prodr., 10:198. 1846. Type: BRAZIL. Rio de Ja-

neiro: Organ Mountains, 1841, Gardner 5826

(holotype, K).

Brunfelsia hydrangeiformis var. glabriscula J. A.

Schmidt in Mart., Fl. Bras. 8(1):256. 1864. Lec-

totype (designated here): BRAZIL, prope Rio,

Gardner 563 (isotypes, BM, F, G, K, NY, p, w).

Lectoparatype: BRAZIL. Rio de Janeiro: in

montibus prope Nova Friburgo, Jul. 1832, Riedel

85 (LE).

Leaves with blade 13-26 cm long, 3-11 cm

wide, oblong to oblanceolate, rarely lanceolate,

acute to acuminate at apex, glabrous on upper sur-

face, puberulent beneath at nerves or glabrous,

smooth to sulcately nerved, rarely reticulate ru-

gose. Pedicel 4-8 mm long. Calyx 14-35 mm

long, 5-10 mm in diameter, tubular to tubular-

campanulate, often somewhat inflated, glabrous or

sparsely glandular pubescent, light green, teeth 2-

4(7) mm long, ovate to ovate-lanceolate. Corolla

tube rarely exserted from calyx.

DISTRIBUTION — Brazil (Espirito Santo, Minas

Gerais, Rio de Janeiro, Sao Paulo).

SPECIMENS EXAMINED — BRAZIL. Espirito

Santo: Santa Barbara de Caparao, 4 Dec. 1929.

Mexia 4089-a (uc, us). Minas Gerais: Estasoao

Experimental de Cafe, Coronel Pacheco, 29 Nov.

1940, Heringer 460 (RB); without locality, Ker-

FIG. 31. Brunfelsia hydrangeiformis subsp. capitata comb, et stat. nov. Isotype specimen in the herbarium of The

Natural History Museum, London (BM).

86

FIELDIANA: BOTANY

Del. bl T. PlowiMB Wl - <V & . "JI

, /</; ;

PLOWMAN: REVISION OF BRUNFELSIA

87

mann 270 (BR), 1824, Riedel 113 (K, LE), 1816-

1821, St. Hilaire B-58 (P). Rio de Janeiro: Pico

do Papagaio, Alto da Boa Vista, Oct. 1928, Brade

s.n. (R), 29 Nov. 1928, Ducke s.n. (RB); Nova Fri-

burgo, Nov. 1842, Clamsen s.n. (F), 6 Dec. 1881,

Glaziou 14173 (c, P); Serra dos Orgaos: Terezo-

polis, 1888, Brunei s.n. (R); Estrada do Fagundes,

9 Dec. 1948, Duarte & Pereira s.n. (GH, RB);

Parque Nacional da Serra dos Orgaos entre km

8.5-9, 11 Jan. 1960, Plaster 56 (R); Picada Sau-

dade, 700 m, Jul. 1915, Lutzelburg 6495 (M); Val

do Rio Beijaflor, Mato entre Abr. 1 e 2, 1400 m,

30 Dec. 1932, Markgraf & Apparicio 10495 (RB),

U.S. Exploring Expedition s.n. (us); caminho para

Pedra do Sino, 2000 m, 2 Nov. 1952, Vidal

11-5435 (R, uc). Petropolis, Morin, 20 Jan. 1883,

Glaziou 14174 (c, F, LE, P, R, us); Quitandinha,

790 m, 29 Dec. 1939, Lutz 1551 (R); Valance, 16

Sep. 1878, Glaziou 9558 (GH); Cantagallo, 1859,

T. Peckolt 71 (BR); Serra de Estrella, 11 Dec.

1823, Riedel s.n. (LE), 1844, Weddell 818 (NY);

Itatiaia, lote 90, 8 Jan. 1947, Aparicio & Edmundo

868 (RB); Monte Serrat, 2 Jan. 1929, Campos Por-

to 1879 (B, RB); without locality, 1838, Miers s.n.

(F, P), 1844, Widgren s.n. (LD, s, us); *Serra dos

Orgaos, Gardner 563 (BM, F, G, K, NY, OXF, P, us,

w); *Municipio Santa Maria Madalena; Santa Ma-

ria Madalena, Pedra Dubois, Plowman & de Lima,

H.C., 12901 (F). Sao Paulo: Alto da Serra, Jan.

1904, Edwall 15183 (SP); Serra da Bocaina, Lutz

Boc 14 (R). Without state: Caminho de Lucina-

dos, 24 Aug. 1879, Franklin & Glaziou 5319 (R);

Glaziou 4909 (c), 8471 (c), Lobb s.n. (K), Dec.

1837, Miers s.n. (NY, us), Sellow 719 (G), Sellow

s.n. (BM, F, LE, M), Widgren 596 (s).

Brunfelsia hydrangeiformis is a beautiful shrub

of southeastern Brazil. It grows in moist, shady

primary forests between 700 and 2000 m eleva-

tion. It is now a rare plant and can be found only

in certain protected natural areas, such as the Ser-

ra da Estrella and Serra dos Orgaos in the State

of Rio de Janeiro and perhaps elsewhere where

patches of old forests have survived.

This brunfelsia shows considerable morpholog-

ical variation, and two subspecies are readily dis-

tinguished. The typical subspecies, hydrangeifor-

mis, as described by Pohl (Fig. 30) has a long,

tubular calyx that is hirsute or rarely glandular

pubescent in varying degrees. The leaves of this

species are usually broad, often subspathulate, and

deeply sulcately veined, giving a rough, reticulate

pattern to the surface. The nerves and principal

veinlets are sunken in shallow furrows on the up-

per surface.

A plant related to B. hydrangeiformis was de-

scribed as Franciscea macrophylla Cham, and

Schltdl. (Chamisso and Schlechtendal, 1827) and

later transferred to B. macrophylla by Bentham

(1846). The type, collected by Sellow, was de-

stroyed in Berlin, but a photograph and fragment

are preserved at the Field Museum of Natural His-

tory in Chicago. This plant differs from B. hy-

drangeiformis in having a more lax inflorescence

in which the pedicels are elongated and articulat-

ed with a peduncle, similar to that of B. bras-

iliensis. The entire inflorescence and calyx bear a

ferruginous pubescence of sessile glandular hairs,

almost scurfy in appearance. Only one modern

collection (Williams & Assis 7928) from Minas

Gerais matches the type of this plant, which ap-

pears to be a rare aberrant form of subsp. hydran-

geiformis.

Brunfelsia hydrangeiformis subsp. capitata,

originally published by Bentham as B. capitata,

is a well-defined subspecies differing from sub-

species hydrangeiformis in having narrower, ob-

lanceolate leaves; longer pedicels (4-8 vs. 2-4

mm); and a glabrous or sparsely pubescent calyx

that tends to be tubular-campanulate and broader

above than at the middle. In addition, the calyx

teeth in subspecies capitata are usually ovate in-

stead of linear-lanceolate and shorter (2-4 vs. 4-

10 mm). The leaves are often smooth and lack the

rugose furrowing of the nerves.

Intermediates between the two subspecies are

not uncommon (cf. Gardner 5065) and indicate

that hybridization probably occurs between them.

The subspecies of B. hydrangeiformis are not as

well-defined, ecologically or geographically, as

those in other species of Brunfelsia. Subspecies

hydrangeiformis occurs further inland in Minas

Gerais and extends further north than subspecies

capitata. However, field studies of these plants

may reveal ecological differences between them,

such as a difference in the time of flowering. Sub-

species hydrangeiformis flowers from August to

November; subspecies capitata from October to

January, as indicated from herbarium labels.

Brunfelsia hydrangeiformis is most closely re-

lated to B. mire of Bolivia, a plant that is known

from several heterogeneous populations, some of

which are very similar to B. hydrangeiformis.

Their differences are discussed under B. mire.

Brunfelsia hydrangeiformis is perhaps the

showiest of all the South American brunfelsias

and yet has received little attention from horti-

FIELDIANA: BOTANY

FIG. 32. Brunfelsia imatacana. Reproduced courtesy of the Field Museum of Natural History.

culturalists. After a brief period of cultivation in

European conservatories, these plants were ne-

glected, possibly because of difficulties in propa-

gation. As indicated by its specific epithet, B. hy-

drangeiformis bears a superficial resemblance to

Hydrangea hortensis.

16. Brunfelsia imatacana Plowman, Fieldiana,

Bot. n.s. 8: 4, 1981. Type. VENEZUELA.

Bolivar: About 20 km NE of Guasipati on

road to El Miamo along side road just before

Rio Cabeza Mala, alt. 200 m, open woodland

with white sandy soil. Treelet 4-5 m tall,

flowers violet fading pure white with age,

odor of jasmine. N. v. Jazmm del monte,

Juan de la Calle. Used for banos by healers.

25 Sept. 1968. T. Plowman 1918 (holotype,

F- 18 13259; isotypes, F- 1746565, GH, K, NY, p,

uc, us, YEN). Figure 32.

Shrub or treelet 3-5 m tall. Bark reddish or

grayish brown, longitudinally and transversely

cracked, shedding in thin flakes. Branchlets about

2 mm in diameter, pubescent or glabrate, grayish

brown, longitudinally cracked. Leaves scattered

along branchlets, 65-150 mm long, 20-65 mm

wide, oblong-elliptic or elliptic-obovate, short to

long acuminate, the apex itself acute to blunt, cu-

neate to narrowed at base, glabrous or sparsely

glandular-pubescent beneath, especially at midrib,

dull dark green above, dull paler green beneath,

firmly chartaceous to subcoriaceous, the lateral

nerves 6-10, spreading, straight to somewhat

arching, indistinct; petiole short, 5-10 mm long,

1.5 mm in diameter, pubescent. Inflorescence ter-

minal, 1- or 2-flowered, subsessile. Bracts 1-3

mm long, linear-lanceolate or squamiform, pubes-

cent, caducous. Flowers light violet with a white

spot at orifice, fading to pure white with age, fra-

grant with odor resembling jasmine. Pedicel 8-9

mm long, 1 mm in diameter, glabrous, becoming

thickened, warty in fruit. Calyx 15-21 mm long,

7-10 mm in diameter, tubular, terete, glabrous or

with scattered glandular hairs, teeth 4-5 mm long,

subequal, ovate, acuminate; calyx in fruit 18-25

mm long, campanulate, thickly coriaceous, dotted

PLOWMAN: REVISION OF BRUNFELSIA

89

FIG. 33. Distribution of Brunfelsia hydrangeiformis (solid triangle), B. imatacana (solid square), and B. mire

(solid circle).

with lenticels. Corolla tube 30-38 mm long, 1-2

mm in diameter, twice as long as calyx, straight,

with scattered glandular hairs, orifice 3 mm

across, lobes 13-23 mm long, subequal, rounded.

Stamens inserted in upper part of corolla tube;

filaments 1 mm wide, strap-shaped; the longer an-

terior pair 5 mm long, suberect, included in mouth

of corolla tube, the shorter posterior pair 3 mm

long; anthers orbicular-reniform, upper pair 1.5

mm across, lower pair 1.8 mm across. Ovary 1.8

mm long, 2 mm in diameter, broadly ovoid, with

7-8 ovules per locule; style 22-23 mm long; stig-

ma 1.5-2 mm long, briefly bifid. Capsule dry at

maturity and completely enclosed by coriaceous

calyx, 12-17 mm long, 14-16 mm in diameter,

ovoid to subglobose, smooth, light green, pericarp

0.5 mm thick, crustaceous. Seeds 6-10, 6-9 mm

long, 2-5 mm in diameter, oblong-ellipsoid,

somewhat angular, reddish brown, reticulate-pit-

ted.

DISTRIBUTION — Restricted to the region of the

Serrania de Imataca and Serrania de Nuria in the

northeastern part of the state of Bolivar, Venezue-

la. See Figure 33.

SPECIMENS EXAMINED— VENEZUELA. Boli-

var: Reserva Forestal de Imataca, along logging

road to Campamento Paraiso, 28 km NE of Upata,

ca. 100 m, 26 Sep. 1968, Plowman 1919 (ECON,

F, GH, K, s, us, YEN); low, flat woodland E of Rio

Cabeza Mala, 15 km NE of Guasipati, on road to

Miamo, 2 Jun. 1960, Steyermark 86244 (us, YEN);

Altiplanicie del Nuria, 5 km from Hato de Nuria,

E of Miamo, 400 m, 10 Jan. 1961, Steyermark

88290 (NY, YEN); more or less level forest along

pica 105, 40 km S of Tumeremo, E of highway

90

FIELDIANA: BOTANY

between Tumeremo and El Dorado, 29 km N of

El Dorado, 220 m, 23 Jul. 1960, Steyermark

86574 (NY, us, YEN); Reserva Forestal "La Para-

gua," margenes del Rio Asa, Jun. 1970, Blanco

826 (F, YEN); El Dorado, cultivated, Aug. 1957,

Trujillo 3488 (MY).

Previously, I considered B. imatacana to be a

subspecies of the related species B. pauciflora of

southeastern Brazil (Plowman, 1973, 1979). Al-

though the two species are superficially similar,

B. imatacana can be readily distinguished by the

leaf shape; the number of lateral nerves; fragrant

flowers; the shorter pedicel, calyx, and capsule;

and the smaller number and larger size of the

seeds. In addition, the two species are widely dis-

junct by a distance of 3,800 km.

Brunfelsia imatacana is endemic in the north-

easternmost part of Bolivar state in Venezuela, in

the range of low hills known as the Serrania de

Imataca and Serrania de Nuria. This area occupies

the northern rim of the Venezuelan section of the

Guayana Shield. This region has recently been

recognized by Steyermark as a distinct phytoge-

ographic unit of the flora of Venezuela. Most of

the flora found here represents species of the

Guianas at their western limits as well as various

eastern Amazonian species at their northeastern

limits of dispersal (Steyermark, 1968, 1979).

Steyermark (1968) lists 278 plant species known

only from the Serrania de Imataca and another

514 species that are restricted to the even more

limited area of the Altiplanicie de Nuria. The Im-

ataca region has been documented as a major cen-

ter of both plant and animal species endemism

and has been proposed as a Pleistocene forest ref-

uge area (Prance, 1973; Brown, 1979; Steyer-

mark, 1979).

During the collection of the type material, local

guides informed me that the foliage of B. imata-

cana was used by healers (brujos) for herbal

baths. Although this use of the plant was not ob-

served, bulk samples of roots, stems, and leaves

were collected for chemical analysis. The freshly

air-dried leaves and stems were sent to Dr. John

Leary of the Massachusetts College of Pharmacy,

Boston, for testing. Dr. Leary (pers. comm., 1973)

reported that extracts of this material yielded al-

kaloid-positive precipitates based on six different

precipitating agents. In addition, he obtained a

positive test for phytosteroids but negative tests

for flavonoids, tannins, and saponins. Dried root,

stem, and leaf material of this same sample were

also sent to Mr. Jan-Erik Lindgren of the Karolin-

ska Institutet, Stockholm, Sweden. Mr. Lindgren

(pers. comm., 1973) prepared methanol extracts

of the samples, which were examined by thin-lay-

er chromatography. He also performed a complete

alkaloid extraction on the material, followed by

gas and thin-layer chromatography on the ex-

tracts. Mr. Lindgren was unable to detect the pres-

ence of any alkaloids in the samples. It remains

uncertain what, if any, active constituents may be

present in this species. However, isolation and

identification of the active constituents in other

brunfelsias with demonstrated pharmacological

activity have proved to be similarly difficult

(Plowman, 1977; Lloyd et al., 1985).

17. Brunfelsia latifolia (Pohl) Benth. in DC.,

Prodr., 10:199. 1846. Figure 34.

Franciscea latifolia Pohl, PI. Bras. Ic. 1: 3, t. 1. 1826.

Type: BRAZIL. Rio de Janeiro: Tijuca, habitat

in umbrosis, inter frutices, Schott 6161 (holo-

type, w).

Brunfelsia maritima Benth. in DC. Prodr., 10: 200.

1846. Type: BRAZIL. Rio de Janeiro: pr. Rio-

Janeiro, habitat in maritimis Taipu (holotype, c-

DC, non vidi; photograph of holotype, c; iso-

types, G-DC, non vidi).

Franciscea maritima (Benth.) Miers, Ann. Mag. Nat.

Hist, sen 2, 5: 250. 1850.

Low shrub 0.2-1.0 m tall, much branched from

near base. Branches often somewhat zigzag,

leafy. Bark thin, light brown to grayish. Branch-

lets light brown, with internodes about 1 cm long.

Leaves subsessile, variable in form, blade 4-9 cm

long, 2.2-5.5 cm wide, elliptic to oblong, ovate

or obovate, apically acute, blunt or rounded,

sometimes with a short acumen, or emarginate,

basally obtuse to cuneate, rarely truncate, gla-

brous, at times minutely pubescent beneath at

midrib, dull or shiny above, dark green, dull and

pale green beneath, firmly membranaceous to co-

riaceous, lateral nerves 5-8, mostly straight; pet-

iole 1-4 mm long, short, erect, glabrous, dark pur-

ple. Inflorescence terminal, compact, the axis 3-

7 mm long, glabrous. Flowers 3-20, pale violet

with prominent white thickening at mouth, fading

to white, scentless. Bracts 1-2 per flower, 1-10

mm long, lanceolate, ciliolate, caducous. Pedicels

6-8 mm long, slender, glabrous or sparsely glan-

dular, light green, becoming thicker in fruit. Ca-

lyx 9-13 mm long, tubular-campanulate, obovoid

in bud, glabrous or sparsely glandular, light green,

smooth, teeth 2-3 mm long; erect to recurved,

triangular to lanceolate, acute at apex, calyx in

fruit coriaceous, striately veined, enclosing fruit.

PLOWMAN: REVISION OF BRUNFELSIA

91

./?

92

FIELDIANA: BOTANY

Corolla tube 15-20 mm long, 2 mm in diameter,

twice as long as calyx, sparsely glandular, pale

violet near base, white toward apex; limb 20-30

mm in diameter, spreading, plane, white beneath,

pale violet above fading to white, thickening at

mouth rounded or quadrangular, white, lobes

slightly unequal, the uppermost smaller, 5 mm

long, subtruncate at apex, the others 8 mm long,

rounded at apex. Stamens included in upper part

of corolla tube; filaments slender, cylindric, white,

upper pair 2 mm long, lower pair 1 mm long;

anthers 1 mm long, orbicular-reniform, brownish.

Ovary 2 mm long, ovoid, pale green; style slen-

der, white, curved at apex; stigma briefly bifid,

lobes unequal, light green. Capsule enclosed by

persistent calyx, 11-13 mm long, 8-10 mm in di-

ameter, ovoid, apiculate at apex, smooth, gla-

brous, dark green, pericarp thin-walled, dry at ma-

turity, crustaceous, sparingly dehiscent. Seeds

10-12, 5-6 mm long, 3 mm in diameter, oblong-

ellipsoid, angular, dark brown, reticulate-pitted.

Embryo 4 mm long, nearly straight; cotyledons

1.5 mm long, ovate.

DISTRIBUTION — Brazil (Rio de Janeiro). See

Figure 37.

SPECIMENS EXAMINED — BRAZIL. Rio de Ja-

neiro: Recreio dos Bandeirantes, 22 Oct. 1938,

Alston & Lutz 138 (BM); Pedra de Itauna, BR-6,

J.U. Santos 5790 (us); Jacarepagua, Pedra de Itau-

na, 24 Mar. 1971, Araujo & Kennedy s.n. (ECON);

restinga de Jacarepagua do lado sudoeste da Pedra

de Itauna, 6 Sep. 1969, Sucre 5314 (GH, RB), 7

Sep. 1969, Sucre 5348 (GH, RB); Copacabana,

Praia Grande, 1834, Luschnath s.n. (BR). Lagao

de Piratininga, 28 Dec. 1871, Glaziou 5970 (c);

restinga de Piratininga, 8 Nov. 1922, J.G. Kuhl-

mann s.n. (RB); Barra de Marica at the restinga, 7

Nov. 1972, Rizzini s.n. (GH); prope Taipu, Apr.

1833, Riedel 1308 (c, LE); *restinga de Jacarepa-

gua, Pedra de Itauna, Sucre et al. 6428 (F), *Vidal

311 (F). Without state: Mikan s.n. (F, K, w), Rie-

del s.n. (GH).

Brunfelsia latifolia has long been misinterpret-

ed by taxonomists since Bentham's treatment of

the genus in 1846. Examination of the type

(Schott 6161), which is preserved in Vienna, and

comparison with contemporary collections make

it clear that B. latifolia is both morphologically

and ecologically distinct from any other species.

Brunfelsia maritima, a species often confused

with B. grandiflora, is clearly a synonym of B.

latifolia.

Brunfelsia latifolia is a diminutive shrub en-

demic to the state of Rio de Janeiro in Brazil, in

the environs of the city of Rio de Janeiro. A rare

plant, this species grows only in sandy restingas,

a maritime formation of low shrubs occurring

along the sea coast. It flowers from October to

April. Brunfelsia latifolia is distinguished by its

low habit; the rather small, elliptic leaves; and

small, pale violet flowers. It seems to be most

closely related to B. bonodora, an equally rare,

erect-growing shrub of the mountains near Rio de

Janeiro. This species has narrower, lanceolate

leaves and larger flowers than B. latifolia. Some

specimens (Mikan s.n.) are somewhat intermedi-

ate between these two species and indicate a close

relationship and possibly hybridization. However,

this remains to be determined when additional

material is available.

In the horticultural trade, the name B. latifolia

is frequently misapplied to at least two species,

B. australis and B. grandiflora. The confusion be-

tween B. australis and B. latifolia stems from

1841, when Hooker published an illustration of B.

australis that he misidentified as B. latifolia. In

reality, B. latifolia has never been cultivated as an

ornamental.

18. Brunfelsia macrocarpa Plowman, Bot. Mus.

Leafl. 23(6):251, t. 16. 1973. Type: ECUA-

DOR. Esmeraldas: Rio Hoja Blanca con Rio

Hualpi, bosque humedo primario, 75 m, arbol

pequena del sotobosque, de 7 m, 7 cm d.a.p.

Corteza lisa, gris, corola morada, fruto co-

mestible, guayabilla, 14 Sep. 1965, Little &

Dixon 21064 (holotype, GH; isotype, us

2639557). Figure 35.

Shrub to small tree 2-7 m tall. Branches strag-

gling, bare, with smooth grayish bark. Branchlets

somewhat crooked, terete, glabrous, yellowish

brown. Leaves sparse toward tips of the branch-

lets, blade 14-25 cm long, 6.5-11.5 cm wide,

broadly elliptic, occasionally oblong, abruptly

acuminate at apex, the acumen often subfalcate,

FIG. 34. Brunfelsia latifolia (originally as Franciscea latifolia). From Pohl, J. E., Plantarum Brasiliae Icones et

Descriptiones (1826).

PLOWMAN: REVISION OF BRUNFELSIA

93

BRUNFELSIA macrocarpa Plowman

FIG. 35. Brunfelsia macrocarpa. Reproduced courtesy of the Botanical Museum of Harvard University.

94

FIELDIANA: BOTANY

blunt or abruptly cuneate at base, glabrous, dark

green above, yellowish green beneath, charta-

ceous, lateral nerves 4-6, spreading; petiole 6-17

mm long, glabrous. Inflorescence terminal at tips

of branchlets, 1 -flowered, bracteate. Flowers fra-

grant, showy, violet turning white with age.

Bracts few, 2-3 mm long, lanceolate, concave,

puberulent. Pedicel 8-10 mm long, erect, thick-

ened at apex, glabrous. Calyx 20-22 mm long,

6-8 mm in diameter, tubular, ovate in bud, weakly

inflated, more or less punctate, firmly membra-

naceous, teeth 4-8 mm long, subequal, ovate-lan-

ceolate, acute to acuminate at apex; calyx in fruit

greatly accrescent, 4-7.5 cm long, to 4 mm thick,

coriaceous, somewhat shiny, with punctate lenti-

cels. Corolla tube 40-45 mm long, 3-5 mm in

diameter, twice as long as calyx, angled in cross

section, glabrous; limb 50-55 mm in diameter,

spreading, the lobes 18-22 mm long, rounded,

subequal, the uppermost a little larger, overlap-

ping at lateral margins. Stamens included; fila-

ments subligulate, the upper pair 2 mm long, the

lower 3 mm long; anthers 2 mm in diameter, or-

bicular-reniform. Ovary 3 mm long, conical; style

36-38 mm long, slender, curved and thickened at

apex; stigma 2 mm long, briefly bifid, gaping, the

upper lobe larger. Fruit subbaccate, enclosed by

accrescent calyx, 4-5 cm long, 4-5 cm in diam-

eter, broadly ovoid or subglobose, smooth, yel-

low-ochre, pericarp 6-8 mm thick, exocarp leath-

ery, mesocarp thick, fleshy, endocarp thin, carti-

laginous, conspicuously veined. Seeds 10-20,

10-13 mm long, 5-7 mm in diameter, dark

reddish brown, reticulate-pitted. Embryo 10-11

mm long, straight; cotyledons 4-5 mm long,

broadly ovate.

DISTRIBUTION — Colombia (Narino), Ecuador

(Esmeraldas). See Figure 37.

SPECIMENS EXAMINED — COLOMBIA. Narino:

Isla Gorgona, 11-14 Jan. 1837, Barclay 911 (BM);

Costado Oriental, 21 Jun. 1950, Fernandez 365

(us), 7 Sep. 1924, Longfield 367 (K).

Known from only a few collections, this inter-

esting treelet occurs in the Island of Gorgona,

which lies just off the Pacific coast of southern

Colombia, and in the coastal forests of northern

Ecuador, where it grows as a small tree in the

understory of primary forests. The large, showy

flowers are said to be fragrant with the odor of

honeysuckle (Longfield 367). The fleshy yellow

fruit is edible and bears the common name guay-

abilla, or "little guava," which it resembles.

Brunfelsia macrocarpa is a very distinctive

species in having a large, fleshy, accrescent calyx

that encloses the fleshy yellow fruit. It most close-

ly resembles B. dwyeri of Panama in having large,

solitary, purple flowers that fade to white and a

thick-walled fruit. However, the fleshy fruit and

leaves are also very similar to those of B. cho-

coensis, which is placed in a different section.

This resemblance may be due to convergent evo-

lution occurring in these species, which have been

isolated for a long period in the lowland Choco

forest under the pressures of a similar and uniform

environment. No other species of section Fran-

ciscea has such a large and fleshy fruit, and B.

macrocarpa is the only known species of Brun-

felsia with a large, accrescent calyx.

19. Brunfelsia mire Monach., Phytologia 4(5):

342. 1953. Type: BOLIVIA: La Paz: Mapiri

Region, San Carlos, 850 m, baumchen 1 m,

bliiten violett, 25 Jan. 1927, Buchtien 1298

(holotype, NY; isotypes K, us 1399371 and

1399372). Figure 36.

Shrub 0.2-1.0 m tall, rarely to 2 m, mostly

unbranched. Stem to 8 mm in diameter, naked,

glabrous. Bark thin, light to dark brown, rugu-

lose. Leaves crowded toward apex of stem, sub-

verticillate, to about 6 per whorl, short petiolate

to subsessile, blade 12-28 cm long, 4-8 cm wide,

oblanceolate to oblong, rarely ovate-oblong, acute

to acuminate, often subfalcate at apex, cuneate to

blunt at base, glabrous, dull, dark green above,

pale green beneath, subcoriaceous to coriaceous,

smooth, lateral nerves 8-13, straight or somewhat

arching, spreading, rarely divaricating, veinlets

conspicuously reticulate with broad areolae; peti-

ole stout, 2-12 mm long, roughish. Inflorescence

terminal, arising from upper crown of leaves, ses-

sile, capituliform to corymbiform, usually dense,

the axis to 10 mm long. Flowers 8-30, showy,

violet fading white with age. Bracts 2-4 mm

long, lanceolate, ciliolate, caducous. Pedicel 3-10

mm long, slender, glabrous. Calyx 10-18 mm

long, 4-8 mm in diameter, tubular to tubular-cam-

panulate, glabrous, rarely sparsely glandular-pu-

bescent, reticulate-veined, teeth 3-7 mm long,

ovate to ovate-lanceolate, acuminate; calyx in

fruit subcoriaceous, the teeth becoming deeply

cut, dotted with lenticels. Corolla tube 25-38 mm

long, 2-3 mm in diameter, 2-3 times as long as

calyx, slender, glabrous or with scattered glandu-

lar pubescence; limb 25-50 mm across, spreading,

lobes to 1 5 mm long, broad, rounded, overlapping

PLOWMAN: REVISION OF BRUNFELSIA

95

BRUNFELSIA mire

Monachino

FIG. 36. Brunfelsia mire. Reproduced courtesy of the Botanical Museum of Harvard University.

96

FIELDIANA: BOTANY

at lateral margins. Stamens included in upper part

of corolla tube; filaments slender, upper pair 4 mm

long, lower pair 3 mm long; anthers ca. 1.5 mm

long, reniform-hippocrepiform. Ovary 2 mm

long, ovoid; style filamentous, equaling the fila-

ments; stigma briefly bifid, gaping, lobes ca. 1

mm long, subequal. Fruit 13-20 mm long, 12-18

mm in diameter, subglobose to ovoid, apiculate at

apex, smooth, pericarp 0.5 mm thick, cartilagi-

nous, dry at maturity, tardily dehiscent. Seeds 10-

15, 5-6 mm long, 3 mm in diameter, oblong-el-

lipsoid, angular, dark brown, reticulate-pitted.

DISTRIBUTION — Peru (Cuzco, Puno), Brazil

(Acre, Mato Grosso, Para, Rondonia), Bolivia

(Beni, Cochabamba, La Paz, Santa Cruz). See

Figure 32.

SPECIMENS EXAMINED — PERU. Cuzco: Rio

Marcopata, 20 km above Quince Mil, 780 m, 15

Jan. 1973, Madison 986 (GH); Prov. Convencion:

Echarate, 900 m, 2 Feb. 1939, Stork, Morton &

Vargas 10461 (G, K, uc); Sahuayaco 920 m, 17

Jan. 1947, Vargas 6311 (us). Prov. Paucartambo:

Pilcopata, Villa Carmen, 720 m, 17 Nov. 1964,

Vargas 15767 (us); *Prov. La Convencion; Sa-

huayaco, Rio Challpimayo near Pacchar, Plow-

man & Davis 4852 (COL, F, GH, K, IAN, MO, p, u,

us, USM); *Prov. Paucartambo; near Pilcopata,

Plowman & Davis 5000 (F, GH, K); *Prov. Pau-

cartambo, Plowman & Davis 5036 GH, K). Puno:

*near Puno, Soukup 446 (F). BRAZIL. Acre: Rio

Acre, Porvir, 14 Nov. 1923, J. G. Kuhlmann 809

(RB); Rio Acre, Monte Mo, Dec. 1911, Ule 9750

(G, K, MG). Mato Grosso: Arinos, caminho do

Porto Velho, Nov. 1914, J. G. Kuhlmann 1326 (R),

J. G. Kuhlmann 1327-k (SP, us); Utiarity (Papa-

gaio), Apr. 1918, .7. G. Kuhlmann 2287 (R, RB);

Mato proximo do Rio Sacre, 17 Nov. 1944, A. S.

Lima 7859 (SP); *Municipio de Colider; estrada

Santarem-Cuiaba, 30 km de Guarantan, Serra do

Cachimbo, /. L. Amaral et al. 803 (F); *Municipio

de Colider; estrada Santarem-Cuiaba, Serra do

Cachimbo a 30 km da cidade de Guaranta, M. N.

Silva et al. 17 (F). Para: *Alto Tapajos, Rio Cu-

ruru, Anderson 11129 (NY); *Municipio de Cas-

tanhal; Colonia 3 de Outubre, Guedes 265 (IAN);

*Belem; IPEAN, Pires & Silva 11811 (IAN). Ron-

donia: Rio Madeira, Porto Velho, 10 Jan. 1930,

Ducke s.n. (RB); 8 km NE of Porto Velho, 7 Nov.

1968, Prance et al. 8265 (wis); Machado River

region, source of Jatuaraha River, Angostura, Dec.

1931, Krukoff 1532 (NY); *proximo de Vilhena,

Vieira et al. 641 (F). BOLIVIA. Beni: *Prov.

Ballivian; Rio Chimane, environs Fatima, Davis

& Marshall 1106 (F). Cochabamba: *Prov. Cha-

pare; N of Villa Tunari, Besse et al. 513 (SEL). La

Paz: Uchimachi-Coroico, 20 Jul. 1894, Bang

2352 (BM, F, G, GH, K, LE, M, MICH, MO, NY, PH, us,

w); Hacienda Simaco sobre el camino a Tipuani,

1400 m, Feb. 1920, Buchtien 5640 (GH, us); Ha-

cienda Casano, 1,400 m, 27 Jan. 1923, Buchtien

7615 (GH, uc); near Yungas, 4000 ft, 1885, Rusby

1030 (F, NY, PH, us); Mapiri, 2,500 ft, May 1886,

Rusby 2611 (NY); Huachi, head of Beni River,

3000 ft, Aug. 1921, White 1072 (GH, NY); Beni

River, Jul. 1886, Rusby 862 (NY); *Prov. Larecaja;

Mapiri, Beck 4937 (F); *Apolo, Salinas- Yanaloma

trail, Boeke 1477 (F); *Prov. Larecaja, road from

Caranavi to Guanay, beyond Alcoche, Plowman

& Davis 5172 (F, GH, K, p, u, USM); *1.8 km SW

of Yolosa on road to Chuspipata, Solomon et al.

12022 (F); *14.3 km SW (above) Yolosa on road

to Chuspipata, Solomon et al. 12071 (F). Santa

Cruz: Lagunillas, Cordillera of Tucahuasi, 900 m,

Aug. 1934, Cardenas 2813 (F); Prov. Gutierrez,

Fraile Buenavista, 450 m, 3 Jan. 1926, Steinbach

7404 (GH, NY, s).

Brunfelsia mire is a small, unbranched shrub

rarely attaining more than 1 m in height. It grows

in humid, often swampy forests in deep shade,

mostly between 400 and 1400 m elevation. Its

area extends along the eastern flank of the Cor-

dillera from central Bolivia to Peru and north and

east into parts of Brazil. The flowers are violet

and fade to white with age; they appear between

November and February.

Still known from relatively few collections, B.

mire shows considerable variation in the size and

shape of the leaves and calyx. Populations from

central Bolivia, exemplified by the type collec-

tion, have typically large, oblanceolate leaves

with divaricating lateral nerves and a relatively

short, tubular-campanulate calyx. Populations

from Mato Grosso (/. G. Kuhlmann 1326) and

southern Peru (Stork et al. 10461) have ovate-lan-

ceolate leaves with arching, diverging lateral

nerves. Still other collections from the state of

Acre in Brazil (Ule 9750, Kuhlmann 809) and ad-

jacent Peru (Madison 986) have oblong to oblan-

ceolate leaves with spreading nerves and a longer,

narrowly tubular calyx. These last populations

strongly resemble B. hydrangeiformis of south-

eastern Brazil, a closely related species. The full

range and nature of variation within B. mire is still

incompletely known because of the paucity of

herbarium material and lack of field studies. At

this time, however, it seems best considered as a

PLOWMAN: REVISION OF BRUNFELSIA

97

polymorphic species exhibiting several different

phenotypes in different parts of its range.

The affinities of B. mire lie with B. chiricaspi

of Colombia and B. hydrangeiformis of south-

eastern Brazil, especially B. hydrangeiformis

subsp. capitata. The similarity between the latter

species and B. mire is striking in certain popula-

tions mentioned above and suggests that these

plants may have had a more or less continuous

distribution at some time in the past. This pattern

of distribution is also seen in B. uniflora and B.

bonodora-B. grandiflora. Subsequent differentia-

tion of populations of B. mire in western South

America has given rise to the heterogeneous com-

plex that we see today. Introgressive hybridization

may have occurred with B. grandiflora, the flow-

ers of which closely resemble some populations

of B. mire.

Brunfelsia mire may be distinguished from B.

hydrangeiformis subsp. capitata by the smoother

leaves, by the smaller calyx, and by the corolla

tube, which is exserted at least twice the length

of the calyx.

Brunfelsia mire also shows affinities with B.

chiricaspi, which also has large oblong-obovate

leaves and a corymbiform inflorescence. Brunfel-

sia mire may be distinguished primarily by the

longer corolla tube (25-38 vs. 22-25 mm), by the

spreading corolla lobes, and by usually having

many flowers per inflorescence.

The specific epithet of B. mire is taken from

the native name of the plant. It is used as a strong

medicine by the Indians of central Bolivia to ex-

pel cutaneous parasites and has been the object of

some pharmacological-pharmacognostic investi-

gation [see Plowman, 1977 — Eds.].

20. Brunfelsia obovata Benth. in DC., Prodr., 10:

199. 1846. Type: BRAZIL: Minas Gerais,

near Parahybuna, 3 ft high in a marsh in wa-

ter 2 ft deep, 1841, Gardner 5064 (holotype,

K; isotypes, BM, CGE, F, G, GH, K, NY, us

1067018, w).

Franciscea obovata (Benth.) Miers, Ann. Mag. Nat.

Hist. Sen 2, 5: 250. 1850.

Shrub to 1 m tall, much branched. Branches

spreading, naked, glabrous. Bark light brown.

Branchlets short, slender, sparsely pubescent to

villous, yellowish brown. Leaves mostly toward

the tips of branchlets, rarely scattered, blade 2-

9 cm long, 1-3.5 cm wide, obovate to oblong-

obovate, blunt to rounded at apex, at times sub-

acute or with short acumen, somewhat revolute

at margin, cuneate to narrowed at base, upper

surface usually glabrous, smooth, dull, dark

green, lower surface rugulose (in siccitate), pu-

bescent-villous, especially at midrib, yellowish

green, firmly membranaceous to thickly coria-

ceous, lateral nerves often obscure, 4-7, straight

or somewhat curved; petiole 2-5 mm long, more

or less villous, rarely glabrous. Inflorescence

terminal on current year's branchlets, sessile.

Flowers 1-10, showy, bluish to lilac, fading to

white with age. Bracts 2-10 mm long, small,

linear-lanceolate, sometimes concave, villous,

ciliolate at margin, caducous. Pedicel 2-9 mm

long, slender, thickened toward apex, angular-

grooved, glabrous, rarely glandular. Calyx 10-

22 mm long, 6-10 mm in diameter, tubular to

tubular-campanulate, oblong-ovoid in bud,

strongly inflated, 5-angled at sinuses, concave

between sinuses at lobes, glabrous, smooth, light

green, firmly membranaceous, sometimes stria-

tely nerved, teeth short, 1-4 mm long, broadly

ovate, rarely lanceolate-ovate, blunt to subacute

at apex; calyx in fruit enlarging somewhat to in-

clude capsule, nitid, coriaceous, prominently

striate-nerved. Corolla tube 20-30 mm long,

1.5-2 mm in diameter, 1-3 times as long as ca-

lyx, glandular, especially toward apex; limb 25-

45 mm long, spreading, lobes 9-16 mm long,

broadly rounded. Stamens inserted in upper part

of corolla tube; upper pair of filaments suberect,

3 mm long, anthers briefly exserted from tube;

lower pair included, 2 mm long; anthers 1.5 mm

long, reniform. Ovary 2 mm long, conical-

ovoid; style 18-23 mm long, slender; stigma 1

mm long, briefly bifid, gaping. Capsule nearly

enclosed by persistent calyx, 8-10 mm long, 8-

12 mm in diameter, subglobose, apiculate at

apex, rugose (in siccitate), greenish brown, peri-

carp coriaceous, 0.5-1 mm thick, sparingly de-

hiscent. Seeds 4-8, 4 mm long, 2.5 mm in di-

ameter, oblong, somewhat narrowed at ends, an-

gular, dark, reddish brown, reticulate-pitted. Em-

bryo unknown.

DISTRIBUTION — Brazil. See Figure 37.

Key to the Varieties of Brunfelsia obovata

Benth.

1 . Leaves 4-9 cm long, firmly membranaceous to

subcoriaceous; flowers (1)2-5; pedicels 1-8

mm long; calyx 16-22 mm long, tubular-in-

98

FIELDIANA: BOTANY

FIG. 37. Distribution of Brunfelsia latifolia (solid square), B. macrocarpa (solid triangle), B. obovata (open circle),

and B. pilosa (solid circle).

flated; corolla tube 1-1V2 times as long as calyx

20a. Brunfelsia obovata var. obovata

1. Leaves 2.5-7.5 cm long, coriaceous, stiff;

flowers 3-10 (rarely reduced to 1-2); pedicels

5-12 mm long; calyx tubular-campanulate; co-

rolla tube 1^-3 times as long as calyx

20b. Brunfelsia obovata var. coriacea

20a. Brunfelsia obovata var. obovata. Figure

38.

Leaves blunt to rounded at apex, or short acu-

minate, 4-9 cm long, 1.8-3.5 cm wide, firmly

membranaceous to subcoriaceous, with 5-7 lateral

nerves. Flowers (1)2-5. Pedicels 1-8 mm long.

Calyx 16-22 mm long, tubular. Corolla tube 20-

25 mm long, 1-1 J/2 times as long as calyx.

DISTRIBUTION — Brazil (Goias, Minas Gerais,

Sao Paulo).

SPECIMENS EXAMINED — BRAZIL. Goias:

*Municipio Alto Paraiso; Chapada dos Veadeiros,

Hatschbach & Kummrow 37257 (GH). Minas

Gerais: Dist. de Carangola, Alto da Serra da Yra-

ma, 19 Apr. 1935, /. G. Kuhlmann 141 (RB); Ca-

manducaia, margem do Rio do Sellado, 1800 m,

14 Nov. 1961, Handro 984 (SP); Sapucai-Mirim,

Serra da Mantiqueira, 1400 m, 6 Nov. 1953, M.

Kuhlmann 2926 (SP); without locality, 1816-1821,

St. Hilaire B-2051 (P). Sao Paulo: Sao Paulo,

1835, Lund 755 (c, G-DC, non vidi; photographs,

F, PH); in silvis paludosis inundatisque pr. Sao

Jose, pr. Sao Paulo, Oct.-Nov. 1833, Riedel 1467

(GH, K, LE); Est. Campo Grande (Sao Paulo Rail-

way), 17 Nov. 1893, Edwall 1981 p.p. (c, SP).

PLOWMAN: REVISION OF BRUNFELSIA

99

BRIJNFELSLA

100

FIELDIANA: BOTANY

20b. Brunfelsia obovata var. coriacea J. A.

Schmidt in Mart., Fl Bras. 8(1): 259. 1864.

Type: BRAZIL: Piaui, in silivis ad fl. Gur-

guea, May 1819, Martius s.n. (holotype, M).14

Figure 39.

Brunfelsia silvicola Taub. in Bot. Jahrb. Syst. 21: 450.

1896. Type: BRAZIL: Goias, in silvarum mar-

ginibus in ditione fluminis superioris Maranhao,

Sep., without year, Vie 3019 (type destroyed at

Berlin). Representative specimen and possible

topotype, same locality, Sep. 1892, Vie 77 (RB-

621, P, R-66267).

Leaves 2.5-7.5 cm long, 1-3.3 cm wide, blunt

to subacute at apex, coriaceous, with 4-6 lateral

nerves. Flowers 3-10, rarely 1-2. Pedicels 5-12

mm long. Calyx 10-16 mm long, tubular-cam-

panulate. Corolla tube 20-30 mm long, 1^-3

times as long as calyx.

DISTRIBUTION — Brazil (Bahia, Distrito Federal,

Goi£s, Piaui).

SPECIMENS EXAMINED — BRAZIL. Bahia: Jato-

ba (Chapada), 1913, Lutzelburg 4019 (M); *Cha-

padao Occidental da Bahia, 12 km N of Corren-

tina on road to Inhaumas, Harley 21898 (F). Dis-

trito Federal: cerrado perto Corrego Guara, 5

Mar. 1962, Castellanos 23294 (R); Brasilia, Fun-

dacao Zoobotanica, 8 Sep. 1961, Heringer 8662

(SP); Sobradinho, 7 Sep. 1962, Heringer 8993/

1197 (SP, us), 21 Mar. 1963, Heringer 9096/1290

(us); Corrego Samambaia, near Taguatinga, 1100

m, 10 Sep. 1965, Irwin et al. 8152 (wis); road

Brasilia to Taguatinga, 12 Sep. 1964, Prance &

Silva 59030 (wis); ca. 5 km S of Brasilia, on road

to Belo Horizonte (Rio Gama), 1175 m, 21 Sep.

1965, Irwin, Souza & Reis dos Santos 8514 (wis);

*Brasilia, Duarte & Santos 102 (F, NY); *Nucelo

Bandeirante, Heringer 10556 (F); *Bacia do Rio

Sao Bartolomeu, Heringer et al. 4281 (F); *ca. 5

km SE of Brasilia on road to Belo Horizonte, Ir-

14 In this thesis (Plowman, 1973) and in the manu-

script we received, this taxon was cited as subspecies

coriacea stat. nov. However, in the herbarium at F, all

later aquired collections were annotated by Plowman in

the mid-1980s as variety coriacea J. A. Schmidt, indi-

cating to us that he had changed his mind as to the status

of the taxon. We have chosen to use the status of variety

as this was apparently a more recent decision of Plow-

man's.— Eds.

win et al. 13126 (F); *Brasilia; Corrego de Ban-

anal, Plowman 10020 (F). Goias: entre cabeceira

do Rio das Pedras et le campement du Corto, 2

Sep. 1893, Glaziou 21920 (BR, K, LE, P, R); *Cha-

pada dos Veadeiros, Anderson 6751 (F, MO);

*Cristalina, Hatschbach 43843 (F); *Formosa,

Heringer 10719 (us).

Brunfelsia obovata was described by Bentham

in 1846 in de Candolle's Prodromus, citing a

specimen collected by George Gardner. In the

same work, however, he cited Lund 755, which

also belongs to B. obovata, as a syntype of B.

calycina.

Brunfelsia obovata is a distinct species readily

recognized by the obovate, apically blunt leaves

and by the inflated, angular calyx. It is closely

allied with B. cuneifolia of southern Brazil and

has been confused with this species. Brunfelsia

cuneifolia differs in having the leaves cuneate to

acuminate at the apex; longer, lanceolate calyx

lobes; and a well-developed indument.

From an ecological standpoint, B. obovata is

somewhat specialized. It is almost always found

in aquatic or semi-aquatic environments, often

standing in water. It occurs in brejos (swampy

secondary forest), along the flooded margins of

rivers, and in humid capoeiras, from about 300 to

1800 m elevation.

Two varieties are distinguished within B. obov-

ata. The typical variety occurs in Minas Gerais

and Sao Paulo. Variety coriacea grows in the in-

terior states of Goias and Piuai and is distin-

guished by smaller, more coriaceous leaves; more

flowers (up to 10) per inflorescence; a shorter ca-

lyx (10-16 vs. 16-22 mm); and the corolla tube

relatively longer than the calyx (1^-3 vs. 1-1 ^

times as long). In addition, the calyx of var. cor-

iacea is more coriaceous and more strongly an-

gled than in var. obovata.

Few collections of B. obovata are known, and

only one (St. Hilaire B-2051) shows characters

somewhat intermediate between the two varieties,

indicating possible hybridization between them.

Variety obovata flowers in October and Novem-

ber. Its fruits are as yet unknown. One common

name, broma, has been reported for the species

(Peckolt, 1909).

Brunfelsia obovata var. coriacea corresponds

FIG. 38. Brunfelsia obovata var. obovata. From Martius, Flora Brasiliensis (1846).

PLOWMAN: REVISION OF BRUNFELSIA

101

w

..1812226

FIELD MUSEUM

OF

NATURAL HISTORY

THE NEW YORK BOTANICAL GARDEN

Plants of the Planalto do Brasil

Distrito Federal

No. 851U

• ••< Isin nbov.i*,a ssn. c -i-iac«»a

.dot. w<*n» la

sri>;^HtLsS-n?:°d-

rJ^llia^o^roaiTto'Bclo Horizonte (Rio

). Elevation 1175 ">•

i. KelS OOS OUHl^Ja / 1 ^1H" ..••rujrr^ J- . • — i

-,d ™k ~.-~^ «* °- t'srsSrti'ri.'s A«";"«"'^' A4sft

s«riKTRsr* sssMts^esai

102

FIELDIANA: BOTANY

well to B. silvicola published in 1896 by Taubert,

which is here placed in synonymy. The type of B.

silvicola was destroyed in Berlin in 1945, but an-

other collection (Ule 77) from the same locality

is extant and may serve as a representative col-

lection and may possibly be a topotype.

Brunfelsia obovata var. coriacea flowers in the

month of September, somewhat earlier than var.

obovata; fruits have been collected in March. Va-

riety coriacea appears to grow under the same

ecological conditions as var. obovata (swampy ar-

eas) but is found in a generally much drier region

of cerrado vegetation, probably being confined to

localized wet areas.

21. Brunfelsia pauciflora (Cham. & Schltdl.)

Benth. in DC., Prodr. 10: 199. 1846. Figure

40.

Franciscea pauciflora Cham. & Schltdl., Linnaea 2:

600. 1827. Type: BRAZIL. E Brasilia meridion-

ali, without locality or date, Sellow s.n. (holotype

destroyed at Berlin; fragment, F 621699; photo-

graphs, F, NY; isotypes, M).

Besleria inodora Veil., Fl. Flum.: 261. 1829

("1825"), Ic. 6, t. 81. 1831 ("1827"). Lectotype

(designated here): plate 81 of Vellozo.

Brunfelsia calycina Benth. in DC., Prodr. 10: 199.

1846. Lectotype (designated here): BRAZIL.

Santa Catarina: Insula Santa Catarina, 1832,

Baron de Colchester s.n. (K). Lectoparatype:

BRAZIL. Santa Catarina: Insula Santa Catarina,

Bade s.n. (specimen not located). Lund 755, also

cited by Bentham, is excluded as a type. It is B.

obovata Benth. var. obovata.

Franciscea calycina (Benth.) Miers, Ann. Mag. Nat.

Hist. ser. 2, 5: 250. 1850.

Brunfelsia pauciflora var. calycina (Benth.) J. A.

Schmidt in Mart., Fl. Bras. 8(1): 257. 1864.

Franciscea eximia Scheidw. ex Moore & Ayres,

Card. Mag. Bot. 1: 16. 1850. Lectotype (desig-

nated here): Moore & Ayres, Gard. Mag. Bot. 2:

177 + t. 1850. Representative collection: culti-

vated at Royal Botanic Gardens, Kew, Herbari-

um Hookerianum, 1 867, without collector (K).

Brunfelsia eximia (Scheidw.) Bosse, Handb. Blu-

meng. 524. 1859.

Brunfelsia calycina var. eximia (Scheidw.) L. H. Bai-

ley & Raffill, Stand. Cyclop. Hort. 1: 581. 1914.

Franciscea macrantha Lem., Jard. Fleur. 3: 349, t.

249. 1853. Lectotype (designated here): Lemai-

re, Jard. Fleur. 3: t. 249. 1853. Representative

collection: Royal Botanic Gardens, Kew, culti-

vated in temperate house, 1 1 Nov. 1952, Souster

s.n. (K).

Brunfelsia calycina var. macrantha (Lem.) L. H. Bai-

ley & Raffill, Standard Cyclop. Hort. 1: 581.

1914.

Franciscea lindeniana Planch., Belgique. Horticole.

15: 100, t. 16. 1865. Type: Planchon, Belgique.

Horticole. 15: 100, t. 16. 1865. Representative

collection: Royal Botanic Gardens, Kew, culti-

vated in temperate house, 16 Jun. 1953, Souster

s.n. (K) (see Fig. 42).

Brunfelsia lindeniana (Planch.) G. Nicholson, 111.

Diet. Gard. 1: 216. 1884.

Brunfelsia calycina var. lindeniana (Planch.) Raffill,

Gard. Chron. ser. 3, 83: 370. 26 May 1928.

Brunfelsia calycina var. floribunda L. H. Bailey &

Raffill, Standard Cyclop. Hort. 1: 581. 1914.

Representative collection: New Jersey, Ruther-

ford, cultivated at Julius Roehrs Co., 29 Mar.

1918, Bailey s.n. (BH).

Shrub 1-3(5) m tall. Bark somewhat smooth

or longitudinally cracked, grayish brown. Branches

sparse, erect or spreading. Bra nth lets stout, gla-

brous, rarely pubescent or glandular, smooth, dark

green. Leaves scattered or crowded near branch

tips, blade 6.5-16 cm long, 2-6.5 cm wide, ob-

long to oblong-lanceolate, more rarely elliptic-ob-

long or obovate-oblong, apically acute to short

acuminate, sometimes blunt or emarginate, cune-

ate to attentuate at base, glabrous, punctate or

rarely glandular pubescent beneath, chiefly at

midrib, upper surface dark green, dull to nitid,

lower surface pale green, firmly membranaceous

to subcoriaceous, lateral nerves 5-11, spreading,

straight or broadly arching; petiole 5-12 mm long,

glabrous or pubescent. Inflorescence terminal,

sessile or subsessile, with 1-11 flowers. Flowers

showy, deep, reddish purple fading to pale lav-

ender or white with age, with white spot at mouth

sometimes edged with violet. Bracts small, 1-3

per flower, 1-8 mm long, linear-lanceolate, con-

cave, narrowly acute apically, glabrous or glan-

dular ciliolate, caducous. Pedicel 1 1-25(35) mm

long, stout, broader toward apex, glabrous or

sparsely glandular, in fruit becoming thicker and

corky-verrucose. Calyx 18-32 mm long, 6-10

mm in diameter, tubular, rarely tubular-campanu-

late or inflated, punctate to glandular pubescent or

completely glabrous, firmly membranaceous to

subcoriaceous, pale to dark green, teeth 3-8 mm

long, ovate to ovate-lanceolate, incumbent, acute

to acuminate at apex; calyx in fruit campanulate-

urceolate, completely enclosing fruit, becoming

slightly larger and thick-coriaceous, conspicuous-

FlG. 39. Brunfelsia obovata var. coriacea.

PLOWMAN: REVISION OF BRUNFELSIA

103

3 cm

BRUNFELSIA

pauci flora (Cham.K Schlecht] Benth.

FIG. 40. Brunfelsia pauciflora. Reproduced courtesy of the Linnean Society, London.

104

FIELDIANA: BOTANY

FIG. 41. Distribution of Brunfelsia pauciflora (solid circle) and B. rupestris (solid square).

ly striate-nerved with light brown, lenticular-ver-

rucose outgrowths. Corolla tube 28-36 mm long,

1.5-3 mm in diameter, 1-2 times as long as calyx,

reddish purple above, paler beneath, ring at mouth

4-5 mm long, elliptic, white, sparsely glandular;

lobes 15-30 mm long, broadly rounded to subel-

liptic, rounded-truncate or blunt at apex, some-

what accrescent with age, more or less overlap-

ping laterally. Stamens included in upper part of

tube; filaments subterete, anteriorly canaliculate,

white, upper pair 4-6 mm long, lower pair 2-4

mm long; anthers 1.5-2 mm long, orbicular-reni-

form, greenish brown. Ovary 2-3 mm long, 1.5

mm in diameter, conical-ovoid, pale green; style

25-30 mm long, filamentous, lavender; stigma 1

mm long, briefly bifid, lobes subequal, white.

Capsule included in calyx, 20-22 mm long, 15-