ISSCA

International Society for the Study

and Conservation of Amphibians

(International Society of Batrachology)

SEAT

Laboratoire des Reptiles et Amphibiens, Muséum national d'Histoire naturelle, 25 rue Cuvier, 75005

Paris, France. — Tel.: (33).(0)1.40.79.34.87. — Fax: (33).(0)1.40.79.34.88. — E-mail: dubois@mnhn.fr.

BOARD

President: C. Kenneth Dopp, Jr. (Gainesville, USA).

General Secretary: Annemarie OHLER (Paris, France).

Deputy Secretary: Alain PAGANO (Angers, France).

Treasurer: Stéphane GROSJEAN (Paris, France).

Deputy Treasurer: Rafael DE SÀ (Richmond, USA).

Councillors: Lauren E. BROWN (Normal, USA); Alain DuBois (Paris, France); JIANG Jianping (Chengdu,

China); Esteban O, LAviLLA (San Miguel de Tucumän, Argentina): Thierry LODÉ (Angers, France):

Miguel VENCES (Amsterdam, The Netherlands).

TARIFF FOR 2004

INDIVIDUALS

[Regular 2004 subscription to Alytes (volume 22) + ISSCA + Cürcalytes 50€ or$]

Student 2004 subscription to Alyres (volume 22) + ISSCA + Circalytes 25€0r$

SrecaL orrer: gifthalf-pricesubscriptionof one year for a colleagueof yourchoice: Half of price above

Regular 2004 subscription to Alpes (volume 22) alone FE 0rS

Student 2004 subscription to Apres (volume 22) alone 23€or$

Back issues of Alytes: any single issue 15€ or$

Back issues of Alyres: any double issue 25€0r$

Back issues of Alptes: any complete volume (4 issues) 40€ or$

Back issues of Alyres: complete set of volumes 1 to 21 (1982-2004) 630€ or$

Regular five-ear (2004-2008, volumes 22 to 26) individual subscription to Alpes +

ISSCA + Circalytes 25€ or$

Regular five-year (2004-2008, volumes 22 to 26) individual subscription to Alytes 200€ or$

SPeciaL orrer: five-year (2004-2008, volumes 22 to 26) individual subscription to Alytes + ISSCA +

Circalytes, with complete set of back issues of A/yres (1982-2004, volumes 1 to 21) 15€ or$

Life individual subscription to A/ytes from 2004 on 1080 € or $

Life individual subscription to Alytes + ISSCA + Circalytes from 2004 on 1200 € or $

Patron individual subscription to Alyres from 2004 on 2160 € or $ or more

Patron individual subscription to Alytes + ISSCA + Circalyres from 2004 on 2400 € or $ or more

Important notice: from 1996 on, any new life or patron individual subscriber to Alpes is offered a free

complete collection of back issues of Alytes from the first issue (February 1982) until the start of her/his

subscription.

INSTITUTIONS

2004 subscription to Alyres (volume 22) + ISSCA + Circalytes 100€ orS|

2004 subscription to Alyres (volume 22) alone 92€ or$

Back issues of Alyres: any single issue 30€ or

Back issues of Alyres: any double issue 50€ or$

Back issues of Alytes: any complete volume (4 issues) 80€ or$

Back issues of Alyres: complete set volumes 1 to 21 1260 € or $

SPeciaL orrer: five-year (2004-2008, volumes 22 to 26) subscription to A/ytes + ISSCA + Fee

with complete set of back issues of A/ytes (1982-2004, volumes 1 to 21) 1450€or$

Circalytes is the internal information bulletin of ISSCA. Back issues of this bulletin are also available:

prices can be provided upon request by our Secretariat.

MODES OF PAYMENT

— In Euros, by cheques drawn on a French bank payable to “ISSCA”, sent to our secretariat (address

above).

— In Euros, by direct postal transfer to our postal account: “ISSCA”, Nr. 1398-91 L, Paris; if you use this

mode of payment, add 2.30 € to your payment for postal charges at our end.

21, US Dollars, by cheques payable 10 “ISSCA”, sent Lo Rafael O. DE SA. Associate Professor

epartment of Biology, University of Richmond, Richmond, VA 23173, USA, (e-mail,

rdesa@richmond.edu: fax: (804)289-8233). Bout ie Paris

AINTES

INTERNATIONAL JOURNAL OF BATRACHOLOGY

December 2004 Volume 22, N° 1-2

Alytes, 2004, 22 (1-2): 1-14.

The higher nomenclature

of recent amphibians

Alain DUBoIs

Vertébrés: Reptiles & Amphibiens,

USM 0602 Taxonomie & Collections,

Département de Systématique & Evolution,

Muséum national d'Histoire naturelle,

25 rue Cuvier, 75005 Paris, France

<dubois@mnhn.fr>

The absence of rules in the International Code of Zoological Nomen-

clature for nomenclature of taxa above superfamily is a source of instability

and confusion, especially with the recent increase in number of higher

taxa following multiplication of phylogenetic analyses. À recent proposal

concerning such rules, submitted elsewhere, is briefly presented here,

and its consequences regarding nomenclature of higher taxa of recent

amphibians are summarised. The class nomen AMPHiBiA should be credited to

De BLAINVILLE (1816) instead of LINNAEUS (1758). The nomen LISSAMPHIBIA

Haeckel, 1866 is an invalid junior synonym of BATRACHIA Brongniart, 1800,

that applies to one of the superorders of the subclass including all recent

amphibians. The valid nomen of this subclass is NEOBATRACHI Sarasin &

Sarasin, 1890. The three orders of recent amphibians should be known as

ANURA Duméril, 1806, UroDELA Duméril, 1806 and GYMNOPHIONA Rafinesque-

Schmaltz, 1814. The nomina SALIENTIA Laurenti, 1768, CaUDATA Scopoli,

1777, APora Oppel, 1811, ARCHAFOBATRACHIA Reig, 1958 and NEOBATRACHIA

Reig, 1958 are invalid and should no longer be used,

To be able to study and designate living organisms, systematists have devised a system of

scientific classification of these organisms into taxa (taxonomy) and a system of rules

pertaining to designation of these taxa (nomenclature). The latter system allows any taxon

to be universally designated by all biologists worldwide by a single scientific name or

nomen (DUBOIS, 2000). However, the current /nternational Code of Zoological Nomenclature

(ANONYMOUS, 1999; cited below as “the Code”), only deals with nomina of some taxa, from

subspecies to superfamily, excluding taxa of lower and higher ranks. Nomenclature of higher

zoological taxa above superfamily (“class-series nomina” according to Duois, 2000) should

Bibliothèque Centrale Muséum

LIL

3 3001 00216001 7

| jource : MNHN, Paris

ALYTES 22 (1-2)

be fixed by consensus among workers. However, in many zoological groups, no such consen-

sus exists, even for well-known and non-controversial taxa, as is examplified by the three

orders of recent amphibians, for which the Zoological Record, in its recent editions, uses

double denominations: “ANURA (= SALIENTIA)”, “CAUDATA (= URODELA)” and “GYMNO-

PHIONA (= APODA)”. This absence of rules is a source of confusion and instability in scientific

literature, especially given recent development of phylogenetic analyses and multiplication of

higher zoological taxa. For this reason, a set of formal rules for this nomenclature, based on

a detailed rationale, was recently proposed (Dugois, submitted). This proposal, which is much

more precise and consistent than a previous one (Dugois, 1984b), still has to be considered

and discussed by the international community of zoologists before its possible inclusion, most

likely after some changes, in the Code. The major criteria on which the proposed system is

based are as follows:

(C1) As requested in the Preamble of the Code (p. 2), the rules should respect “the

freedom of taxonomic thought or actions”. This means that these rules should not tie

nomenclature to any fixed classification of animals, and, more importantly, to any given

philosophy of taxonomy (e.g., phylogenetic).

(C2) Just like those of the Code for other nomina, these rules should work automatically,

without need of a permanent recourse to a committee, board or court, so that they allow any

taxonomist worldwide to find the valid nomen of any given taxon under any taxonomic

system.

(C3) Therefore, the status (taxonomic allocation) of any nomen should be based on the

original extension (content) of the taxon to which this nomen was first applied, irrespective of

the intension (definition) then provided for the taxon, and of subsequent uses of the nomen,

except in a few exceptional cases, as explained under (C5) below.

(C4) Like those of all other taxa, nomina of higher taxa should have been published after

1757 and their validation should follow a rule of priority (i.e.. among several nomina

proposed for the same taxon, the first published should be the valid one) and a rule of

homonymy (i.e., any nomen homonymous with a previously published nomen should be

invalid).

(CS) However, in order to avoid unnecessary instability, genuine well-known nomina, i.e.,

nomina widely used outside specialised scientific literature dealing with taxonomy and evolution,

should be protected and stabilised, even if they are junior synonyms or homonyms of other

more obscure nomina. An objective criterion is proposed to recognize nomina that should be

so protected, and this is presence of these nomina in a high number (100) of ritles of

non-taxonomic publications dealing with these animals after 1900. This is justified by the fact

that use of a nomen in a title is relevant only if this nomen is well-known to most potential

readers, and not only to specialists.

(C6) A number of criteria and rules need to be added to have a complete functional set of

rules allowing automatic and universal allocation of nomina to taxa and validation of one of

them among several competing nomina for the same taxon. In particular, whenever a couple

or set of sister-nomina Was proposed for taxa created in the process of splitting an earlier

higher taxon (such as GRADIENTIA-SALI A-SERPENTIA, CAUDATA-ECAUDATA or ANURA-

URODELA), these sister-nomina should be validated or rejected together, instead of validating

a mixture of nomina from two or more such different couples or sets.

Source : MNHN, Paris

Dusois 3

Pending publication of this long work (Dugois, submitted), its discussion by the

international community and its possible formal inclusion in the Code, a process which is

likely to take years, it may be useful to provide all batrachologists with general data and

conclusions concerning higher nomenclature of the most important groups of recent amphib-

ians.

In the recent decades, various discussions have been published concerning phylogenetic

relationships of recent amphibian groups (i.e., taxa represented by at least one species in the

extant fauna: frogs, salamanders and caecilians), both among themselves and with other

groups of fossil amphibians and other tetrapods. No consensual opinion has been reached on

most of these questions, and further discussions, based on new information, can be expected

in the future. Thus, higher taxonomy and nomenclature cannot be stabilised for the time

being. The discussion below will be restricted to the few higher taxa which do not appear

controversial and are likely to remain valid whatever the future developments of phylogenetic

studies. Given this likely taxonomic stability, it is relevant to propose stabilisation of the

nomina of these taxa for future works. Among higher taxa (above superfamily) that include

recent amphibian groups, the taxa concerned are only those of the following ranks: class,

subclass and orders. Although still controversial, the superorders will also be included in the

discussion below.

THE CLASS

Universal agreement currently exists among zoologists for recognising a class that

includes all three groups of recent amphibians (frogs, salamanders and caecilians), as well as

several all-fossil groups. Although some authors still used the nomen BATRACHIA for this class

until the end of the 20" century, most current authors now use the nomen AMPHIBIA (see e.g.

Dugois, 1984b: 10, tab. 1). In particular, this nomen was largely used in many titles of books

and other publications, both in scientific and non-scientific literature, and should therefore be

preserved according to criterion (CS).

The nomen AMPHIBIA was introduced in scientific literature by LINNAEUS (1758). Howev-

er, Linnaeus’s original taxon was quite different from the taxon now known under this nomen.

It contained many more reptile and “fish” than amphibian species and genera: only 2 of the

16 genera originally included in the taxon (Caecilia and Rana) are currently considered to

belong in it. It was split in three orders, two of which (REPTILIA and SERPENTES) included

amphibians, but these two nomina were later historically associated with reptilian groups. The

traditional division into two classes called respectively AMPHIBIA and REPTILIA, in the sense

they have retained for about two centuries, was not immediate after LINNAEUS (1758). It was

first established by DE BLAINVILLE (1816), and adopted progressively by subsequent authors.

Probably the etymological meaning of the term AMPHIBIA (“animals with a double life”)

played a rôle in final stabilisation of this term to designate frogs, salamanders and caecilians.

Since then, the nomen AMPHIBIA has been used in zoological taxonomy with various

meanings, but always for a taxon including these three groups and excluding all groups of

recent “reptiles” and “fishes”. Pending consensus among authors on cladistic relationships

between major vertebrate groups, the taxon AMPHIBIA is here used in the sense most often

Source : MNHN, Paris

4 ALYTES 22 (1-2)

found in the scientific literature, that of ZrrreL (1888), i.e., for the whole “batrachomorph”

clade as recognized e.g. by TUDGE (2000). This is the sense of the term in thousands of

publications, in most textbooks of biology and paleontology, and in all volumes of Zoological

Record since 1927. Authorship of this nomen must however be credited to DE BLAINVILLE

(1816), and the earlier homonymous nomen AMPHIBIA Linnaeus, 1758 must be rejected as

invalid. This interpretation is not new, as it had already been proposed e.g. by KUHN (1965:

12), who however incorrectly cited LATREILLE (1825) instead of DE BLAINVILLE (1816) as the

author of the current concept of the taxon.

THE SUBCLASS

Although phylogenetic relationships and taxonomy of entirely fossil groups of amphib-

ians are still controversial (see e.g.: MILNER, 1988; TRUEB & CLOUTIER, 1991; LAURIN, 1998;

SancHiz, 1998; TUDGE, 2000), consensus exists among most current authors for allocation of

allliving amphibians, and their close relative fossil forms, into a single subclass including three

orders (frogs, salamanders and caecilians). This subclass is not a taxon that can be considered

well-known or widely used by authors who are not taxonomists or evolutionary biologists, as

it was rarely mentioned in titles of non-systematic publications. Therefore its valid nomen

should be established from original contents of taxa for which nomina were coined, not by

any subsequent incorrect uses of these nomina by specialists.

For this subclass, some recent authors (e.g.: DUELLMAN & TRUEB, 1985; MILNER, 1988;

TRUEB & CLOUTIER, 1991; LAURIN, 1998; TUDGE, 2000) used the nomen LISSAMPHIBIA

Haeckel, 1866, whereas DuBois (1984b) supported use of the nomen BATRACHIA Brongniart,

1800. However, both opinions are unquestionably incorrect, as both nomina BATRACHIA and

LiSSAMPHIBIA were coined for a taxon including frogs and salamanders but expressly exclud-

ing caecilians. These two nomina are therefore available for a taxon of lower rank and will be

considered below. So, what is the valid nomen of the subclass?

The first taxonomic recognition of a taxon encompassing the three current orders of the

subclass containing all recent amphibians, and only them, was by OPPEL (1811a-f), under the

nomen NUDA. However, this nomen is invalid, for several reasons, in particular as it is a junior

homonym of NUDI Batsch, 1788.

The valid nomen for this subclass is NEOBATRACHI Sarasin & Sarasin, 1890, a nomen that

was clearly mentioned by KUHN (1967: 30) and DuBois (1983: 272; 1984b: 12, 29) as a senior

homonym of NEOBATRACHIA Reig, 1958, making the latter nomen invalid. The nomen

NEOBATRACHI Was proposed for a subclass including all recent amphibians (frogs, salaman-

ders and caecilians) as opposed to the all-fossil amphibian groups, for which SARASIN &

SARASIN (1890) used the nomen STEGOCEPHALIA. It should be used as the valid nomen for the

taxon including all recent amphibians and closely related groups, for which the nomen

LISSAMPHIBIA cannot be conserved.

Source : MNHN, Paris

Dugois 5

THE SUPERORDERS

To designate the subclass of recent amphibians, the nomen LISSAMPHIBIA Haeckel, 1866

has had growing use in the last two decades (see DuBois, 1984b: 10), although almost

exclusively in systematic publications. Few (if any) of the recent authors who used this nomen

examined HAECKEL'S (1866) book where it was first published, because if they had they would

have realised that the original taxon designated under this nomen is different from that

understood by recent authors.

HAECKEL (1866: exxx-exxxii) recognized a class AMPHIBIA, with two subclasses, for

which he proposed the nomina PHRACTAMPHIBIA and LISSAMPHIBIA. The PHRACTAMPHIBIA

were composed of three orders, two containing only fossil taxa (GANOCEPHALA and LABY-

RINTHODONTA) and one (PEROMELA) composed of the caecilians. The LISSAMPHIBIA

contained three orders of living taxa, two of which (SOZOBRANCHIA and SOZURA) embraced

the current tailed amphibians, whereas the third one, ANURA, contained the tailless amphib-

ians. Therefore, HAECKEL's (1866) LISSAMPHIBIA were exactly equivalent to BRONGNIART’S

(18004) BATRACHIA, and not to the latter plus the GYMNOPHIONA, as stated by several recent

authors. This remained the opinion of Haeckel apparently for his entire life, as in all his

subsequent works (e.g., HAECKEL, 1868, 1870, 1872, 1873, 1902) the LISSAMPHIBIA always

only contained the current ANURA and URODELA, whereas the GYMNOPHIONA were classed in

the PHRACTAMPHIBIA.

The recent confusion traces back to PARSONS & WiLLIAMS (1963: 27), who resurrected the

long-forgotten nomen LISSAMPHIBIA for a new taxon they erected for all living amphibians.

Although they acknowledged that HAECKEL (1866) had clearly excluded the GYMNOPHIONA

from his LISSAMPHIBIA, they stated that they were following Gapow’s (1901) use of the latter

nomen for all recent amphibians, a significant change for which Gapow (1901: xi, 10, 84-274)

did not provide any explanation. As GADOW (1901: 9-10) was clearly aware of the original

content of the LISSAMPHIBIA, as well as of existence of the nomen NEOBATRACHI, his choice

of the former for the taxon may be explained only by its etymological meaning (“smooth

amphibians”). He may have considered it more appropriate to designate a taxon for which he

provided the following diagnosis: “Amphibia without dermal armour”’ (GAbow, 1901: 84).

KUHN (1967: 27) did not recognize LISSAMPHIBIA as a valid taxon but wrote incorrectly about

it: “für Caudata, Gymnophiona und Salientia; heterogen”. Most other subsequent authors

seem to have simply followed PARSONS & WILLIAMS (1963) in ai ting this nomen. It was

used by ROMER (1966: 364), and adopted since then by several authors for a subclass

containing all three recent orders of amphibians, but, as first noted by DuBois (1983, 1984b)

it should be treated as a strict junior synonym of BATRACHIA Brongniart, 1800, which

furthermore has had a dramatically larger use in zoology. This latter nomen thus deserves a

detailed discussion.

Contrary to the statement by STEINEGER (1904), and as shown by DuBois (1984b: 11,24),

the familial nomen Barracu Batsch, 1788 is not available in the class-series, and BRONGNIART

(1800) must be credited with authorship of the cl series nomen BATRACHIA (as BATRA-

. The first post-1757 published use of this widespread nomen, based on the Greek term

ass

Source : MNHN, Paris

6 ALYTES 22 (1-2)

batrachos (“frog”), under the spelling Barracui, was by BATSCH (1788), who gave family rank

to this taxon. BATSCH (1788) was the first author to use the category family in classification of

the amphibians. This was a high category in his taxonomic system, between order and genus.

He recognized families throughout the entire animal kingdom. Some nomina he coined for

these families were based on stems of available generic nomina, whereas others were not. In

his class AMPHIBIA, BATSCH (1788) recognized four families, three of which (Barraci,

LACERTAE and SERPENTES) contained amphibians. The nomen 7£srupives has long been

recognized, under the form Zesrupimipar Batsch, 1788, as the valid nomen of the family of

land turtles including the genus Testudo Linnaeus, 1758 (e.g.: Bour & DuBois, 1985; IVERSON,

1992; ROGNER, 1996; MERCHAN FORNELINO & MARTINEZ SILVESTRE, 1999; LAPPARENT DE

BROIN, 2001; VETTER, 2002). The same should be done for the family nomen LACERTIDAE,

erroneously credited in recent herpetological literature either to OPEL (1811e) (e.g., PÉREZ-

MELLADO, 1998), to GRAY (1925) (e.g., EsTes et al., 1988: 211; Cet, 1993: 58; ZHao et al., 1999:

219) or to Core (1864) (e.g., TAYLOR, 1963: 928: DOWLING & DUELLMAN, 1978: 84.1).

However, the nomina Barracni and S£RPENTES, not based on available generic nomina,

are incorrectly formed as family-series nomina according to the Code, and are therefore

nomenclaturally unavailable.

The nomen Barracui Batsch, 1788 being unavailable, the author who made this nomen

available, as a nomen of order, was BRONGNIART (1800a). He created four orders in the class

REPTILES: BATRACIENS, CHÉLONIENS, OPHIDIENS and SAURIENS. These four nomina were

latinized the same year by LATREILLE (1800: xxxvii, xi, xviii, xiii), respectively as BATRACHII,

CHELONN, OPHIDH and SAURII (spellings that soon became unused, except for CHELONIT),

and shortly after by Ross & MACARTNEY (in CUVIER, 1802: tab. 3), respectively as BATRACHIA,

CHELONIA, OPHIDIA and SAURIA. Except for CHELONIA, these latter spellings have been

universally used by later authors and should be retained as correct spellings of these nomina.

BRONGNIART (18004) was the first author to remove the salamanders from the lizards, where

they had been placed by all his predecessors. He grouped them with the frogs in his new order

BATRACIENS. He also expressed doubts (BRONGNIART, 1800b: 91) about the caecilians being

properly referred to the order which he called OPHIDIENS (that included snakes, limbless

lizards and amphisbaenians), but he kept them unallocated to order and did not refer them

formally to his BATRACIENS, 50 that the latter taxon is less inclusive than the NEOBATRACHI of

SARASIN & SARASIN (1890).

The nomen BATRACHIA has been long used in zoology, but in an ambiguous sense, as it

has been employed to designate the class of amphibians (e.g.. BOULENGER, 1910), or its

subclass containing all recent amphibians (e.g., DuBoIs, 1983, 1984b), or a superordinal taxon

including only the two orders of frogs and salamanders, considered sister-taxa (e.g.: MILNER,

1988; TRUEB & CLOUTIER, 1991; ZaRDOYA & MEYER, 2001). The latter opinion is correct, as

the original extension of the taxon covered only our current frogs and salamanders. TRUEB &

CLOUTIER (1991: 295) wrote about BATRACHIA: “we restrict it to include only the Urodela and

Salientia”. Actually this is not a restriction, but a return to the original definition of the taxon.

There currently exists no general consensus on the validity of this taxon, although recent data,

both morpho-anatomical (TR: & CLOUTIER, 1991) and molecular (ZARDOYA & MEYER,

2001) strongly support it. Under this interpretation, adopted here, the nomen BATRACHIA is

the valid nomen of a superorder including frogs and salamanders, and the superorder

containing the caecilians should bear the nomen GYMNOPHIONA (see below). Under an

Source : MNHN, Paris

DuBois {i

alternative interpretation where the salamanders and caecilians are sister-taxa (e.g., FELLER &

HepGes, 1998), the nomen BATRACHIA should be kept as the valid nomen of the subclass

including all recent amphibians. The nomen NEOBATRACHI Sarasin & Sarasin, 1890 would

then become its junior synonym. In such an arrangement, the superorders should be known

respectively as ANURA Duméril, 1806 for frogs (see below) and UROPHORA Hogg, 1839 (senior

synonym of the unnecessary nomen PROCERA Feller & Hedges, 1998) for the order containing

the URODELA and GYMNOPHIONA.

THE ORDERS

In the second half of the 20° century, a few authors (e.g., Goix & GoiN, 1962) still

recognized an order (TRACHYSTOMATA Cope, 1866) for the single family SzrENIDAE Gray,

1825. Currently, there seems to be general consensus to recognize only three orders (frogs,

salamanders and caecilians) among recent amphibians, and the SIREMIDAE are now universally

included among the salamanders (DUELLMAN & TRUEB, 1985; FROST, 1985; LAURENT, 1986;

Dusgois, 1985; ZUG, 1993).

A few words only will be devoted here to the suborders of frogs and salamanders. No

consensus currently exists among authors regarding these taxa. Furthermore, the nomencla-

ture of these suborders raises a number of complex problems, the discussion of which would

require too much space here. These problems will be discussed at length in the forthcoming

publication (DuBois, submitted). Let us just stress again here (after e.g. KUHN, 1967, and

Dugois, 1984b) that, anyway, the nomina ARCHAEOBATRACHIA Reig, 1958 and NEOBATRA-

CHIA Reig, 1958 cannot be retained as valid for two suborders of ANURA, being junior

homonyms of ARCHAEOBATRACHI Sarasin & Sarasin, 1890 and NEOBATRACHI Sarasin &

Sarasin, 1890, respectively. Reig's nomina have never been used outside systematic literature,

and therefore cannot be protected on the basis of usage. Pending the publication of the

detailed analysis of this case, the best solution for authors who wish to recognise these two

suborders (a still controversial matter) may be to use the nomina DISCOGLOSSOIDEI and

RANOIDEI proposed for them by SOKOL (1977), followed and expanded by DuBois (1984b,

1985).

CAECILIANS

The first available nomen for an order including only the caecilians is APODA Oppel,

1811. In his order NUDA, OPrEL (1811a-f) recognized three taxa: APODA, CAUDATA and

ECAUDATA. The last two will be discussed below. Because of its priority, the nomen APODA

has been used by a number of subsequent authors to designate the order of caecilians or

another higher taxon containing the caecilians. However it cannot be valid for this taxon,

being a junior homonym. This nomen is preoccupied by several earlier nomina: an ordinal

nomen of fish of LINNAEUS (1758: 241): three identical nomina proposed by LATREILLE (1804:

73, 75, 103) for three different orders of fishes; and several ordinal nomina proposed by

FiscHER (1808: [13, 25, 28]), including one as a replacement nomen for OPHIDIA Brongniart,

Source : MNHN, Paris

8 ALYTES 22 (1-2)

1800 (1.e., a taxon that did not include caecilians). Therefore the nomen APODA cannot be used

for an order containing only caecilians. OpreL’s (1811c: 409) use of APODA for an order

containing the single genus Caecilia must be considered as a new nomen for a new taxon, and

therefore an invalid junior homonym. This nomen was not used enough in non-systematic

works to qualify for conservation under criterion (CS). It should therefore be definitively

abandoned in the higher taxonomy of amphibians, and cannot be retained, even as a

subdivision of the GYMNOPHIONA, as suggested e.g. by TRUEB & CLOUTIER (1991: 296).

The nomen GYMNOPHIONA should be retained for the order of caecilians. This nomen

was first used under this spelling by MÜLLER (1831), but, as established by DuBois (1984a),

this should be considered an emendation of the nomen GYMNOPHIA proposed by

RAFINESQUE-SCHMALTZ (1814b: 104). The latter author proposed many new nomina for

higher taxa of vertebrates, especially reptiles and amphibians (RAFINESQUE-SCHMALTZ,

1814a-b; RAFINESQUE, 1815), which he divided in 5 orders and 15 families. His order GYMNO-

PHIA contained a single genus, Cecilia Rafinesque-Schmaltz, 1814, an emendation of Caecilia

Linnaeus, 1758. MÜLLER’s (1831: 198) spelling GYMNOPHIONA, which has been used by many

subsequent authors, must be kept as the valid spelling of this taxon.

Finally, within the frame of the taxonomy of recent amphibians presented below, and as

a result of the rule of coordination adapted to class-series nomina (for details, see DuBoIs,

submitted), the nomen GYMNOPHIONA Rafinesque-Schmaltz, 1814 is also the valid nomen for

the superorder including this single order.

FROGS AND SALAMANDERS

Whereas many current authors agree on use of GYMNOPHIONA for the order of caeci-

lians, consensus is not as good for the other two orders of extant amphibians, salamanders

and frogs, which have received many different nomina. The most frequently used ones are

respectively CAUDATA and URODELA, and SALIENTIA and ANURA. Considerable usage of each

of the alternative nomina in non-purely systematic literature can be documented, so that none

of these four nomina can be protected against one another, and original contents of the taxa

must be used as the criterion for allocation of these nomina to our current taxa.

Most authors have long been aware that limbed amphibians were composed of two

different groups, tailed salamanders and tailless frogs, and accordingly several early authors

proposed couples of nomina for these groups. The three most noteworthy of these couples of

nomina were proposed by LAURENTI (1768), SCOPOLI (1777) and DUMÉRIL (18064). According

to the rules proposed (DUBoIs, submitted), two such nomina can be validated together, but a

combination of nomina from different couples is not acceptable.

In his class REPTILIUM, LAURENTI (1768) recognized three orders, two for which he

provided new nomina (SALIENTIA and GRADIENTIA) and one (SERPENTIA) for which he used

a nomen from LINNAEUS (1758). AI three orders included amphibians, but only the first was

homogeneous in this respect. LAURENTS (1768) nomen SALIENTIA Was proposed for the

order including frogs, and its sister-nomen GRADIENTIA for the order including salamanders.

However, both taxa were heterogeneous in this original work, especially as one genus

(Proteus) was straddling both orders, a very exceptional situation indeed in taxonomy,

Source : MNHN, Paris

Dugois 9;

contradictory to the principles of dichotomy and hierarchy used in Linnaean taxonomy. The

SALIENTIA were almost homogeneous, as they contained four genera of frogs (Bufo, Hyla,

Pipa, Rana), but also a single species that was referred to the genus Proteus. Two other species

of the latter genus were referred to the GRADIENTIA, along with two other genera of

salamanders (Salamandra, Triton) and one of frogs (Caudiverbera), but also with one of

crocodilians (Crocodylus) and nine of lizards. Probably because of this heterogeneity, the

nomen GRADIENTIA, apart from limited use in the 19% century (e.g., MERREM, 1820; GRAY,

1850; BOULENGER, 1882), was rejected by most subsequent authors, and was never used as

valid since 1900, whereas the nomen SALIENTIA was continually considered valid by many

authors. Because of the original extension of the taxon it designated (including both reptiles

and amphibians), the nomen GRADIENTIA cannot be the valid nomen for the order of

salamanders. Consequently, its sister-nomen SALIENTIA also cannot be retained as the valid

nomen for the order of frogs. Furthermore, as the taxon SALIENTIA Laurenti, 1768 included

(although in part only) the genus Proteus, the nomenclatural status of which is fixed by its

type-species (Proteus anguinus Laurenti, 1768, a salamander), the nomen SALIENTIA applies

to the taxon of rank superorder for which the valid nomen is BATRACHIA Brongniart, 1800

(see above). Therefore, the nomen SALIENTIA should not be used as valid for frogs, as

suggested e.g. by TRUEB & CLOUTIER (1991).

ScopoLi (1777) published a classification of the animal kingdom in 12 “tribus”, corre-

sponding mostly to taxa proposed by LINNAEUS (1758) either for classes or orders. Each

“tribus” could be divided in several taxa of rank “gens”, the latter in taxa of rank “divisio”,

the latter in taxa of rank “ordo” and the latter in taxa of rank “genus”. Within the divisio

REPTILIA of his gens LEGITIMA, SCOPOLI (1777) recognized two new orders: CAUDATA for the

genera Draco, Lacerta, Siren and Testudo, and ECAUDATA for the single genus Rana. Only the

second of these taxa corresponds to a group now considered homogeneous. However, only the

first of these nomina was retained by subsequent authors, while the second was forgotten

almost entirely shortly after the introduction by DUMÉRIL (1806a) of two replacement nomina

for the two nomina of ScopoLi (1777) (see below). Despite its subsequent use for the order of

salamanders by several authors, the nomen CAUDATA Scopoli, 1777 does not apply to this

taxon according to criterion (C3), as the least inclusive taxon that contains all its originally

included genera covers both reptiles and amphibians.

The first author who clearly separated salamanders from lizards, and classified them with

frogs, was BRONGNIART (1800a-b). As mentioned above, he created an order BATRACIENS for

the genera Bufo, Hyla, Rana and Salamandra. Shorty thereafter, DUMÉRIL (18064) adopted

this order (as BATRACII) and divided it in two taxa, ANOURES and URODÈLES, corresponding

to tailless and tailed amphibians. This was the first couple of taxa clearly created to separate,

within the order of living amphibians, salamanders, and only them (excluding the lizards),

from frogs, which was not the case with GRADIENTIA and CAUDATA. DUMÉRIL (18064)

introduced his two new nomina as French translations of the Latin nomina ECAUDATI and

CaUDATI which he also mentioned for the same taxa. The question may be posed, whether

DuMÉRIL’S (18064) nomina ECAUDATI and CAUDATI were new nomina, and therefore invalid

junior homonyms of ECAUDATA and CAUDATA proposed earlier by ScopoLt (1777), or new

acceptations and spellings (aponyms, sensu DuBois, 2000) for the latter nomina. In the first

four texts published by DuMÉRIL (1806a-b, 1807a-b) where this author used the nomina

ECAUDATI and CAUDATI, he did not mention SCOPOL'S (1777) text and nomina, but he did so

Source : MNHN, Paris

10 ALYTES 22 (1-2)

in later works (DuMÉRIL, 1808: 312; DUMÉRIL & BIBRON, 1834: 242), so there is little doubt

that he simply used Scopoli’s nomina but provided new definitions and contents for the taxa

designated by them.

The taxon ECAUDATI as used by DUMÉRIL (18064) included four genera, Bufo, Hyla, Pipa

and Rana. The last was the only genus originally mentioned by ScoPoLi (1777) as a member

of his ECAUDATA, a nomen of which Duméril's ECAUDATI must therefore be considered as an

emendation. However, the situation is different concerning CAUDATI. As used by DUMÉRIL

(18064), this taxon included four genera: Proteus (as Protoeus), Salamandra, Triton and Siren.

Only the last of these genera was part of the genera originally included in the CAUDATA

Scopoli, 1777, which also included reptiles, so CAUDATI Duméril, 1806, which applies to a

distinct taxon, must be considered a junior homonym created for a different taxon.

Whatever the interpretation chosen for the status of Duméril's nomina with respect to

those of Scopoli, the nomina of the latter cannot be validated for the orders of frogs and

salamanders: (1) if Duméril’s nomina are considered as two new nomina, both are invalid,

being junior homonyms of Scopolis nomina; (2) if, as supported here, they are interpreted as

subsequent uses of Scopolis nomina, only the nomen ECAUDATI, as an emendation of

ECAUDATA, could possibly be considered valid, whereas CAUDATI Duméril, 1806, designating

a distinct new taxon, is an invalid junior homonym of CAUDATA Scopoli, 1777. But then,

because they are sister-nomina, ÉCAUDATI also must be rejected as invalid.

Let us finally consider DUMÉRIL’s (18064) new nomina ANOURES and URODÈLES. They

were proposed as replacement nomina of ECAUDATI and CAUDATI, thus having the same origi-

nal definitions as the nomina ECAUDATA Scopoli, 1777 and CAUDATI Duméril, 1806. These two

nomina were later latinized, as ANURA and URODELA, and used as valid nomina by many

authors. As both these nomina have remained in wide use by many biologists since their cre-

ation, they fully qualify for validation for the two orders of batrachians. However, their reten-

tion as valid nomina imposes rejection of the nomina ECAUDATA Scopoli, 1777 (of which

ANURA is a replacement nomen) and CAUDATI Duméril, 1806 (already rejected as a junior

homonym). It is therefore not possible to maintain uses of both CAUDATA and URODELA as

valid taxa, with the former including the latter or the contrary, as was done by some

recent authors (e.g., respectively: MizER, 1988; TRUEB & CLOUTIER, 1991). Similarly,

the nomen SALIENTIA cannot be used for a taxon including the ANURA, as done also by

several authors (e.g.: MILNER, 1988; TRUEB & CLOUTIER, 1991). Validation of both nomina

ANURA and URODELA definitively rejects the couples of sister-nomina SALIENTIA-

GRADIENTIA and ECAUDATA-CAUDATA. These last four nomina should no longer be used in

higher nomenclature.

HIGHER NOMENCLATURE OF RECENT AMPHIBIANS

This review of amphibian nomenclature is but one example of the difficulties arising

from lack of rules governing nomenclature of higher taxa. Hopefully, the new proposed rules

(Dusois, submitted) will remedy this chaos. On the basis of this analysis, the nomenclature of

the major taxa of recent amphibians is as follows:

Source : MNHN, Paris

Dugois 11

Classis AMPHIBIA De Blainville, 1816

Subclassis NEOBATRACHI Sarasin & Sarasin, 1890

Superordo BATRACHIA Brongniart, 1800

Ordo ANURA Duméril, 1806

Ordo URODELA Duméril, 1806

Superordo GYMNOPHIONA Rafinesque-Schmaltz, 1814

Ordo GYMNOPHIONA Rafinesque-Schmaltz, 1814

ACKNOWLEDGEMENTS

For bibliographic information and constructive comments on the manuscript of this paper, I am

grateful to Roger Bour, Lauren E. Brown, Patrick David, Darrel R. Frost, W. Ronald Heyer, Annemarie

Ohler, Don Shepard and an anonymous reviewer.

LITERATURE CITED

ANONYMOUS [International Commission on Zoological Nomenclature], 1999. — International code of

zoological nomenclature. Fourth edition. London, International Trust for zoological Nomencla-

ture: i-xxix + 1-306.

BATSCH, À. J G. C., 1788. — Versuch einer Anleitung, zur Kennniss und Geschichte der Thiere und

Mineralien. Erster Theil. Jena, Akademische Buchhandlung: i-viii + 1-528, pl. 1-5.

BOULENGER, G. A., 1882. - Catalogue of the Batrachia Gradientia s. Caudata in the collection of the British

Museum. Second edition. London, Taylor & Francis: i-ix + 1-127, pl. 1-9.

1910. - Les Batraciens, et principalement ceux d'Europe. Paris, Doin: i-ix + 1-305 + i-xi.

R. & Dumois, A., 1985. - Nomenclature ordinale et familiale des Tortues (Reptilia). Studia

geologica salmanticensia, “1984”, vol. especial 1, Studia palaeocheloniologica, À, Salamanca,

Ediciones Universidad: 77-86.

BRONGNIART, A., 18004. — Essai d’une classification naturelle des reptiles. I partie. Etablissement des

ordres. Bull. Sci. Soc. philom., 2 (35): 81-82.

me 1800b. — Essai d’une classification naturelle des reptiles. II partie. Formation et disposition des

genres. Bull. Sci. Soc. philom., 2 (36): 89-91, pl. 6.

Cm, J. M. 1993. — Reptiles del noroeste, nordeste y este de la Argentina. Herpetofauna de las selvas

subrropicales, puna y pampas. Torino, Museo regionale di Scienze naturali, Monogr. 14: 1-949,

pl. 1-126.

D. 1864. - On the characters of the higher groups of Reptilia Squamata — and especially of the

Diploglossa. Proc. Acad. nat. Sci. Philadelphia, 6: 224-231.

— 1866. — On the structure and distribution of the genera of the arciferous Anura. J Acad. nat. Sci.

Philadelphia, (2), 6: 67-112.

Cuvier, G., 1802. - Lectures on comparative anatomy. Translated from the French of G. CUVIER (.…) by

William Ross, under the inspection of James MACARTNEY ( .…..). Vol. 1. On the organs of motion.

London, Longman & Rees: i-x1 + 1-542, tab. 1-9.

DE BLAINVILLE, H., 1816. - Prodrome d'une nouvelle distribution systématique du règne animal. Bull.

Sci. Soc. philom. Paris, juillet 1816: “105-112” [actually 113-120] + 121-124.

H. G. & DuëLLMAN, W. E., 1974-1978. - Systematic herpetology: a synopsis of families and

igher categories. New York, HISS Publications, Publications in Herpetology, 7:i-vii + L.1-118.3+

i-vii.

Duois, A., 1983. - Classification et nomenclature supragénérique des Amphibiens Anoures. Bull. mens.

Soc. linn. Lyon, 52: 270-276.

ee 1984a. — Miscellanea nomenclatorica batrachologica (V). Alytes, 3: 111-116.

Core,

Source : MNHN, Paris

12 ALYTES 22 (1-2)

Miscellanea nomenclatorica batrachologica (VII). Alytes, 4 (2): 61-78.

- 2000. - Synonymies and related lists in zoology: general proposals, with examples in herpetology.

Dumerilia, 4

- submitted. Proposed Rules for the incorporation of nomina of higher zoological taxa in the

International Code of Zoological Nomenclature, with the example of the higher nomina of recent

Amphibia.

DurLLMAN, WE. & TRUEB, L., 1985. Biology of amphibians. New York, MeGraw-Hill, “1986”: i-xix +

1-670.

DuvériL, A. M. C., 1806a. - Zoologie analytique, où méthode naturelle de classification des animaux,

rendue plus facile à l'aide de tableaux synoptiques. Paris, Allais: i-xxxiii + 1-544.

La 1806. — Analytische Zoologie. Aus dem Franzôsischen, mit Zusätzen von L. F. FRoRIEP. Weimar,

Landes-Industrie-Comptoir: [i] + i-vi + 1-346.

Lee 1807a. — Traité élémentaire d'histoire naturelle. Seconde édition. Tome 2. Paris, Deterville: i-xii +

1-360, 20 pl.

—— 1807b. — Sur la division des Reptiles Batraciens en deux familles naturelles, Nouveau Bulletin des

Sciences, par la Société philomathique, 1 (3): 62-63.

—— 1808. — Mémoire sur la division des Reptiles Batraciens en deux familles naturelles. Magasin

encyclopédique, ou journal des Sciences, des Lettres et des Arts, 1808 (2): 308-329.

DuvériL, A.-M.-C. & BIBRON, G., 1834. - Erpétologie générale ou histoire naturelle complète des Reptiles.

Tome 1. Paris, Roret: i-xxiv + 1-447.

Esres, R., DE QueIRoz, K. & GAUTHIER, J., 1988. - Phylogenetic relatioi within Squamata. : R.

Esres & G. PREGILL (ed.), Phylogenetic relationships of the lizard families, Stanford, Stanford Univ.

Press: 119-281.

FELLER, A. E, & HEDGES, S. B,, 1998. - Molecular evidence for the early history of livingamphibians. Mol.

Phyl. Evol., 9 (3): 509-516.

Fiscuer, G., 1808. — Tableaux synoptiques de zoognosie. Moscou, Imprimerie de l'Université Impériale:

[1-60] + 1-186, 6 pl.

FRosr, D. R. (ed.), 1985. - Amphibian species of the world. Lawrence, Allen Press & Assoc. Syst. Coll.

Hi-iv] + iv + 12732.

Gabow, H., 1901. — Amphibia and reptiles. London, Macmillan & Co: i-xiii + 1-68, 1 pl.

Gox, C. I. & Goin, O. B., 1962. - Introduction to herpetology. San Francisco & London, Freeman & Co.:

ix + 1-341.

Gray, JL. E., 1825. - A synopsis of the genera of Reptiles and Amphibia, with a description of some new

species. Ann. Philos., (2), 10: 193-217.

_ 1850. — Catalogue of the specimens of Amphibia in the collection of the British Museum. Part IL.

Batrachia Gradientia, etc. London, Spottiswoodes & Shaw: 1-72, pl. 3-4.

HaëckeL, E., 1866. - Generelle Morphologie der Organismen. Zweiter Band. Allgemeine Ennvickelngs-

geschichte der Organismen. Berlin, Georg Kramer: i-clx + 1-462, pl. 1-8.

= 1868. Narürliche Schôpfungsgeschichre. Berlin, Reimer: i-xvi + 1-568, 11 pl.

- 1870. - Natürliche Schôpfumgsgeschichte. Zweite, verbesserte und vermehrte Auflage. Berlin, Reimer:

i-xxxii + 1-688, 15 pl.

2 D he Schôpfngsgeschichte. Dritte verbesserte Auflage. Berlin, Reimer: i-xIviii + 1-688,

liche Schôpfungs-Geschichte. Zehnte verbesserte Auflage. Zweiter Theil. Allgemeine

Srammes-Geschichte ( Phylogenie und Anthropogenie). Berlin, Reimer: i-v + 371-832, 23 pl.

HoGG, JL. 1839. - On the classifications of the Amphibia. Mag. nat. Hist., (n.s.), #:

IVERSON, J. B., 1992. — 4 revised checklist with distribution maps of the rurtles of the world. Richmond

Earlham College: i-xiii, 1-363.

KUHN, O., 1965. - Die ‘Amphibien. Krailling bei München, Oeben: -102.

SE 1967. — Amphibien und Reptilien. Stuttgart, Gustav Fisch

LAPPARENT DE BROIN, F. DE, 2001. - The European turtle f

Dumerilia, 4 (3): 155-218.

auna from the Triassie to the present.

Source : MNHN, Paris

Dugois 13

LATREILLE, P. A., 1800. — Histoire naturelle des salamandres de France, précédée d'un tableau méthodique

des autres reptiles. Paris, Villier: i-xlvii + 1-63, pl. 1-6.

--- 1804. — Tubleau méthodique des poissons. In: Tableaux méthodiques d'histoire naturelle, [i] + 1-238, in:

Nouveau dictionnaire d'histoire naturelle, Tome 24, Paris, Deterville: 71-105.

ee 1825. - Familles naturelles du règne animal. Paris, Baillière: [i-v] + 1-570.

LAURENT, R. E, 1986. - Sous-classe des Lissamphibiens (Lissamphibia). Systématique. Zn: P.-P. Grassé &

M. DeLsoL (ed.), Traité de Zoologie, 4, Amphibiens, fasc. I-B, Paris, Masson: 594-796.

LAURENT, J. N., 1768. — Specimen medicum, exhibens synopsin Reptilium emendatam cum experimentis

circa venena et antidota Reptilium austriacorum. Viennae, Joan. Thom. Nob. de Trattnern:

1-215, pl. 1-5.

LaURIN, M., 1998. - The importance of global parsimony and historical bias in understanding tetrapod

evolution. Part 1. Systematics, middle ear evolution and jaw suspension. Annales des Sciences

naturelles, Zoologie, (13), 19: 1-42.

Linnaëus, C., 1758. — Systema Naturae per regna tria naturae, secundum classes, ordines, genera, species,

cum characteribus, diferentiis, synonymis, locis. Editio decima, reformata. Tomus I. Holmiae,

Laurentii Salvii: [i-iv] + 1-824.

MERCHAN FORNELINO, M. & MARTINEZ SILVESTRE, A., 1999. — Tortugas de España. Madrid, Antiquaria,.

1-400.

MERREM, B., 1820. — Versuch eines Systems der Amphibien. Tentamen systematis amphibiorum. Marburg,

Iohann Christian Krieger: (i-xv + 1-191) xX2, 1 pl.

Mixer, A. R., 1988. — The relationships and origin of living amphibians. /n: M. J. BENTON (ed.), The

phylogeny and classification of the tetrapods, Oxford, Clarendon Press: 59-102.

MÜLLER, J., 1831. - Beiträge zur Anatomie und Naturgeschichte der Amphibien. Z. Physiologie, 4:

190-275, pl. 18-22.

Orrez, M. 1811a. - Ordre II. Reptiles à écailles. Section IL. Ophidiens. Annales du Muséum d'Histoire

naturelle, 16 (94): 254-295.

—— 1811b. - Suite du [°° mémoire sur la classification des Reptiles. Annales du Muséum d'Histoire

naturelle, 16 (95): 376-393.

-- 1811c. - Second mémoire sur la classification des Reptiles. Annales du Muséum d'Histoire naturelle,

16: 394-418.

—— 18114. — Sur la classification des Reptiles. Paris: i + 1-84.

Pan 181le. - Die Ordnungen, Familien und Gattungen der Reptilien als Prodrom einer Naturgeschichte

derselben. München, Lindauer: i-xii + 1-87.

= 1811f. — Sur la classification des Reptiles. Nour. Bull. Sci., Soc. philom., 2 (46): 301-306.

PaRsONS, T. S. & WiLLiaAMs, E. E., 1963. - The relationships of the modern Amphibi re-examination.

Quart. Rev. Biol., 38 (1): 26-53.

PÉREZ-MELLADO, V. 1998. - Familia Lacertidae Oppel, 1811. Jn: A. SALVADOR (ed.), Reptiles, Madrid,

Museo Nacional de Ciencias Naturales, CSIC, Fauna Iberica, vol. 10: 161-166.

RAFINESQUE, C.S., 1815. Analyse de la nature ou Tableau de l'univers et des corps organisés. Palerme, Jean

Barravecchia: 1-124, 1 pl.

UE-SCHMALTZ, C. S., 1814a. - O quadro del metodo sinottico di somiologia. Specchio Sci,

Giorn. encicl. Sicilia, À (1): 11-15.

— 1814b.— Principes fondamentaux de somiologie, ou les loix de la nomenclature et de la classification de

l'empire organique, ou des animaux et des végétaux, contenant les règles essentielles de l'art de leur

imposer des noms immuables et de les classer méthodiquement. Palerme, Franc. Abate: 1-52.

Re, O., 1958. — Proposiciones para una nueva macrosistemätica de los Anuros (nota preliminar).

Physis, 21: 109-118.

ROGNER, M., 1996. — Schildkrüten. 2. Hürtgenwald, Heidi-Rogner-Verlag. 1-265.

ROMER, A. S., 1966. — Vertebrate paleontology. Third edition. Chicago & London, The University of

Chicago Press: i-ix + 1-468.

SancHiz, B., 1998. — Salientia. /n: P. WELLNHOFER (ed.), Handbuch der Paläoherpetologie, Teil 4,

München, Friedrich Pfeil: i-xii + 1-275.

SARASIN, P. & SaRasiN, EF. 1890. -Zur Entwicklungsgeschichte und Anatomie der ceylonesischen

Blindwühle /chthyophis glutinosus, L. Vierter Theil. /n: Ergebnisse naturwissenschafilicher For

chungen auf Ceylon, Zweiter Band, Wiesbaden, Kreidel: 151-263, pl. 15-24.

Source : MNHN, Paris

14 ALYTES 22 (1-2)

Scorout, I. A., 1777. - Introductio ad historiam naturalem, sistens genera lapidum, plantarum, edt

animalium hactenus detecta, caracteribus essentialibus donata, in tribus divisa, subinde ad leges

naturae. Pragae, Gerle: [i-x] + 1-506 + i-xxxvi.

Sokor, O. M., 1977. — À subordinal classification of frogs (Amphibia: Anura). J. Zool., London, 182:

505-508.

STEINRGER, L., 1904. - Amphibia versus Batrachia. Science, (2), 20: 924-925.

TAYLOR, E. H., 1963.- The lizards of Thailand. University of Kansas Science Bulletin, 44 (14): 687-1077.

TRUE, L. & CLOUTIER, R., 1991. — À phylogenetic investigation of the inter- and intrarelationships of the

Lissamphibia (Amphibia: Temnospondyli). /n: H.-P. SCHULTZE & L. TRUE (ed.), Origins of the

higher groups of tetrapods: controversy and consensus, Ithaca, New York, Cornell Univ. Press:

223-313.

Tune, C., 2000. - The variety of life. Oxford, Oxford University Press: i-xv + 1-684.

Tarralog- Turtles of the World. Vol. 1. Africa, Europe and Western Asia. Frankfurt am

Main, Chimaira. 1-96.

Zarpoya, R. & MEvER, À., 2001. — On the origin of and phylogenetic relationships among living

amphibians. Proc. nat. Acad. Sci. USA, 98 (13): 7380-7383.

Zuao, E., ZHao, K., ZHou, K. et al. (ed.), 1999. — Fauna sinica. Reptila. Vol. 2. Squamata Lacertilia.

Bejjing, Science Press: i-xi + 394, pl. 1-8.

Zivrez, K. A., 1888. Vertebrata (Pisces, Amphibia, Reptilia, Aves). /n: Handbuch der Palacontologie, II.

Band, [third installment], München & Leipzig, R. Oldenbourg: 257-436.

Zu, G.R., 1993. — Herpetology. San Diego, Academic Press: i-xv + 1-527.

Corresponding editor: Lauren E. BROWN.

Source : MNHN, Paris

Alytes, 2004, 22 (1-2): 15-18. 15

Rediscovery and redescription

of the holotype of Mantella maneryv

…s

Miguel VENcEs", Cindy WooDHEAD"”, Parfait Bora""" & Frank GLAW

* Zoological Museum, University of Amsterdam, Mauritskade 61,

1092 AD Amsterdam, The Netherlands

** Durrell Institution of Conservation and Ecology, Department of Anthropology,

Eliot College, The University of Canterbury, Canterbury, CT2 7NS, United Kingdom

*#* Université d'Antananarivo, Département de Biologie Animale,

Antananarivo 101, Madagascar

*##+ Zoologische Staatssammlung, Münchhausenstr. 21, 81247 München, Germany

The Malagasy poison frog Mantella manery Vences, Glaw & Bôhme,

1999 was described on the basis of color slides of a specimen deposited in

the collection of the Département de Biologie Animale, Université d’Anta-

nanarivo, which is the only voucher of this species known to date. The

holotype of this species was not available for morphological examination at

the time of the description but has been rediscovered by us in 2004. Its

catalogue number is UADBA 7273 and its snout-vent length is 22.7 mm. We

here provide an updated description of Mantella manery, based on mor-

phological examination of the holotype.

INTRODUCTION

The genus Mantella Boulenger, 1882 is composed of 15 species currently recognized

(GLaw & VENCES, 2003). These colorful diurnal animals are usually named Malagasy poison

frogs (Day et al., 1996) and are important for the pet trade, ecotourism, and as flagship

species for conservation (BEHRA, 1993; ZIMMERMANN, 1996; VENCES et al., 2004). After GLAW

& VENCES (1994) first mentioned and figured an unnamed species of Mantella from the

Marojejy Massif in north-eastern Madagascar, hobbyists have used various invalid (condi-

tional) names to refer to this species, such as “Mantella marojezyi or “ Mantella marojezy

To avoid an accidental description similar to the case of Mantella milotympanum Stanis-

zewski, 1996, the species was described as Mantella manery by VENCES et al. (1999), based on

photographs and field data only. The holotype was said to be “a single specimen of this species

(...) in the herpetological collection of the Zoological Institute of the Antananarivo Univer-

sity, Madagascar”. Because this specimen was not found in the Antananarivo collection, the

ription of Mantella manery was based “on color slides of this specimen”” alone

al., 1999).

In a recent effort of contributing to the inventory of the herpetological collection in the

Département de Biologie Animale, Université d’Antananarivo, Madagascar (UADBA), we

Source : MNHN, Paris

16 ALYTES 22 (1-2)

rediscovered the holotype of Mantella manery in February 2004. In the following we provide

a redescription of this species and focus on the previously unavailable morphological features

of the holotype. Terminology follows VENCES et al. (1999).

Mantella manery Vences, Glaw & Bôühme, 1999

Mantella manery Vences, Glaw & Bôhme, 1999. - Name-bearing type: holotype by original designation

(VENCES et al. 1999: 15), its catalogue number here first reported as UADBA 7273.

Usage of the name subsequent to the original description:

Mantella manery: VENCES et al., 1999; GLAW & VENCES, 2000, 2003; SCHAEFER et al., 2002;

VENCES & GLAW, 2003.

Mantella manery n. sp. (1999): STANISZEwWSKI, 2001.

Morphology of holotype. - Adult specimen in moderate state of preservation. Several cuts

through ventral skin for gonad examination. Some tissue removed from left femur for DNA

extraction. Probably a male, but gonads not sufficiently recognizable due to poor preservation

and dark color of inner organs. Body relatively stout for a Mantella; head clearly longer than

wide, slighly narrower than body; snout rounded in dorsal and lateral views; nostrils directed

laterally, very slightly protuberant: canthus rostralis distinct, concave; loreal region slightly

concave; tympanum distinct, rounded, its diameter 57 % of eye diameter; supratympanic fold

distinct, slightly curved; tongue narrow and longish-ovoid, very slightly notched posteriorly;

vomerine and maxillary teeth absent. Forelimbs slender; subarticular tubercles single; inner

and outer metacarpal tubercles distinct; fingers without webbing; comparative finger length 1

<2<4 < 3; finger discs moderately enlarged; nuptial pads absent. Hindlimbs slender; when

hindlimbs are adpressed along body, the tibiotarsal articulation reaches the posterior eye

corner; lateral (outer) metatarsalia strongly connected; a large inner and a distinct outer

metatarsal tubercles; webbing between toes absent; comparative toe length 1 <2<5<3<4,

third toe clearly longer than fifth toe. Skin on dorsal surface, throat and chest smooth; slightly

granular on venter; shanks ventrally granular, possibly marking an area of indistinet and not

sharply delimited femoral glands.

Measurements of holotype. — Al in mm. Snout-vent length, 22.7 (estimated as 25 mm by

VENCES et al. 1999); maximum head-width, 7.7; head length from tip of snout to maxillary

articulation, 9.0; horizontal eye diameter, 2.8; horizontal tympanum diameter, 1.6; distance

from anterior edge of eye to center of nostril, 1.9; distance from center of nostril to snout tip,

1.1; distance between centers of nostrils, 2.6; hand length, 6.0; forelimb length, 14.4; hindlimb

length, 33.8; foot length including tarsus, 14.9; foot length, 9.6; tibia length, 10.4.

Color of holotype in life. — See VENCES et al. (1999). Figure 320 in GLaw & VENCES (1994)

shows the ventral side of the holotype but is mirrored horizontally.

Color of holotype in preservative. - After almost 10 years, the pattern of the holotype is still

fully recognizable (fig. 1). The greenish dorsal and blue ventral color has partly faded and is

much less vivid than in life.

Source : MNHN, Paris

VENCES et al. 17

Fig. 1. — Holotype of Mantella manery (UADBA 7273) in ventral and dorsal view, as photographed in

February 2004, before the application of ventral cuts for gonad examination and tissue removal

from shank muscle. The scale bar represents 10 mm.

PB acknowledges the support by the Volkswagen Foundation through a grant to curate the

amphibian collection of the University of Antananarivo. We are grateful to O. Ramilijaona who granted

ss to this collection, and to the Malagasy authorities for research permits.

LITERATURE CITED

Beura, O., 1993. - The export of reptiles and amphibians from Madagascar. Traffic Bull., 13 (3): 115-116.

DALY, J. W., ANDRIAMAHARAVO, N. R., ANDRIANTSIFERANA, M. & Myers, C. W., 1996. - Madagascan

poison frogs (Mantella) and their skin alkaloids. Am. Mus. Novit., 3177: 1-34

GLaw, F, & VENCES, M., 1994. — 4 ficldguide ro the amphibians and reptiles of Madagascar. 2°% edition.

Küln, Vences & Glaw Verlag: 1-480, 48 pl.

2000. - Mantella manery, M. nigricans und M. milotympanum. Aquarien- & Terrarien-Z., 5 (1): 36-39.

2003. Introduction to Amphibians. Jn: S. M. GoobMAN & J. P. BENSTEAD (ed.), The Natural History

of Madagascar, Chicago & London, The University of Chicago Press: 883-898.

ScHaërer, H.-C., VENCES, M. & Vera, M. 2002. - Molecular phylogeny of Malagasy poison frogs, genus

Mantella (Anura: Mantellidae): homoplastie evolution of colour pattern in aposematie amphi-

bians. Org. Divers. Evol., 2: 97-105

Sraniszewskt, MS. 2001. - Mantellas. Frankfurt, Edition Chimaira, 1-229.

Source : MNHN, Paris

18 ALYTES 22 (1-2)

Vences, M., CHiaRt, Y., RAHARIVOLOLONIAINA, L. & MEYER, A., 2004. — High mitochondrial diversity

within and among populations of Malagasy poison frogs. Mol. Phylogener. Evol., 30: 295-307.

Vexces, M. & GLaw, F., 2003. - Mantella. In: $. M. GooDMAN & J. P. BENSTFAD (ed.), The Natural

History of Madagascar, Chicago & London, The University of Chicago Press: 913-916.

Vexces, M. GLAw, F. & BôHME, W., 1999. — À review of the genus Mantella (Anura, Ranidae,

Mantellinae): taxonomy, distribution and conservation of Malagasy poison frogs. Alytes, 17: 3-72.

ZIMMERMANN, H., 1996. — Der Schutz des tropischen Regenwaldes und ein kleines Frôschchen in

Ost-Madagaskar. Srapfia, 47: 189-218.

Corresponding editor: Alain DUROIS.

SCA 2004

Source : MNHN, Paris

Alytes, 2004, 22 (1-2): 19-37. 19

Developmental pathway, speciation

and supraspecific taxonomy in amphibians

1. Why are there so many frog species

in Sri Lanka?

Alain DuBois

Vertébrés: Reptiles & Amphibiens,

USM 0602 Taxonomie & Collections,

Département de Systématique & Evolution,

Muséum National d'Histoire Naturelle,

25 rue Cuvier, 75005 Paris, France

<dubois@mnhn.fr>

Sri Lanka (and probably also southern India) harbours an unusually

high number of frog species, especially of the direct-developing rhacophorid

genus Philautus. An hypothesis is proposed to try and account for the

exceptional radiation in these frogs: these direct-developers would be

submitted to “familial”, rather than “individual”, mortality, which could

tend to increase allele fixation in isolated populations. Possible ways of

testing this hypothesis, which is neither supported nor rejected by meta-

taxonomic data (mean number of species per genus), are discussed. If

confirmed, this hypothesis could account, at least in part, for some rapid

and massive evolutionary radiations in some zoological groups, like cichlid

fishes, birds and mammals.

INTRODUCTION

Several recent publications have pointed out the discovery that many new species of frogs

remain to be described in Sri Lanka (DUTTA & MANAMENDRA-ARACHCHI, 1996; PETHIYA-

GODA & MANAMENDRA-ARACHCHI, 1998; MEEGASKUMBURA et al., 2002a-b; PENNISI, 2002;

Bossuyr et al., 2004) and probably also in southern India, especially in the Western Ghats

2002). If confirmed, these findings would much more than double the number of frog

n Sri Lanka, and increase significantly the number of amphibian species in India.

Most of these new species are members of the genus Philautus Gistel, 1848, a group of small

tree-frogs belonging, according to the taxonomy adopted, either to the subfamily Rhacopho-

rinae of the Ranidae (DuBois, 1992: Bossuyr & Dugois, 2001) or to the family Rhacophori-

dae (VENCES & GLAW, 2001; WiLkINSON, 2003). These frogs lay egg-clutches in terrestrial

shelters (in leaf litter, under stones or barks, etc.), where these large unpigmented eggs

undergo direct development

Source : MNHN, Paris

20 ALYTES 22 (1-2)

The information so far published on these findings is quite insufficient and unsatisfac-

tory. The only hard data available are cladograms based on genetic sequences in 57 “species”

from Sri Lanka and neighbouring areas (MEEGASKUMBURA et al., 2002a; BossuyT et al., 2004).

These molecular data are not to be found in the papers themselves, but in “Supporting online

material” (SOM) which most readers are unlikely to ever see (see DuBois, 20034). More

importantly, the “new species” are yet to be properly compared (not only from a molecular

point of view, but also in morphology, behaviour, bioacoustics, etc.), diagnosed, described

and named, and the genus Philautus as a whole is still in bad need of a taxonomic revision

(Dusois, 20044). However, despite the paucity of genuine scientific evidence, the high number

of undescribed species in Sri Lanka and southern India is certain. Pending a serious generic

revision of the Sri Lankan and Indian rhacophorines, and the proper description of the

unnamed species, we have to face the fact that Sri Lanka currently harbours more than five

times more frog species than had been believed by former authors (e.g.: GÜNTHER, 1864;

BOULENGER, 1890; KIRTISINGHE, 1957; Durra & MANAMENDRA-ARACHCHI, 1996), and that

probably many more species were present there still one century ago, before the massive

deforestation of this island in the 20°" century (PETHIYAGODA & MANAMENDRA-ARACHCHI,

1998; BAHIR et al., 2002). A similar, although perhaps less extreme, trend also no doubt exists

in southern India, especially in the Western Ghats (Buu, 2002). These two regions (Sri Lanka

and the Western Ghats) have long been considered a single biodiversity region and hotspot,

although they show important faunal differences and should rather be considered two distinct

hotspots (BossuyT et al., 2004).

The discovery that Sri Lanka harbours a batrachofauna much richer than most other

ones in the world, including in various other tropical regions, and possibly richer than any of

them (see PETHIYAGODA & MANAMENDRA-ARACHCHI, 1998: 4), is puzzling, as highlighted by

the journal Science (PENNISI, 2002). The comments on this finding published by this journal,

however, are disappointing, as they do not suggest a serious scientific hypothesis to try and

account for this fact. MEEGASKUMBURA et al. (2002a) simply wrote in this respect: “the

persistence of so many species is striking and may be attributable to a combination of

terrestrial eggs, direct-developing embryos, and high fecundity (up to 91 ova per clutch)”.

How a combination of these three “factors” might explain the unusual high number of frog

species of this region remains a mystery. Most of the comments from “experts” provided by

Science (PENNISI, 2002) on this discovery are not more enlightening regarding the question

“Why are there so many frog species in Sri Lanka?”, a single one being relevant in this respect:

“[Their] water-free lifestyle ‘gives species a lot more latitude, McDiarmid explains, and ‘lends

itself to geographic isolation and speciation”* (PENNISI, 2002: 341). This suggests that terres-

trial direct-development might favour speciation through [ecological?] “latitude” and “geo-

graphic isolation”, but evidence for these two suggestions, and even a detailed explanation of

“how it could work”, are wanting.

To the best of my knowledge, two alternative hypotheses trying to explain the high

number of Philautus species in Sri Lanka have been published. Interestingly, they are quite

opposite. The first one (PETHIYAGODA & MANAMENDRA-ARACHCHI, 1998: 4) relies on the

restricted dispersion abilities of these frogs: “a feature remarkable among the Sri Lankan

Rhacophoridae is the exceedingly small range of distribution of many species, often less than

0.5 km. (...) Das (in litt.) suggests that the high diversity observed might be in part

attributable to their reproductive mode (direct development), which probably restricts their

Source : MNHN, Paris

Dugois 21

dispersion, unlike in species with aquatic eggs or larvae, which could disperse with flooding or

flowing water (high diversity and local endemicity are also observed in the Neotropical frogs

of the genus Eleutherodactylus (Leptodactylidae), many of which breed in phytothelms).”

The second hypothesis, co-signed by the same authors (MEEGASKUMBURA et al., 2002b: 12),

states exactly the contrary: “It appears that direct-developing species have the potential to

undergo rapid adaptive radiation in part through being independent of aquatic habitats,

permitting their dispersal throughout the available expanse of humid-forest.”

As a rule, breeding Philautus populations seem to be quite small (much smaller, at least,

than populations of most frog species in open habitats) and tend to have strongly patchy

distributions, with groups of males calling in bunches of close bushes, separated by large areas

without calling males (repeated personal observations in forests of Sri Lanka, southern India,

Nepal, Thailand and Yunnan). Thus, these frogs are not uniformly distributed on the forest

floor. However, in this genus virtually nothing is known on the population size, distribution,

behaviour and dispersal of non-breeding individuals, in particular of imagos!. The fact that

these frogs do not depend on water bodies for the deposition of their eggs would rather seem

to speak for the absence of natural barriers between populations, which should rather be more

liable than water-breeding species to meet and mix at breeding time in forested areas, but

breeding populations appear to be rather isolated from each other and it is not known whether

some individuals may disperse from one population to another and, if so, what are the

quantitative parameters of such events (frequency, proportions of individuals involved, etc.).

Pending detailed eco-ethological works on these frogs, which are currently wanting, the only

possibility is to make general conjectures. Direct development probably plays a rôle in the

observed phenomenon, associated with the small size, very limited range and semi-isolation,

of many Philautus populations. It would seem that beside the possible, but yet precisely

undocumented, limited population sizes and dispersion abilities of these frogs, another factor

may play a significant rôle in their high speciation rate.

The present paper is devoted to the presentation of an hypothesis that could possibly

account, at least in part, for the seemingly unexpected discovery and, of possible ways of

testing this hypothesis. In a second related paper (DuBois, 2004c), comments are offered on

related matters, in particular regarding amphibian generic taxonomy.

1. An imago (Latin term meaning “image, portrait”; see DURoIs, 1978, 19976) is a specimen similar in aspect 10

the adult, but smaller and sexually immature, which results either from metamorphosis (in species with tadpoles)

or from hatching (in species which develop inside egg capsule). This term should be preferred to the term

“metamorph” sometimes found in the literature for several reasons: (L) it has more generality, as it applies 10

species with “direct development” which do not show proper metamorphosis, but rather a continuous develop-

ment from embryo to imago: (2) the term “metamoph "is unclear in meaning and confusing. This latter term has

never been properly introduced into scientific literature as a new technical term, but simply used, without formal

definition, but then in three distinet senses: (a) to designate specimens during the process of metamorphosis: (2)

to designate metamorphosed specimens as opposed 10 larvae; (3) to designate metamorphosed specimens as

opposed to “neotenic” or “’paedomorphic” ones, in species or genera that show both kinds of developments.

Similarly, the unambiguous adjective maginal (derived from image) should be used instead of the term

“metamorphic”, which is primarily a geological term referring 10 metamorphism and whose use in Z0ology is

confusing for the reasons mentioned above.

Source : MNHN, Paris

22 ALYTES 22 (1-2)

ARE THERE INDEED MORE FROG SPECIES IN SRI LANKA THAN ELSEWHERE?

Before discussing a possible hypothesis for the facts observed, the first question to ask is

whether these facts are indeed exceptional. Although in the first part of the 20° century a

number of biologists, including some zoologists, seemed to be confident that most of the

living animal species of our planet had been discovered and named, except in a few “obscure”

groups considered to be “of little interest”, this idea is now completely abandoned. In the last

decades, a number of studies have been devoted to this question and, although estimates are

difficult and poorly reliable, it is now widely acknowledged that only a small proportion of

these species have yet been recognized by zoologists: a conservative estimate in this respect is

that, with about 1.75 million species currently recognized as “valid” by taxonomists (although

not really “known”, see DuBois, 2003c), the latter have only surveyed about 10 % of the total

number of animal species still living on our planet, perhaps even much less (HAMMOND et al.,

1995). This general estimate covers a very heterogeneous situation, as only a few groups of

vertebrates (particularly the birds) can be considered “well surveyed”, most higher taxa being

“poorly” or “very poorly surveyed”. Vertebrates as a whole are often considered to be “rather

well surveyed”, and, a few decades ago, many authors would have considered that this applies

in particular to the living Amphibia, whose total number was believed to be rather low, a few

thousands only. This was merely a reflect of the bad standard of amphibian taxonomy

worldwide. In the second half of the 20" century, a strong increase in the number of known

species followed the increase of field work in various parts of the planet, expecially in tropical

regions, and the introduction of new taxonomic concepts and methods (Dugois, 1998). As

shown in table 1, the number of species recognized as valid by taxonomists has drastically

increased in the last decades, and this trend should go on, at least as long as research positions

and funds are available for this work, which is not certain (see DuBois, 1998, 2003c). Another

way to realize how bad the amphibian species of our planet are known is to consider that, of

4536 amphibian species described by zoologists by the end of 2000, no less than 20.9 % were

only known from a single locality, and only 75.8 % from more than two localities (tab. 2): had

not a little more than 1000 localities been visited at least once, the number of amphibian

species recognized by taxonomists would be one quarter lower than now. Furthermore, an

important number of the species yet reported from a single locality (the type-locality) are

currently known from a single specimen (the holotype); however, the sources used to compute

the figures in tab. 1-2 are too incomplete to allow a reliable quantitative estimate in this

respect.

In 2003, 5441 amphibian species were recognized (4761 Anura, 515 Urodela, 165

Gymnophiona), but, given the current rate of increase (tab. 1), it is reasonable to predict that

z0ologists have not yet collected, studied, described and named half of the amphibian species

that still live on our planet, perhaps even much less, and since many of these species are

currently threatened with extinction, a large proportion of them will probably disappear

during our century before having been even encountered by man, or at least by taxonomists

(Dugois, 19974, 2001, 2003a).

Source : MNHN, Paris

Table

1. — Number of species of living amphibians considered valid by taxonomists at different dates, according to several checklists or checklist updates, and average rate of

increase in this number per year, during the history of amphibian taxonomy (see DUBOIS, 1987b: 101). The estimate for the year 2000 was obtained by adding the species

reported in the Zoological Record as having been described as new from 1997 (GLAW et al, 1998) to the end of the year 2000, Date: last year covered by the checklist or the

checklist update.

oral number [Average year increase | Average proporional | Average yearly increase! Average proportional | Average yearly increase

Date Reference of species im species mumber since | year inerese since |in species number since| early increase since | in species mumber since

Amphibia preccding date receding date 1768 1768 1969

1768 | LAURENTI, 1768 57 = _ = = = = ©

1854 | DuméRiL et al., 1854 234 2.06 361% 2.06 361% = = ë

1882 | BourENGER, 1882a-b 1003 27.46 114% 830 14.56% = = ë

1969 | Goram, 1974 3343 26:90 2.68 % 16.35 28.68 % = =

1984 | Frosr, 1985 4015 44.80 1.34% 18.32 32.14% 44.80 134%

1992 | DuELLMAN, 1993 4522 6338 1:58 % 19:93 34.96 % 5126 153%

1997 | GLAW et al., 1998 4975 90.60 2.00 % 21.48 37.68 % 58.29 1.74 %

2000 | This paper 5208 77.67 1.56 % 22.20 38.95 % 60.16 1.80 %

2003 | DueLMaN & ScHLAGER, 2003 5441 77.67 149% 291 40.19% éi71 1:85 %

Source : MNHN, Paris

24 ALYTES 22 (1-2)

Table 2. — Information on the number of localities from where 4536 amphibian species had been reported at the

end of 2000. This table was computed from the same sources as in tab. 1, where the relevant data are

lacking for many species, hence the total number of species lower than in tab. 1.

Number of localities from which the species has been reported Number of species | Percentage of species

A single locality (type-locality) 949 209%

nity”, or to localities only 151 33%

Type-locality and “

More than two localities 3436 758%

Thus the question may be asked, whether the situation encountered in Sri Lanka (and

possibly also in southern India) is indeed exceptional, or only results from the amphibian

fauna of these areas having been particularly neglected until now, which is certainly true

(Duois, 1999, contra INGER, 1999). A tentative reply can be obtained by looking at some

figures. According to GORHAM (1974), 3343 amphibian species were recognized as valid by

taxonomists in 1969, and this number has raised to 5441 in 2003 (tab. 1): thus the increase over

this 34-year period was of 2098 species, i.e. 62.8 % of the 1969 figure. The number of species

occurring in Sri Lanka considered as valid by KiRTISINGHE (1957; followed by GORHAM, 1974)

was 35; according to DuTraA & MANAMENDRA-ARACHCHI (1996), this number had risen to 53;

now, according to PETHIYAGODA & MANAMENDRA-ARACHCHI (1998), the inclusion of the new

species discovered in Sri Lanka before 2000 (but not yet described) is about 131, i.e. a increase

of about 274.3 % of the 1969 figure over the 34-vear period 1969-2003. Even if these figures

are approximate and possibly exaggerated (but also possibly underestimated), it is quite clear

that the order of magnitude in the increase of species is much higher in Sri Lanka than the

average rate over the whole planet. A similar trend was identified in southern India

(Buu, 2002). A similar increase seems to have been observed in a single other region of

the world, central and southern America, where a major contribution to this increase is due

to the description of many new species of the genus Æleutherodactylus over the recent

decades.

However, a strong increase in the number of recently discovered species has also been

observed in other tropical regions of the world, and is therefore not by itself evidence that the

total number of species of Sri Lanka and southern India is exceptionally higher. Evidence in

this respect comes from a rough estimate of the number of known species per surface in a few

“megadiversity” countries of the world, as presented by PETHIYAGODA & MANAMENDRA-

ARACHCHI (1998): the species density per 1,000 km? was estimated as 0.06 in Brazil and India,

0.09 in Zaire, 0.13 in Indonesia, 0.22 in Venezuela, 0.36 in Colombia, 1.3 in Ecuador, 2.75 in

ica and 3.9 in Sri Lanka. Even if such estimates are not directly comparable, as they

do not take into account various parameters that are likely to influence species diversity (such

as latitude, altitude, climate or vegetation type), they also point to a difference in the order of

magnitude in the number of species for a given surface between Sri Lanka (and southern

India) and other tropical countries.

Another important consideration is that, of the 131 species estimated by PETHIYAGODA.&

MANAMENDRA-ARACHCHI (1998), 93 (i.e., 71 %) are reported to be “rhacophorid species”,

and that the vast majority of the latter are likely to be members of the genus Philautus, as

Source : MNHN, Paris

Duois 25

defined by Dugois (1987) and Bossuyr & DuBois (2001). It is therefore very likely that the

exceptional amphibian radiation observed in Sri Lanka is mostly, if not only, due to unusual

species diversity in this genus, but not in all other genera, including endemic ones of Sri Lanka

(Adenomus, Lankanectes, Nannophrys) (DUTTA & MANAMENDRA-ARACHCHI, 1996;

MANAMENDRA-ARACHCHI & PETHOYAGODA, 1998; VENCES et al., 2000; DuBois & OHLER,

2001a). The situation is similar in southern India, at least in the Western Ghats (Buu,

2002).

For the purpose of the present discussion, we will consider it very likely that Sri Lanka

(and possibly southern India), mostly on account of the genus Philautus, just like central and

southern America on account of the genus Æ/eutherodactylus, do indeed harbour exception-

ally high numbers of amphibian species, many of which are very similar in aspect and have a

very limited distribution, both factors that certainly contributed to the long underestimation

of the number of frog species in these areas. If we consider this fact as most likely, what could

be its explanation?

AN EVOLUTIONARY HYPOTHESIS

The vast majority of the new frogs recently discovered in Sri Lanka (and southern India)

belong in a single genus, the tree-frog genus Philautus Gistel, 1848. As redefined by DUBois

(1987, 1992) and reviewed by Bossuyr & DuBois (2001), this genus now only includes

direct-developing frogs. In frogs, “direct development”, sometimes called “endotrophy”

(e.g., MCDrarMiD & ALTIG, 1999), designates a mode of development that skips the usual

free larval stage of anurans, the embryo’s growth and differentiation being supported only by

the resources that were available from the start within the envelopes of the egg, as vitelline

reserves. In the genus Philautus, such eggs are not deposited isolated, but as groups or

“clutches” of eggs usually hidden under terrestrial shelters (under stones, leaf litter, tree

barks, or in holes). During the whole development of the eggs, the latter remain together

in this shelter; at hatching, the imagos leave the eggs and disperse on the ground and

in the surrounding vegetation. The hypothesis proposed here is that these developmental

particularities, by themselves, constitute particular ecological conditions likely to facilitate

speciation, through a mode of mortality that is different from that usually encountered in

frogs.

et al. (2002) presented as à novelty the finding, shown in their molecular cladogram, that

ally” referred to the rhacophorid genera Theloderma Tschudi, 1838 and Rhacophorus

Kuhl & Van Hasselt, 1822 are not closely related 1 the other species of these two genera but are closely related

to those of the Sri Lankan species of the genus Philautus. This statement deliberately ignored several previous

publications where the same hypothesis had already been proposed, without any use of molecular data: thus,

Peters (1860), AuL (1931) and KiRTISINGHE (1957) had already placed the species Polhpedates schmarda Kelaart.

1854 (referred to Theloderma by Liëm, 1970, DUTrA & MANAMENDRA-ARACHCHI, 1996 and BOSSUYT & DUOIS,

2001) in the group now known as Philautus, and Durois (1987, 1992, 1999: Boss pis. 2001) had

already removed all Sri Lankan species placed by earlier authors in Rhacophorus from that genus, 10 place them

in Philautus. Actually, maintaining these latter species in Rhacophorus (as done e.g. by DUTTA & MANAMENDRA-

ARACHCHI, 1996 and PETHIYAGODA & MANAMENDRA-ARACHCHI, 1998) was already obsolete much before the

Science paper (DUOIS, 1999), and the later should rather have stated that it confirmed the validity of this action

rather than presenting it as new.

the Sri Lankan speci

Source : MNHN, Paris

26 ALYTES 22 (1-2)

This hypothesis was already proposed earlier, as follows: “The particularities of intra-

and interspecific variation in [the genus Philautus] (intraspecific variability often higher than

morphological differences between related species), where ‘sibling’ species (dualspecies)

often have very different calls (personal observations in southern India), might be related to a

particular mode of natural selection, connected with the reproductive and developmental

modes of these species. As a matter of fact, in the species that lay numerous eggs in water, the

tadpoles later disperse more or less, and are all submitted similarly to selection, which results

in a roughly Gaussian distribution of characters in the population. In contrast, in Philautus

and in other groups with terrestrial clutches, containing a small number of eggs, the latter are

certainly submitted to largely random but massive mortality: a given clutch, deposited by a

female, runs the risk of being discovered by a predator, which then can destroy it completely,

but it can also remain undiscovered and reach safely overall eclosion.” (Dugois, 1987: 71,

translated).

For more clarity, we may consider an hypothetical and very simplified example. Let us

compare the sympatric populations of two different frog species of the same size, having

similar demographic conditions, i.e. a reproductive population of 5 males and 5 females, each

female pairing with a single different male and laying 10 eggs, that will develop into 5 males

and 5 females, and all adults dying after first reproduction. Let us further hypothesize that

both populations are completely isolated, i.e. without immigration or emigration during the

period considered. Species A lays its eggs in water, where they hatch after embryonic

development, giving birth to tadpoles that spread in the water body, where they live randomly

distributed, until they metamorphose into imagos. Species B lays eggs clutches under terres-

trial shelters, where the eggs undergo direct development until they hatch as imagos. Let us

now consider that, in both populations, mortality between egg-laying and the stage imago is

80 %: i.e., in both populations, 50 eggs are laid, 10 of which only reach the stage imago. Let us

consider that this mortality is caused by predators, e.g. snakes. In population A, snakes will eat

40 tadpoles among the 50 randomly distributed in the pool, whereas in population B they will

discover and eat 4 egg-clutches out of 5. It is quite clear that, if the only surviving clutch bears

special characters, these will be widely distributed in the frogs resulting from this clutch, much

more than in the population with tadpoles.

In some extreme situations, one generation may be enough to result in the total replace-

ment of one allele by another in a population. This is the case e.g. if a mutation takes place in

a sex-linked gene borne by the heterogametic chromosome, especially if this mutation occurs

very early in the germ-line, ideally in the first primordial cell at the origin of the whole

germ-line of an embryo. In anurans both male and female heterogamy do occur (DUELLMAN

& TRUEB, 1985: 447, 450). The situation in Philautus in unknown, but let us hypothesize that

in this group, like in several studied ranids, the heterogametic sex is male (XY/XX type). If a

mutation » occurs in the Y chromosome of the first primordial cell of an early embryo, all

spermatozoa resulting from the divisions of this cell and bearing the Y chromosome (i.e., half

of the spermatozoa of this individual) will bear the "1 allele, and all males resulting from

fertilization of eggs by these spermatozoa will bear the mutation 1. So, among our 5

hypothetical females, one will produce 10 embryos, all 5 males of which will bear m, whereas

the 20 males produced by the other nine females will not. Now, under the schematic model

developed above, the fate of the 5 m-bearing males will be very different in the two species. In

the species with tadpoles, mortality among the 25 males will be random, and the probability

Source : MNHN, Paris

Dugois 27

that the 5 surviving tadpoles bear m will be 5/25 x 4/24 x 3/23 x 2/22 x 1/21 = 120/6,375,600

= 0.000019: thus the complete fixation of m in one generation will be a very unlikely event. On

the other hand, in the direct-developing species, the probability that the 5 surviving males be

bearers of m will be 1/5 = 0.20. Thus, in this very special case, a single generation could easily

allow fixation of a mutation in a population in a direct-developing species, whereas the same

event would be very unlikely in a tadpole-developing species. As it is known that, in some

cases, speciation can result from a single mutation in a single locus (see references and

discussion in Dumois, 1988: 42), it is obvious that, in this example, speciation could be

facilitated by the mode of mortality, which may be qualified of “familial” in direct-developing

frogs, vs. “individual” in species with tadpoles.

Of course, this example is very schematic and simplistic, as the same result would not be

obtained if an autosomic or homogametic sex chromosome was involved: in this case, even

with the same demographic figures, several generations would be needed to result in the

fixation of a new allele in the population, and then many other factors would interfere, such

as population effective breeding size, population range, dispersal (immigration and emigra-

tion), longevity, “selective values” of the initial allele and of the mutation m, etc. Many

models could be computed using various values for all these parameters, but they would be of

little interest as long as we do not have more information on the actual values of these

parameters in the populations of frogs considered. It is clear, however, that familial predation

on all eggs of a female at once (or survival of all these eggs altogether) entails different results

from random mortality of individuals in a mixed population. Could this factor explain the

seemingly higher speciation rate in Sri Lankan Philautus than in other frog groups? There are

several ways to test this hypothesis. One is to have a look at some metataxonomic data (as

defined by Dugois & OHLER, 2001).

DEVELOPMENTAL MODE AND SPECIATION IN FROGS

Early anuran development can follow several rather different pathways (see eg.

MCDiaRMID & ALTIG, 1999). A majority of anuran species have free aquatic tadpoles that are

“exotroph”, ie. that feed on bacterial, vegetal or animal resources found in the aquatic

environment where they live. As this mode of feeding requires a behavioural and energetic

investment for foraging, it can also be called ergotrophy (from the Greek ergon, work”). The

transition from the egg-enclosed embryo to the imago through such a free larval stage with

active feeding is widespread, dominant and probably plesiomorphic in amphibians (but see

BoGarT, 1981), whereas other developmental modes are all apomorphic relative to the

former. These derived modes of development are often collectively designated as “endotro-

phy” (eg, THIBAUDEAU & ALTIG, 1999), which is incorrect as in some of them only the

feeding is really internal (inside the egg), whereas in some others it comes from the parent or

from brothers and sisters, i.e. from outside the egg (although inside one of the parents). It

seems better to use the unambiguous term /ecithotrophy (WouRMs, 1981) for feeding only

upon the internal vitelline resources of the egg. For the more general category of all

developmental modes that are not dependent from foraging for external feeding, I propose the

new term argiotrophy (from the Greek argia, “idleness, inaction”). This category includes

Source : MNHN, Paris

28 ALYTES 22 (1-2)

species whose development takes place either within the genital tract or another pouch in one

of the parents, or within the egg capsules, the eggs being deposited in some terrestrial or

arboreal shelter. As discussed in more detail in a second paper (Dupois, 2004c), this category

is heterogeneous as far as developmental pathways are concerned, but from an ecological

point of view and for the purpose of the present discussion, it is a relevant category, as in all

these cases the following conditions are met: all eggs of a clutch remain together during a large

portion of their development, either as a clutch hidden in some shelter, or kept within the

adult; during all this part of their development, these eggs are likely to be either discovered

and destroyed altogether, or to remain undiscovered and safe. Thus all these cases are

submitted to familial, not to individual, mortality.

The development of many species of anurans being still unknown, no complete review of

the two major ecological categories of frogs regarding developmental mode is possible for the

time being, but the information available, as gathered by ALTIG & MCDirARMID (1999), is

presented in table 3. The taxonomy of amphibians being in constant change, the precise

figures of such a table are bound to be obsolete before being published, but the general trends

are likely to remain the same, at least for a few years. To prepare this table, a taxonomy slightly

modified from the list in DUELLMAN & SCHLAGER (2003: 456-484) was followed, and each

anuran genus was referred to either of four ecological categories, defined as follows: (T)

genera known to have free aquatic tadpoles (at least briefly described in at least one species):

(A) genera known to have another mode of development (argiotrophy), without free aquatic

tadpoles (at least briefly described in at least one species): (B) genera with both categories

(among the species currently referred to the genus, at least one is known to have free aquatic

tadpoles, and one to be argiotroph); (U) unknown (the development of all species of the genus

is currently unknown).

Information on the development is available for at least one species of 325 anuran genera.

Among them, 227 genera (i.e., 69.8 %) are known to have at least one species with free

tadpoles but no reported argiotroph species; 93 genera (i.e., 28.6 %) are known to have

argiotroph species but no reported species with free tadpoles; and only 5 genera (1.e. 1.5 %)are

considered to include both kinds of species.

The argiotroph species are not randomly distributed among anurans. The latter are

divided by a number of recent authors (e.g., SOKoL, 1977) in two groups or suborders, the

Discoglossoidei and the Ranoideï*. Interestingly, argiotrophy is much rarer in the Discogl

soidei, where it is known in 7.7 % of the genera (2/26) against 30.2 % (98/325) in the Ranoïdei,

3. This list is unreliable for several groups, as some taxa appear twice in different parts of the classification (e.g.,

Syncope or Ingerana baluensis), some species are misplaced according to the classification chosen (e.g.. in the

genera Hoplobatrachus, Limnonectes, Megophrys, Philautus or Rana), some names (e.g.. Bombina) are lacking

altogether whereas others are listed as valid without explanation although they are currently considered junior

subjective synonyms (e.g.. in the genera Amolops, Bufo, Limnonectes, Philautus or Rana). Strangely enough, this

st is not always consistent with the taxonomies presented for the families in the chapters of the book itself

His et al., 2003). For example, in the Ranidae the information concerning several taxa (e.g.. Amolops.

Elachyglossa, Fejervarva, Ingerana, Limmonectes, Occido=yga, Odorrana, Sphaerotheca or Strongylopus) are not

compatible with those in the chapter devoted to this family (Dupois, 2003b). In tables 3-4 here, the family

Ranidae is understood as including the eleven subfamilies listed in the latter chapter, as well as the subfamilies

Mantellinae and Rhacophorinae, This conservative approach seems best until a robust phylogenetic hypothesis

is agreed upon by many workers concerning the relationships between al hese groups

4. These suborders are sometimes called (e.g.. FELLER & HEDGES, 1998) Archacobatrachia and Neobatrachia,

but ne two names are invalid, being junior homonyms (Dusois, 1984, 2004b).

Source : MNHN, Paris

Table 3. - Some data on the higher taxa (suborders and families) of anuran amphibians: number of

known genera and species (slightly modified from DUELLMAN & SCHLAGER, 2003; see note

3), developmental

modes (slightly modified from ALTIG & MCDIARMID, 1999).

Developmental modes of genera (see text for details): T, ergotroph with free tadpoles; A,

argiotroph; B, both argiotroph and ergotroph with free tadpoles developmental modes

reported in genus: U, unknown.

Number of Number of genera (and of species in these genera)

Suborder Family genera with given developmental mode

(and species) 5 + 5 ml

Discoglossoidei Ascaphidae 10) 1) 0 0 0

Bombinatoridae 2 (10) 18) 0 0 1@)

Discoglossidae 2(10) 2(10) 0 0 0

Leiopelmatidae 16) o 1) 0 o

Megophryidae 11027) HG27 o 0 0

Pelobatidae 304) 31) L 0 0

Pelodytidae 16) 16) L) 0 0

Pipidac 560) 403) Q 1) o

Rhinophrynidae 14) 10) o 0 o

Total 27 (198) 24(185) 14) 10) 1)

Ranoïdei Allophrynidae 10) 0 0 o 10)

Antroleptidae 66) 41) 265 0 °

Brachycephalidae 16 0 1@) o 0

Bufonidae 35 (448) 16589) 1553) o 46)

Centrolenidae 3 (56 3 (136) 0 o o

Dendrabatdae | 10 (201) 9(08) ° 1003) 0

Heleophrynidae 16) 16) 0 0 0

Hemisotidae 10) 10) 0 o 0

Hytidae 43823) 35 (736) 460 16) 300)

Hyperoiidae 19 (48) 15043) 0 o 465)

Leptodactylidae 49 (1085) 3325) Us) 1) 20)

Limnodynastidae 1049) 8(45) 2(4) o o

Microhylidae 66 (356) 270118) 36 234) o 44

Myobatrachidae 1373) 6(65) s(6) ® 2(2)

Ranidae 61 (1040) 45 798) 11170) 1 (68) 46)

Rhinodermatidie 1@) ° 1@) 0 0

Sooglossidae 26) 0 26) o o

Toul 323 (4563) 203 (5010) 92 (1296) 4023 2469

Total e 350 (4761) 2276195) 95 (1300) 5030) 2566)

Source : MNHN, Paris:

30 ALYTES 22 (1-2)

a matter that should call future attention from the phylogenetic point of view. The only

two genera of Discoglossoidei in which some species are reported to be argiotroph are

Pipa Laurenti, 1768 (where embryos develop on the back of the female and rely on

their vitelline reserves alone for development) and Leiopelma, with two different kinds of

argiotrophy (with free non-feeding tadpoles in dorsal pouch of father and with direct

development within egg capsule). Besides, THIBAUDEAU & ALTIG (1999: 172) listed the

Megophryidae among the families including at least one “endotroph” species, but this

was based on a misidentification of direct-developing eggs of Philautus aurifasciatus

(Schlegel, 1837) as Xenophrys longipes (Boulenger, 1885), a mistake corrected by LEONG &

CHou (1998).

In contrast, in the Ranoidei, a vast array of argiotroph developmental pathways have

developed. The distribution of argiotrophy within the various families follows no clear or

consistent pattern: this category is found in various groups that have no direct cladistic

relationships, which suggests that these derived modes of development appeared indepen-

dently in these groups and are therefore homoplasic. This was precisely documented in some

cases only (MARMAYOU et al., 2000), but is very likely in several others. In a few cases however,

retention of a silent “direct development program” in tadpole-developing species, or the

reverse, probably occurred (see DuBois, 2004c).

Argiotroph species are reported only in 13 of the 20 families currently recognized in the

Ranoïdei. Among the 299 genera of Ranoïdei for which information is available for at least

one species, 203 (i.e., 67.9 %) are known to include only species with free aquatic tadpoles, 92

(i.e., 30.8 %) are known to include only argiotroph species, and 4 (1.e., 1.3 %) are considered to

include both.

The hypothesis presented above is that taxa (genera, families) including species confron-

ted with “familial” mortality would tend to have higher rates of speciation than taxa with

species submitted to “individual” mortality. An empirical confirmation of this hypothesis

would be provided if anuran genera including argiotroph species had a higher mean number

of species than genera with free tadpoles. As a first apparent confirmation of this trend, the

most speciose anuran genus is the direct-developing Æleutherodactylus Duméril & Bibron,

1841, which, with about 680 species known in 2003 (and perhaps as many yet to be discovered

and described), is also the most speciose genus of all vertebrates. However, this trend is not

confirmed over the whole group of anurans, at least in the current state of knowledge. Over

the 325 anuran genera for which developmental data are available (tab. 3), the mean number

(x + s) of included species is 14.1 + 34.3 (range 1-326) for the 227 genera that include only

species with free aquatic tadpoles, and 15.6 + 69.9 (range 1-682) for the 98 genera that include

at least one argiotroph species. The difference is not statistically significant (Mann-Whitney U

test: U=9776.5, P = 0.09), but this is of little meaning as a large majority of the anuran genera

include very few species. Table 4 gives the number of known species of the 43 most speciose

genera of anurans (i.e., including more than 20 species), with their known modes of develop-

ment: here also, the mean number of ES s higher in the 10 genera including at least one

argiotroph species (114.0 + 201.3, range 22-682) than in the 33 genera known to include only

species with free aquatic tadpoles (66.5 + 69.4, range 21-326), but, given the large variance in

each group, the difference is still not statistically significant (Mann-Whitney U test: U = 155,

P=0.77).

Source : MNHN, Paris

DuBois

Table 4. — Some data on the 43 genera of anurans with the highest numbers of species (from the

same source as in table 3). Developmental modes of genera (see text for details): T,

ergotroph with free tadpoles; A, argiotroph; B, both argiotroph and ergotroph with free

tadpoles developmental modes reported in genus; U, unknown.

Rank Family Genus Number of species _ | Developmental mode

1 Lepiodactylidae Eleutherodachylus Duméril & Bibron, 1841 682 A

2 Hylidae Hyla Laurent, 1768 526 T

3 Bufonidae Bufo Laurent, 1768 247 T

4 Ranidae Rana Linnaeus, 1758. 21 E

5 Hyperoliidae Hyperolius Rapp, 1842 n7 T

6 Hylidue Lisoria Tschudi, 1838 m2 T

T Dendrobatidae Colestethus Cope, 1866 103 B

E Hylidae Scinax Wagler, 1830 #7 T

O Ranidae Philautus Gistel, 1848 EE A

10 Bufonidac Arelopus Duméril & Bibron, 1841 74 %.

in Ranidac Rhacophorus Kuhl & Van Hasselt, 1822 œ T

n Ranidae Phrynobatrachus Gunther, 1862 6 Tv

5 Ranidae Mantidactylus Boulenger, 1895 68 B

4 Leptodactylidae Leptodactylus Fitzinger, 1826 e T

15 Centrolenidae Cochranella Taylor, 1951 a T

16 Ranidae Limnonectes Fitzinger, 1843 5 R

17 Hyperoliidae Lepiopelis Günther, 1859 El T

18 Ranidae Platymantis Gunther, 1859 50 A

1 Ranidae Boophis Tschudi, 1838 47 T

1 Ranidae Prychadena Boulenger, 1917 47 T

19 Leptodactylidse Telmatobius Wicgmann, 1835 47 T

2 Hylidae Gastrotheca Fitzinger, 1843 46 B

3 Lepiodactylidae Physalaemus Fitzinger, 1826 ai T

24 Centrolenidae Centrolene Jiménez de la Espada, 1872 40 T

25 Ranidae Amolops Cope, 1865 36 T

26 Centrolenidae |" Hpatinobatrachium Rutr-Carranza & Lynch, 1991 35 T

27 Megophryidae Scutiger Theobald, 1868 El mn

28 Dendrobatidae Dendrobates Wagler, 1830 5 T

29 Hyperoliidae Afrixalus Laurent, 194 32 T

30 Leptodactylidae Phrynopus Peters, 1874 5 A

3 Ranidae Odorrana Fei, Ve & Huang, 1991 30 T

El Microhylidae Cophixalus Boetiger, 1892 29 A

32 Dendrobatidae Epipedobates Myers, 1987 2 T

32 Hylidae Phyllomedusa Wagler, 1830 2 T

35 Ranidae Paa Dubois, 1976 27 ï

36 Microhylidae Oreophryme Boettger, 1895 26 A

37 Leptodactylidae Cycloramphus Tschudi, 1838 25 S

38 Microhylidae Microhyla Tschudi, 1838 24 1

El Hylidae Nictimystes Stejneger, 1916 2 1

ET Myobatrachidae Uperoleia Gray, 1841 24 T

a Arthroleptidae Schoutedenella de Witte, 1921 2 A

42 Bufonidae Ansonia Stoficzka, 1870 a T

42 Megophryidae Megophrys Kuhl & Van Hasselt, 1822 a 1

Source : MNHN, Paris

32 ALYTES 22 (1-2)

Such an empirical approach to this question has only a very limited value, for several

reasons. First, the category of argiotrophy is ecologically rather homogeneous regarding the

question here posed (at least, all species in this category are likely to be submitted to “familial”?

mortality during development), but rather heterogeneous in developmental terms, as dis-

cussed in more detail elsewhere (Dugois, 2004c). Information available on detailed develop-

mental pathways is currently too scanty in most genera without free aquatic tadpoles to allow

for a more detailed analysis. For the time being, data are insufficient to allow to test

statistically the existence of significant differences regarding mean species numbers in genera

having different developmental pathways within the ecological category of argiotrophy.

Second, comparison of the number of species per genus would make fully sense only if all

taxonomists were using the same “genus concept”. However, despite precise proposals in this

respect (DuUBoIs, 1988), there currently exists no consensus among zootaxonomists about

“what is a genus”, and there is no reason to think that the various genera of anurans are

“equivalent” by any standard (for a detailed discussion of this concept of taxonomic

equivalence, see DuBois, 1988: 59-67). Clearly, some genera (e.g., Hyla, Mantidactylus

or Rhacophorus) are rather heterogeneous assemblages that will most likely be dismantled in

the future, as was the case for Rana in the recent decades (see Duois, 2003b). Others appear

to be more homogeneous groups that may keep their status of genera in the future (e.g., most

of the genus Bufo). This question also is tackled again in more detail elsewhere (DuBois,

2004c).

Another major problem comes from the fact that all genera have not been submitted to

the same effort of work in the recent decades. A striking fact for all experienced taxonomists

is that the taxonomy of some frog genera is more “difficult” than that of others, because they

show both a large overall similarity between species and unusual patterns of variation (with

some of the interspecific variation overlapping intraspecific variation). This no doubt has

acted as a break against their recent taxonomic revision. Among such genera, although not

alone, are some genera of argiotroph species, such as Philautus mentioned above, or the

African Arthroleptis-Schoutedenella complex. The possibility is strong that revision of such

genera, using morpho-anatomical, molecular, bioacoustic and cytogenetic characters, might

disclose the existence of many more species than is actually believed. For these reasons, this

empirical approach does not allow to really test the evolutionary hypothesis presented above.

Finally, and perhaps more importantly, comparisons as made above are likely to be

statistically invalid as they do not rely on phylogenetic information. To be significant, such

comparisons should use cladograms as input or be made between sister-taxa, but the

information available on the phylogenetic relationships between the 325 anuran genera

considered above is too incomplete to be used in this analysis, and restricting the compar

to the few groups of genera for which reliable cladistic data are available would not allow

genuine statistical comparison as the numbers would be much too low. However, this question

should be kept in mind for the future, and considered again when our understanding of

phylogenetic relationships between anuran genera is well improved.

For the time being, there are other possible ways to test the hypothesis presented above.

As suggested above, models utilizing various populational, ethological and ecological

parameters could be devised to investigate the theoretical likeliness that argiotrophy might

facilitate speciation in frogs.

Source : MNHN, Paris

DuBois 33

Another approach would be through biological comparisons within couples of phyloge-

netically related sympatric species of similar size and natural histories (except developmental

mode), one of which lays clutches of eggs that give birth to free aquatic tadpoles, whereas the

other one has another developmental mode, either in some external shelter or in some pouch

of one of the parents. Several parameters may be considered for such comparisons, such as

genetic polymorphism, heterozygosity and “genetic variance”, measured e.g. with the F,;

fixation index of WRIGHT (196$), or also cytogenetic differentiation. If the hypothesis above

is correct, argiotroph species should show a significant tendency to allele fixation in small

isolated populations. This does not necessarily imply that they would show significantly

different mean genetic polymorphisms or heterozygosities than species with free tadpoles,

because if predation on clutches is random the net effect on allele frequencies will be zero over

the course of successive generations. On the other hand, if the populations are indeed quite

isolated and small, they would tend to show local genetic drift and genetic variance between

them should be more important than between similar populations of species with free

tadpoles.

Empirical data to support or refute this hypothesis are lacking, as until now argiotrophy

does not seem to have been particularly discussed as a pertinent factor in speciation rate,

genetic polymorphism and evolutionary patterns in amphibians. WRiGHT'’s (1951) theories on

relationship between population characteristics and genetic structure would seem a good start

for such works. This was the case in INGER et al. (1974)'s study dealing with several popula-

tions of Malaysian bufonids and ranids: evidence was found for lower genetic variation in

species with linear distribution along streams and breeding among neighbours than in species

with large panmictic breeding aggregations. Unfortunately, this nice study was not followed

by others in other areas, that would have allowed to increase the sample size and test the

generality of these findings. More data are available in Urodela, but here also no study has yet

focused on a detailed comparison between related and sympatric ergotroph and argiotroph

species. In plethodontids, argiotroph taxa show great spatial heterogeneity and very high

genetic variance between populations, although local heterozygosity may be relatively low

(LARSON, 1984; LARsON et al., 1984b), which is congruent with the hypothesis presented

above. The highest heterozygosities in argiotroph salamanders have been found in species with

dense populations (HANKEN & WAKkE, 1982; WakE & YANEV, 1986; GaRCiA-PARIS et al.,

2000). Particularly relevant for the present discussion is the recent study by CRAWFORD (2003)

on mitochondrial and nuclear DNA variation in four Central American species of Eleuthe-

rodactylus, which showed considerable values of genetic variance between populations. This

author also found very large effective population sizes in these species. Applying a molecular

clock model, he concluded that the unusually high species diversity in the genus Æleuthero-

dactylus was probably not due to higher speciation rate but to old age, and he suggested that

“the tropics have functioned as a museum of antiquity rather than as a cradle of speciation”

(CRAWFORD (2003: 2537). However, whether the molecular clock model validly applies to

these taxa remains open to question.

As for cytogenetic differentiation, BOGART (1991) pointed out the importance of

demonstrable karyotypic changes involving modification of chromosome number in the

genus Eleutherodactylus. He also remarked that karyotypic diversity seemed larger in “smaller

genera that contain species with terrestrially developing or direct developing eggs” (BOGART,

1991: 242), such as Arthroleptis, Cardioglossa, Fritziana, Leptopelis or dendrobatid genera. In

Source : MNHN, Paris

34 ALYTES 22 (1-2)

such groups, major karyotypic changes would occur by centric fusion and fission in small,

isolated populations where inbreeding would “fix mutational events in a homozygous condi-

tion” (BOGART, 1991: 254). This model would seem more difficult to apply to large popula-

tions.

Such kinds of comparative studies would be worth undertaking both in frogs and in

salamanders. For more generality, such studies could be carried out in several taxonomic

groups and in different regions and kinds of habitats of the world. To come back to the genus

Philautus, which prompted this reflexion and has never been the matter of detailed demo-

graphic, ecological, genetic and cytogenetics studies, it would appear most crucial to develop

such researches to try and throw more lights on its evolutionary patterns.

Should the hypothesis turn out to be supported, it could have far-reaching consequences.

If “familial” mortality indeed facilitates speciation, this fact might explain in part the high

rates of speciation and of evolution observed in some animal groups displaying parental care,

such as the birds or the cichlid fishes (with their striking radiation in the great African lake:

see e.g. JOHNSON et al., 1996) or true viviparity, such as the mammals. Starting from other

premises, other authors (e.g.: WiLsoN et al., 1975; Bus et al., 1977; WyLes et al., 1983;

LaRsON et al., 1984u; SAGE et al., 1984) already discussed the factors possibly involved in such

cases of rapid speciation, and, although they insisted mostly on the rôle of chromosomal

evolution and of social behaviour, their data are not incompatible with the present hypothesis.

If the latter is correct, the unexpected high number of species of Philautus in Sri Lanka as

compared with the number of frog species in other parts of the world would be accounted for

by the fact that these Sri Lankan frogs are not precisely frogs, at least not usual frogs with

aquatic eggs and larvae, but other “kinds” of animals. In fact, if one forgets the numerous

Philautus species, the amphibian fauna of Sri Lanka does not appear in the least exceptional:

rather it would seem poorer than those of other areas of similar latitude, even in the same part

of the world. Is this because of competition with the unusually successful Philautus clade?

ACKNOWLEDGEMENTS

For comments on previous drafts of this paper, I am grateful to Thierry Lodé, Annemarie Ohler,

Miguel Vences, David B. Wake and an anonymous reviewer. Ronn Altig provided information on the

misidentification of a Xenophrys tadpole in the literature. Annemarie Ohler helped for the statistical

analysis, and Roger Bour for the preparation of tables.

LITERATURE CITED

Au, E., 1931. - Anura III, Polypedatidae. Das Tierreich, 58: i-xvi + 1-477.

ALTIG, R. & McDiarmin, R. W., 1999. - Diversity. Familial and generic characterizations. In: MCDiaR-

MID & ALTIG (1999): 295-337.

Baur, M. MERGASKUMBURA, M. PETHIYAGODA, R., MANAMENDRA-ARACHCHI, K. & SCHNEIDER, C. J.,

2002. — Reproduction in captivity of four species of direct-developing frogs from Sri Lanka

(Anura: Ranidae: Rhacophorinae: Pseudophilautus). Frog Leg, 10: 9.

Source : MNHN, Paris

DuBois 35

Buu,S. D., 2002. — 4 synopsis 10 the frog fauna of the Western Ghats, India. Occasional Publication, Indian

Society for Conservation Biology: 1-24.

BOGART, J. P., 1981. - How many times has terrestrial breeding evolved in anuran amphibians? Monit.

zool. ital., (n.s.), 15, suppl.: 29-40.

se 1991. - The influence of life history on karyotypic evolution in frogs. n: D. M. GREEN & $. K.

SESSIONS (ed.), Amphibian cytogenetics and evolution, San Diego, Academic Press: 233-258.

BossuyT, F. & DuBois, A., 2001. — A review of the frog genus Philautus Gistel, 1848 (Amphibia, Anura,

Ranidae, Rhacophorinae). Zeylanica, 6 (1): 1-112.

Bossuyr, F., MEEGASKUMBURA, M., BEENAERTS, N., GOWER, D. J., PETHIYAGODA, R., ROELANTS, K.,

MANNAERT, A., WILKINSON, M., BAHIR, M. M., MANAMENDRA-ARACHCHI, K., NG, P. K. L.,

SCHNEIDER, C. J., OOMMEN, O. V. & MILINKOVITCH, M. 2004. — Local endemism within the

Western Ghats-Sri Lanka biodiversity hotspot. Science, 306: 479-481.

BOULENGER, G. A... 18824. - Catalogue of the Batrachia Salientia s. Ecaudata in the collection of the British

Museum. London, Taylor & Francis: i-xvi + 1-503, pl. 1-30.

—— 1882b. - Catalogue of the Batrachia Gradientia s. Caudata and Batrachia Apoda in the collection of the

British Museum. Second edition. London, Taylor & Francis: i-ix + 1-127, pl. 1-9.

—— 1890. — The fauna of the British India, including Ceylon and Burma. Reptilia and Batrachia. London,

Taylor & Francis: i-xviii + 1-541.

BusH, G. L., Case, S. M., WILsON, A. C., & PATTON, J. L., 1977. — Rapid speciation and chromosomal

evolution in mammals. Proc. nat. Acad. Sci. USA, 74 (9): 3942-3946.

CrawroRD, A. J., 2003. - Huge populations and old species of Costa Rican and Panamanian dirt frogs

inferred from mitochondrial and nuclear gene sequences. Molecular Ecology, 12: 2525-2540.

Dumoïs, A., 1978. - Les principaux stades de développement significatifs en écologie et en génétique des

populations des Amphibiens Anoures. La Terre et la Vie, 32: 453-459.

= 1984. - La nomenclature supragénérique des Amphibiens Anoures. Mem. Mus. nan. Hist. nat., (A)

131: 1-64.

1987. - Miscellanea taxinomica batrachologica (1). A/ytes, 5 (1-2): 7-95.

1988. - The genus in zoology: a contribution to the theory of evolutionary systematics. Mém. Mus.

natn. Hist. nat, (A), 140: 1-123.

=. 1992, - Notes sur la classification des Ranidae (Amphibiens, Anoures). Bull. Soc. linn. Lyon, 61 (10):

305-352.

1997a. — Editorial: 15 years of Alytes. Alytes, 14 (4): 129.

1997b. — Instructions to authors of papers submitted to Alytes. Alytes, 14 (4): 175-200.

1998. -List of European species of amphibians and reptiles: will we soon be reaching “stability”?

Amphibia-Reptilia, 9 (1): 1-28.

--- 1999. - South Asian Amphibi

South Asian nat. Hist., 4 ( 11.

--- 2001. - Les Amphibiens. Encyclopédie Clartés, Janvier 2001, 4390: 1-20.

2003a. — Histoire de la batrachologie. /n: R. DUGUET & F. MELKI (ed.), Les Amphibiens de France,

Belgique et Luxembourg, Mèze, Editions Biotope: 190-195.

2003b. - True frogs (Ranidae). /n: HUTCHINS et al. (2003): 245-264.

2003c. - The relationships between taxonomy and conservation biology in the century of extinctions.

Comptes rendus Biologies, 326 (suppl. 1): S9-S21

20034. - Editorial. Should internet sites be mentioned in the bibliographies of scientific publications?

Alytes, 21 (1-2): 1-2.

20044. - Book review. Amphibians of Nepal: a few words of caution. Ales, 21 (3-4): 173-180.

2004. - The higher nomenclature of recent amphibians. A/ytes, 22 (1-2): 1-14.

2004c. - Developmental pathway, speciation and supraspecific taxonomy in amphibians. 2. Develop-

mental pathway. hybridizability and generic taxonomy. Alytes, 22 (1-2): 38-52.

Duois, A. & OHLER, A., 2001. — Systematics of the genus Philautus Gistel, 1848 (Amphibia, Anura,

Ranidae, Rhacophorinae): some historical and metataxonomic comments. J South Asian nat.

173-186.

DUELLMAN, W. E., 1993. - Amphibian species of the world: additions and corrections. Unix. Kansas Mus.

Nat. Hist. spec. Publ., 2: [ii] + i-iüi + 1-372.

DUELLMAN, W. E. & SCHLAGER, N., 2003. - Amphibians species list. /n: HUTCHINS et al. (2003): 456-489.

a new frontier for taxonomists. Invited editorial / Book review. J

Source : MNHN, Paris

36 ALYTES 22 (1-2)

DuELLMAN, W. E. & TRUEB, L., 1985. — Biology of amphibians. New York, McGraw-Hill, 1986": i-xix +

1-670.

DumériL, A.-M.-C., BIBRON, G. & DUMÉRIL, A., 1854. - Erpétologie générale ou histoire naturelle des

reptiles. Tome 9. Paris, Roret: i-xx + 1-440, 1 pl.

Durra, S. K. & MANAMENDRA-ARACHCHI, K., 1996. — The amphibian fauna of Sri Lanka. Colombo,

Wildlife Heritage Trust of Sri Lanka: 1-232.

FELLER, À. E. & HEDGES, S. B., 1998. - Molecular evidence for the early history of livingamphibians. Mol.

Phyl. Evol., 9 (3): 509-516.

FRosr, D. R. (ed.), 1985. - Amphibian species of the world. Lawrence, Allen Press & Association of

Systematic Collections: [i-iv] + i-v + 1-732.

Garcia Paris, M., Goo, D. A., PARRA-OLEA, G. & Wake, D. B., 2000. — Biodiversity of Costa Rican

salamanders: implications of high levels of genetic differentiation and phylogeographic structure

for species formation. Proceedings of the national Academy of Sciences of the USA, 97 (4):

1640-1647.

GLAW, F, KÔHLER, J., HOFRICHTER, R. & Dugois, A., 1998. - Systematik der Amphibien: Liste der

rezenten Familien, Gattungen und Arten. /n: R. HOFRICHTER (ed.), Amphibien, Augsburg, Natur-

buch Verlag: 252-258.

GorHaM, S. W., 1974. — Checklist of world amphibians up to January 1, 1970. Saint-John, The New

Brunswick Museum: 1-173.

GÜNTHER, A., 1864. — The reptiles of British India. London, Ray Society: i-xxvii + 1-452, pl. 1-26.

HAMMOND, P. M., AGUIRRE-HUDSON, B., DADD, M., GROOMBRIDGE, B., HODGI 4

ME A, M. H. & STEWART GRANT, W. 1995. - The current magnitude of biodiversity. Zn: V. H.

Heywoop & R. T. WaTsON (ed.), Global biodiversity assessment, Cambridge, Cambridge Univer-

sity Press: 113-138.

HANKEN, J. & Wake, D. B.

1982. - Genetic differentiation among plethodontid salamanders in Central

W. E. & SCHLAGER, N., (ed.), 2003. — Amphibians. In: M. HUTCHINS (ed.),

mek's animal life encyclopedia, Second Edition, Volume 6, Farmington Hills, Michigan, Gale

i-xvi + 1-507.

INGER, R. 1999. — Distribution of Amphibians in southern Asia and adjacent islands. Zn: W. E.

DUELLMAN (ed.), Patterns of distribution of Amphibians, Baltimore, John Hopkins University

Press: 445-482.

INGER, R. F., Voris, H. K. & Voris, H. H., 1974. - Genetic variation and population ecology of some

Southeast Asian frogs of the genera Bufo and Rana. Biochemical Genetics, 2 (2): 121-145.

Jonxsox, T. C., ScHoLz, C. A., TaLBoT, M. R., KELTS, K., RICKETTS, R. D., NGOBt, G.. BEUNIG, K.,

SSEMMANDA, I. & MCGiL, J. W., 1996. — Late Pleistocene desiccation of Lake Victoria and rapid

evolution of cichlid fishes. Science, 273: 1091-1093.

KIRTISINGHE, P., 1957. - The Amphibia of Ceylon. Colombo: i-xiii + 1-112, 1 pl

LaRsON, A., 1984. - Neontological inferences of evolutionary pattern and process in the salamander

family Plethodontidae. Evol. Biol., 17: 119-217.

LARSON, À. PRAGER, E. M. & WILSON, A. C., 19844. — Chromosomal evolution, speciation and

morphological changes in vertebrates: the role of social behaviour. Chromosomes Today, 8:

215-228.

LARSON, A., Wake, D. B. & YANEV, K. P., 1984. - Measuring gene flow among populations having high

levels of genetic fragmentation. Generics, 106: 293-308.

LAURENTI, J. N., 1768. — Specimen medicum, exhibens synopsin Reptilium emendatam cum experimentis

circa venena et antidota Reptilium austriacorum. Viennae, Joan. Thom. Nob, De Trattnern: i-ii +

1-215, pl. 1-5.

LEONG, T. M. & CHou, L. M., 1998. - Larval identity of the montane horned fr

Lis, S.S.. 1970. — The morphology, systemat

ridae and Hyperoliidae). Fieldiana: Zool., 57: i-vii + 1-145.

MANAMENDRA-ARACHCHI, K. & PETHIYAGODA, R., 1998. — À synopsis of the Sri Lankan Bufonidae

(Amphibia: Anura), with description of new species. J. South Asian nat. Hist., 3 (2): 213-248.

Source : MNHN, Paris

Dugois 37

MarMAYOU, J., DuBOIs, A., OHLER, A., PASQUET, E. & TILLIER, A., 2000. - Phylogenetic relationships in

the Ranidae. Independent origin of direct development in the genera Philautus and Taylorana. C.

r. Acad. Sci. Paris, Sciences de la Vie, 323: 287-297.

McDiarMib, R. W. & ALTIG, R., (ed.), 1999. - Tadpoles. The biology of anuran larvae. Chicago & London,

University of Chicago Press: i-xv + 1-444.

MEEGASKUMBURA, M., BOSSUYT, F., PETHIYAGODA, R., MANAMENDRA-ARACHCHI, K., BAHIR, M., MILIN-

KOvITCH, M. C. & SCHNEIDER, C. J., 20024. — Sri Lanka: an amphibian hot spot. Science, 298: 379.

MEEGASKUMBURA, M., PETHIYAGODA, R., MANAMENDRA-ARACHCHI, K., BOSSUYT, F. & SCHNEIDER, C.

J.,2002b. — Discovery of a remarkable radiation of direct-developing frogs in Sri Lanka. Frog Leg,

10: 12.

PENNISI, E., 2002. — 100 frogs a-leaping for biodiversity. Science, 298: 339-341.

PETERs, W. 1860. — Über einige interessante Amphibien, welche von dem durch seine zoologischen

Schriften rühmlichst bekannten ôsterreichischen Naturforscher Professor Schmarda während

seiner auf mehrere Welttheile ausgedehnten, besonders auf wirbellose Thiere gerichteten,

naturwissenschaftlichen Reise, mit deren Verôffentlichtung Hr. Schmarda gegenwärtig in Berlin

beschäftigt ist, auf der Insel Ceylon gesammelt wirden. Monatsb. Akad. Wiss. Berlin, 1860:

182-186.

PETHIYAGODA, R. & MANAMENDRA-ARACHCHI, K., 1998. - Evaluating Sri Lanka’s amphibian diversity.

Occ. Pap. Wildlife Heritage Trust, 2: 1-12.

SAGE, R. D., LoIsELLE, P. V., BASASIBWAKI, P. & WiLsON, A. C., 1984. - Molecular versus morphological

change among cichlid fishes of Lake Victoria. In: A. A. ECHELLE & L. KORNFIELD (ed.), Evolution

of fish species flocks, University of Maine at Orono Press: 185-201.

SokoL, O. M., 1977. — A subordinal classification of frogs (Amphibia: Anura). J Zool., London, 182:

505-508.

THIBAUDEAU, G. & ALTIG, R., 1999. - Endotrophic anurans. Development and evolution. /n: MCDraR-

MiD & ALTIG (1999): 170-188.

VeNcEs, M. & GLAwW, FE, 2001. - When molecules claim for taxonomic changes: new proposals on the

classifcation of Old World treefrogs. Spixiana, 24: 85-92.

Wake, D. B. & YANEv, K. P.. 1986. - Geographic variation in allozymes in a “ring species”, the

plethodontid salamander Ensatina eschsholtzii of western North America. Evolution, 40 (4):

702-715.

WiILKINSON, J., 2003. — Asian treefrogs (Rhacophoridae). Zn: HUTCHINS et al. (2003): 291-300.

WILSON, A. C., BUSH, G. L., CASE, S. M. & KING, M.-C., 1975. — Social structuring of mammalian

populations and rate of chromosomal evolution. Proc. nat. Acad. Sci. USA, 72 (12): 5061-5065.

WourMs, J. P.. 1981. - Viviparity: the maternal-fetal relationship in fishes. Am. Zool., 21: 473-515.

WRIGHT, S., 1951. - The genetical structure of populations. Annals of Eugenics, London, 15: 323-354.

The interpretation of population structure by F-statistics with special regard to systems of

mating. Evolution, 19 (3): 395-420.

WYyLES, J. S., KUNKEL, J. G. & WILSON, A.

nat. Acad. Sci. USA, 80: 4394-4397.

[., 1983. — Birds, behavior, and anatomical evolution. Proc.

Corresponding editor: Miguel VENCES.

Source : MNHN, Paris

s, 2004, 22 (1-2): 38-52.

Developmental pathway, speciation

and supraspecific taxonomy in amphibians

2. Developmental pathway,

hybridizability and generic taxonomy

Alain DUBois

Vertébrés: Reptiles & Amphibiens,

USM 0602 Taxonomie & Collections,

Département de Systématique & Evolution,

Muséum National d'Histoire Naturelle,

25 rue Cuvier, 75005 Paris, France

<dubois@mnhn.fr>

Several distinct developmental pathways exist in amphibians: free

tadpoles feeding on external resources, tadpoles or embryos feeding on

secretions from the mother or father, on their brothers or sisters, or on the

internal vitelline reserves of the egg. À new terminology is proposed for

these categories. It is suggested that generic taxonomy should take into

account these developmental pathways, i.e. that species with free feeding

tadpoles and species with other developmental modes should not be

classified in the same genus or subgenus. Artificial hybridization between

cladistically closely related species having different developmental

pathways could provide interesting information both regarding evolution-

ary phenomena and supraspecific taxonomy. Detailed proposals are offered

concerning how developmental pathways and hybridization data, combined

with cladistic information on relationships, can be used in the generic

taxonomy of amphibians. À new term is proposed for the concept of

“relational taxonomic criterion” as defined by Dugois (1988).

INTRODUCTION

In frogs, recent data on unusually high numbers of species of the direct-developing genus

Philautus in Sri Lanka and southern India, as well as of species of the direct-developing genus

Eleutherodactylus in central and southern America, led to the suggestion that such

submitted during their development to “familial”, rather than “individual”, mortality, which

could facilitate allele fixation in isolated populations and thus entail a speciation pattern

different from that of other frogs (DuBois, 2004b). A suggested way of testing this hypothesis

is through using metataxonomic data, the mean number of species per genus. Among the

problems risen by this approach, however, is the fact that no unified “genus concept” is used

by batrachologists and that genera recognized in different groups are not equivalent by any

Ogs are

Source : MNHN, Paris

Duois 39

standard. This problem of the equivalence of genera in different groups was already discussed

at length elsewhere (DuBois, 1988), but these new elements lead me to come back to it under

a new light.

The purpose of taxonomy is not to please taxonomists and phylogeneticists, but to

provide useful information to other biologists and non-biologists, including environmental

biologists, conservationists, ethologists, physiologists, etc. Among taxonomic categories, the

genus plays a particularly rôle in this respect, as the generic nomen is included in the nomen

of all species recognized by taxonomists and used for their works by other biologists (DUBOIS,

1988). If amphibian species do indeed show different patterns of speciation according to their

developmental modes, inclusion of information on the developmental pathway would appear

to be a crucial information to consider when recognizing genera. Among other things, this

inclusion would facilitate the testing of this hypothesis, which is made difficult for the time

being due to the fact that amphibian species bearing the same generic nomen may follow

different developmental pathways.

Before going further, let us briefly explore the diversity of developmental pathways in

amphibians.

CATEGORIES OF DEVELOPMENTAL PATHWAYS IN AMPHIBIANS

Developmental modes are indeed very varied in amphibians, especially in anurans. In

some cases, all the pre-imaginal development! takes place away from the adult, within the egg,

laid in a terrestrial or arboreal shelter: the embryo then depends only on the vitellus of the

eges for its resources. In other cases, the eggs are retained on the skin of the back or in a pouch

of the parent of one sex (dorsal pouch, stomach, oviduct) but does not receive any feeding

from the adult, thus depending also fully on the original vitelline reserves of the egg. Finally,

in a few other cases, the embryo receives some feeding either directly from the adult or through

eating some of the other embryos sharing its shelter within the mother's oviduct.

In the traditional usage of the terms “exotrophy” and “endotrophy” (e.g., THIBAUDEAU

& ALriG, 1999) it is not clear was is considered “outside” and “inside” (designated by the

roots exo- and endo-): if exotrophy is understood as “feeding from a resource external to the

embryo or larva”, then “endotrophy” should designate the opposite situation, i.e., “feeding

from a resource internal to the embryo or larva”, not “internal to the mother or father”.

Strictly speaking, in developmental terms the cases of feeding from resources provided by a

parent or from brothers and sisters do not belong in the category of endotrophy but are in fact

special cases of exotrophy that should better be designated under specific terms. Using a single

category of endotrophy for such a variety of cases unites artificially several non-homologous

modes of development derived independently from the tadpole model. As long as all the

observed situations are not placed in a phylogenetic perspective, comparisons and reviews of

these phenomena based on similarities and analogies (e.g.: LAMOTTE & LESCURE, 1977; WAKE,

1993; THiBAUDEAU & ALTIG, 1999) but not on homologies will be of limited evolutionary

interest. A better understanding of the evolution of these phenomena will require the

1. Development between hatching and metamorphosis (in species with feeding larvae or embryos), or before

hatching (in species in which the embryos relies only upon the egg's vitellus reserves for its development) (see tab.

D), which results in an imago, miniature copy of the adult but sexually immature (see DUROIS, 1978, 2004b).

Source : MNHN, Paris

40 ALYTES 22 (1-2)

obtention of robust cladistic hypotheses for the groups where these special developments

occur, and detailed genetic, biochemical, physiological, ethological and ecological studies of

the species concerned, as generalisation of the observations made on a few species may be

misleading. Another important aspect of such approaches is to have a clear and unambiguous

terminology to designate the various situations encountered in these groups.

Two aspects in particular must be distinguished in this respect: the place of the develop-

ment of the embryo or larva (in the external environment, or within or upon one of the

parents), and the origin of the nutritional resources used by this embryo or larva to reach the

stage imago (in the external environment, or provided by one of the parents or by brothers and

sisters). The place of development is interesting from an eco-ethological and evolutionary

point of view, but by itself it does not provide relevant categories for the comparison of

developmental pathways. For such comparisons, origin of nutritional resources is more

important as it has direct consequences on the ontogenetic trajectory. Free larvae or embryos

feeding on external resources, even within a pouch, differ from embryos maintained inside the

egg capsule in several respects, regarding breathing, locomotion or feeding: thus they require

precocious development of a functional digestive tract, earlier than in embryos feeding

on vitelline resources, etc. Given the importance of trophic resources in developmental

pathways, for more clarity I propose to use WourMs’s (1981) terminology and to expand it, as

follows.

First of all, I propose to abandon the unclear terms “exotrophy” and “endotrophy” and

to replace them, respectively, by ergotrophy (from the Greek ergon, “work”) for species with

free larvae that have to find their food in the external environement, and argiotrophy (from

the Greek argia, “idleness, inaction”) for species whose embryos are provided with

food “passively” or almost so, either from their own vitellus or from the parents, brothers

or sisters (DuBois, 2004b). Within the latter category, several subcategories can be dis-

tinguished.

The term /ecithotrophy (from the Greek Jecithos, “vitellus”) is adequate to designate

pre-imaginal development using only the vitelline reserves of the egg, without external

feeding (Wourms, 1981). Within this subcategory, two infracategories may be recognized:

leipolecitotrophy (from the Greek leipo, “I abandon”), in which the eggs are “abandoned”

by the parents and develop in an external shelter; and sregoecitotrophy (from the Greek

stegos, “roof, house”), in which the eggs are either retained in the female genital tract after

internal fertilization, or kept either upon or within one of the parents, after external fertil-

ization.

The term matrotrophy (from the Greek mater, “mother”) describes development using a

secretion from the mother as nutritional resource (WouRMs, 1981). In frogs this is observed in

the two known species of the bufonid genus Ninbaphrynoides (see e.g.: LAMOTTE & LESCURE,

1977; WAKE, 1993; THIBAUDEAU & ALTIG, 1999). A parallel situation, not considered by

WourMs (1981) as it apparently does not exist in fishes, is patrotrophy (from the Greek pater,

“father”) for nutrition by a secretion from the father. In frogs, this seems to occur in

Rhinoderma darwinii, in which the embryos develop in the male vocal sac and receive feeding

from the father, according to GoICOECHEA et al. (1986). Matrotrophy and patrotrophy are

infracategories of argiotrophy that can be grouped in a more general subcategory of goneitro-

phy (from the Greek goneis, parents”), i.e. nutrition from a secretion by the parents.

Source : MNHN, Paris

DuBois 41

In order to have a set of similarly formed terms, I propose to rename adelphotrophy

(from the Greek adelphos, “brother”) the subcategory recognized by WourMs (1981)

and many others as adelphophagy, for feeding on brothers and sisters inside the mother's

oviduct. According to whether the brothers and sisters are eaten as eggs or as embryos,

‘WourMs (1981) distinguished oophagy from adelphophagy, which does not seem an impor-

tant distinction as in both cases the origin of this nutritional resource is an egg inside the

mother’s oviduct. In contrast, he considered oophagy and adelphophagy as a subdivision of

matrotrophy, which does not recognize the fact that in matrotrophy a specific secretion is

produced by the mother to feed its embryos. It is exact that eggs and embryos eaten in

adelphotrophy were also produced by the mother, but the vitellus of the egg also, so that if

adelphotrophy was to be considered a subdivision of matrotrophy, this should also be the case

for lecithotrophy.

Among all these developmental categories, as far as feeding of the embryo is concerned,

goneitrophy and adelphotrophy are just special cases of “exotrophy”, not of “endotrophy”.

The general ecological and developmental category argiotrophy, including lecithotrophy,

goneitrophy and adelphotrophy, groups all species that are independent from feeding in the

external environment during their development (Dugois, 2004b).

Finally, the fact that in some taxa the embryos are kept within a pouch in one of the

parents is distinct from their nutritional resources. This can be accounted for by use of a

general category of gonciphory (from the Greek phoros, “bearing, carrying”), including

matrophory and patrophory according to which parent is involved, but these are eco-

ethological categories, not categories of developmental pathways.

Table 1 summarizes the major features of each of the latter categories here defined, with

examples in amphibians.

DEVELOPMENTAL PATHWAYS AND GENERIC TAXONOMY

In frogs, it is striking to note that, among 325 anuran genera containing species whose

development has, at least superficially, been described (see table 3 in DUBoIs, 2004b), 320 (i.e.

98.5 %) are homogeneous with respect to their known main ecological and developmental

category, i.e. either ergotrophy with free tadpoles (227 genera) or argiotrophy (93 genera).

This suggests that most frog taxonomists have, perhaps in part “inconsciously”, followed the

“rule” suggested by Dupois (1987: 8-9), according to which frog genera containing two or

more different developmental pathways (such as ergotrophy with free tadpoles, lecithotrophy

in eggs in shelters, lecithotrophy in adult, adelphotrophy or goneitrophy) should be disman-

tled either as distinct genera or as subgenera of the same genus. Recent proposals going in this

direction (e.g.: DuBois, 1987; BossuyT & DuBois, 2001) have been variously accepted by the

community of frog taxonom: some considering that cladistic relationships are more

important than developmental mode as a basis for generic classification. However, it should

be stressed that there is no necessary contradiction between the two approaches. Principles of

“phylogenetic taxonomy” (e.g.. DE QUEIROZ & GAUTHIER, 1992) or “cladonomy” (DUBoIs,

1997) only require that taxa be holophyletic groups (AsHLOCK, 1971; DUBoIs, 1986), but there

is nothing, at least consensually accepted, in cladistic theory to tell us how “high” or “low” in

the cladogram should be placed the limit between species-group, subgenus, genus, tribe, etc.

Source : MNHN, Paris

Table 1. - Categories proposed for developmental pathways of amphibians, with their major synonyms (terms sometimes found in the batrachological

literature for these categories), definitions and examples in amphibians. Rank 2 subcategories are subdivisions of rank 1 categories, and rank 3

infracategories are subdivisions of rank 2 subcategories.

Rank 1 Rank 2 Rank 3 Scan Kind and place of Nutritional resources for E Le

category subeategory | infracategory ynonyr pre-imaginal development pre-imaginal development Wamp)

ser m Free aquatic or terestrial arva between hatching and | Extemal resources of he aquatic y

Betty ut metamorphots or tenestial environment Hi co per Ë

PE | Eidébony Either within a pouch in one of the parents or inside egg | No access to the external <

Angiotropl | Enoophy apsule im terestrial or arborcal habitat resources oflhe environment ES

Lecithotrophy Endotrophy,lcithotroph Inside or outside eg capsule Vitline reserves ofthe eng g

Lcipolecithotrophy | Endotrophy: direct development, , Sa sas im sv ce Arthroleptis, Eleutherodactylus, 4

Lepolecihotophy | niicolous development Reg hab ER le Rae Phleutus JS

cgolcihoophy | Endonophr: voviviar ge | pu Len in a pouchmihin or pon one of parents | Vittine rares of « Asa, Nectphmides, |

Stegolecihonoph} | rooding, exovivipariy, paravivipari| FEB KePin a pouch within or upon one of the parents | Vitelh dite Rheobatrachus &

| Goncirophy Endotronhy vivant, exoviviparit|Fre embryo orlarva thin a pouch in one ofthe parents Sections fiom a parent

| Matrotrophy Endotrophy: viviparity Free embryo or larva within oviduct of mother Secretions from the mother Nimbaphrynoides

Paotophy Endotophy exoviviparty Free embryo or larva thin a pouch Of father Secretions from te father Rhinoderma

Adelphoophy Endotrophy: adelphophagy Free embryo or lava within oviduet of mother Brothers and sisers Salamandre atra

Source : MNHIN, Paris:

DuBoIs 43

Therefore it could well be consensually decided that, as soon as two clades or subclades of

frogs display different developmental modes, they should be treated as distinct genera, or at

least subgenera (see below). This would have a strong advantage, that of delivering the

following clear message to the various categories of non-taxonomists that are users of the

nomina of frog species: “whenever two species bear the same generic (or, in some cases,

subgeneric) nomen, they have (or are believed to have) the same gross developmental mode”.

As shown above, there would be very little to change now to homogenize all frog taxonomy in

this respect, as this is already “almost” done.

The frequently used formula “developmental mode” should be clarified a little further

here. The important point here is to distinguish between different developmental pathways.

What is suggested here is to take into account, in the taxonomic recognition of supraspecific

taxa, the difference between species that follow an ontogenetic trajectory such as that

described in the development table of GosNER (1960), leading to an ergotroph free tadpole,

and those that follow an alternative developmental pathway like those reviewed e.g. by

THIBAUDEAU & ALTIG (1999) and evoked above. The important point is here, and not in the

place of development of the egg (in an external shelter, or inside a pouch in the adult) or even

in the exact developmental stage at which hatching takes place. Thus, it is not suggested here

that taxonomic recognition should be given to differences that can be considered “trivial”?

with respect to the question here addressed, such as the fact that, in some salamander species,

hatching can occur either already within the female’s genital tract or after deposition of the

egg, but with a largely unmodified developmental pathway. In these different populations, at

least according to the published data, hatching occurs in different places but there is no

evidence that it takes place at different developmental stages or that the development table is

modified. Similarly, the term “viviparity”, sometimes used (e.g., GarCIA-PaRIs et al., 2003) to

designate salamander species that give birth to terrestrial imagos, is misleading. This is just a

special case of ovoviviparity, where the embryos start their development with important

viteline reserves, the larvae later may feed by adelphotrophy and development continues very

late within the female genital tract, but without exhibiting a particular pathway. In contrast,

the term “viviparity” should be restricted to situations where, like in the mammals, the egg

does not have important vitelline reserves and the embryos develops thanks to nutrients

provided directly by the female in the genital tract: in amphibians, this situation is known only

in the bufonid genus Nimbaphrynoides.

For the time being, only five anuran genera out of 350 are considered to include both

argiotroph species and ergotroph species with free tadpoles (THIBAUDEAU & ALTIG, 1999): (1)

four American genera: Adenomera Steindachner, 1867 (Leptodactylidae); Colostethus Cope,

1866 (Dendrobatidae); Gastrotheca Fitzinger, 1843 (Hylidae); Pipa Laurenti, 1768 (Pipidae);

(2) one Malagasy genus: Mantidactylus Boulenger, 1895 (Ranidae). In all other regions of the

world, all anuran genera are homogeneous regarding their known developmental pathway.

Detailed comparisons of developmental pathways between members of both groups are

available in some of these cases only (e.g.. WASSERSUG & DUELLMAN, 1984), but in the cases

where the developmental pathways will prove to be significantly different, it is here again

suggested that this should be taxonomically recognized. Nomina are already available to

designate the genera or subgenera that would result from dismantlement of the genera

Colostethus (see DUELLMAN & TRUEB, 1985), Gastrotheca (see DUBoIs, 1987), Mantidactylus

(see GLAW & VENCES, 1994) and Pipa (see GORHAM, 1966).

Source : MNHN, Paris

44 ALYTES 22 (1-2)

Besides, two anuran genera are known to include two different kinds of lecithotroph

development (THIBAUDEAU & ALTIG, 1999), i.e. both stegolecithotroph and leipolecithotroph.

In one case (genus Æleutherodactylus Duméril & Bibron, 1841; Leptodactylidae) the eggs may

develop either within the mother (Eleutherodactylus jasperi) or in an external shelter (all other

known species). In the second case (genus Leiopelma Fitzinger, 1861; Leiopelmatidae),

lecithotroph development may occur within egg (Leiopelma hochstetteri) or in a dorsal pouch

of the father (Leiopelma archeyi and Leiopelma hamiltoni). Detailed study of the development

of these species are needed to establish whether their developmental pathways are similar,

despite the difference of location of the developing egg, or significantly different. In the latter

case, it would also be better to recognize subgenera in these taxa, and here also nomina would

be available both for Eleutherodactylus (see HEDGES, 1989) and Leiopelma (see WELLS &

WELLINGTON, 1985).

DEVELOPMENTAL PATHWAYS AND HYBRIDIZATION

Criteria for recognition of taxa can be sorted into criteria for their délimitation and

criteria for their rank assignation in a hierarchical taxonomic system. As well clarified by

Simpson (1951, 1961), criteria for delimitation of taxa include criteria for inclusion and for

exclusion, and all criteria can be arbitrary or nonarbitrary. The topology of a cladogram, taken

as an accepted hypothesis of relationships between species, can be used as a nonarbitrary

criterion for delimitation of taxa, but it provides by itself no criterion for ranking: the

cladonomic requirement of holophyly of taxa allows to recognize them but not to allocate

them to any category in a hierarchical system. À possible “simplistic” attitude in this respect

is to propose the suppression of taxonomic ranks, but the hierarchical structure of taxonomy

is critical in allowing the latter to play its rôle of a “convenient information storage and

retrieval system” about taxa, their characters, distribution, evolution, relationships, etc.

(MAYR, 1981: 511). It should therefore not be suppressed, but made more useful and more

general in using nonarbitrary criteria for ranking that allow at least a certain equivalence

between taxa of same rank in different groups (see e.g.: DuBois, 1988: 66-73, and references

therein; AVISE & JOHNS, 1999).

Among other criteria, several authors (VAN GELDER, 1977; DUBoIs, 1981, 1988; PLA-

TEAUX, 1981) supported the use of hybridizability as a nonarbitrary criterion for inclusion of

different species in the same genus. Interestingly, beside being a criterion for taxa delimitation,

thisis also a criterion for ranking. On the other hand, Dugois (1988) insisted that this criterion

should never be used for exclusion. In other words, according to this criterion, the fact that two

species are able to give birth to viable true diploid adult hybrids is to be used as evidence that

these two species belong in the same genus, whereas the absence of hybridizability provides by

itself no useful information for the generic allocation of two species. It is important to stress

here that hybridizability of species, as strictly defined by DuBois (1988), is a taxonomic

criterion but not a phylogenetic criterion, as there is no direct correspondence between

hybridizability and cladistic relationships: hybridizable species are not necessarily cladisti-

cally sister-species, but may be quite distantly related (see e.g. the case of European green frogs

of the subgenus Pelophyla. s for this are easy to understand, as

this is linked to the necessity for closely related species to develop isolation mechanisms in

Source : MNHN, Paris

Dugois 45

sympatry ot parapatry, but not in allopatry (see Dugois, 1988). Hybridizability (or its

absence) between two species is not a “character” of any of these species, and is therefore

neither plesiomorph nor apomorph: if it were so, each species would have millions of

characters, according to its potential hybidizability with all other living species. It is rather a

“relational taxonomic criterion”’ (DUBoIs, 1988), or more shortly a relacter (from the Latin

relatio, in the sense of “relation between two things”, and character, in the sense of “character,

mark that distinguishes something”). Relacters are of various kinds, as discussed in detail by

Dusois (1988): e.g., sympatry-parapatry-allopatry, parasitic specificity, ecological competi-

tive exclusion, presence-absence of a hybrid zone and of a gene flow between two parapatric

entities, etc. Using such a relacter as hybridizability to build up taxonomies is a way to

acknowledge that taxonomy does not rely only on characters and relationships, but on other

kinds of information: similarly, the absence of gene flow in the field between two parapatric

entities is a way to establish the specific status of these two entities, although the two kinds of

information on which this decision is taken (parapatric geographic distribution and absence

of gene flow) do not pertain to any of the two entities taken by itself, but characterizes their

relation.

Just like the criterion of similar developmental pathway discussed above, the principle

of hybridizability as a nonarbitrary criterion for inclusion in a genus can perfectly be

used within the frame of a system of phylogenetic taxonomy: one just has to place the

“bar” of the genus rank just at the level of hybridizable species pairs, and use consistently

the principles of cladonomy for all other taxa. Advantages of this system upon any

other arbitrary or “consensual” delimitation of genera were discussed at length elsewhere

(Dusois, 1988). The new question that may be asked here is: what can be the relationships

between this criterion of hybridizability and the criterion of similar developmental

mode?

Although a number of artificial hybridizations have been carried out in the past in

amphibians (reviews in: MONTALENTI, 1938; MOoRE, 1955; BLaAIR, 1972), none of these

reported experiments involved argiotroph, particularly lecithotroph, anuran species, either

between themselves or with species of the same groups having free tadpoles. A rapid a priori

thinking might suggest that there is no need to try such crossings, because of course the

“developmental program” of a species with tadpole is unlikely to be compatible with that of

a lecithotroph species, and such a combination appears bound to fail at a rather early stage of

development. However, until the experience is carried out in different anuran groups including

both kinds of species, this possibility cannot be theoretically ruled out. In amphibians,

hybridization can at least partially succeed between species with rather different develop-

ments (e.g.. MARTINEZ Rica et al., 1984), and in fishes it can be successful, at least up to a

certain point, between species that are considered only distantly related (e.g., WHITT et al.,

1973).

Particularly interesting in this respect are the works on the frog genus Gastrotheca by

several authors (DEL PINO, 1980; SCANLAN et al., 1980; DEL PINO & ESCOBAR, 1981: WASSER-

suG & DUELLMAN, 1984) which suggest that in this genus lecithotroph development was

plesiomorphic, but that, in several distinct groups of high altitude populations, a reversal to

a development through a free tadpole stage occurred. Under such a scenario, rather than a

replacement of a developmental program by another, what would have occurred is the

Source : MNHN, Paris

46 ALYTES 22 (1-2)

appearance, possibly through phenomena of duplication of regulatory genes (GouLD, 1977;

RaArr & KAUFMAN, 1983), of a new developmental program beside the initial one, which would

be conserved in the genome, but unused, “in dormancy”, in some species. The possibility of a

“switch” from one program to another, on the occasion of speciation events, would allow

these frogs to adapt to new ecological conditions or to conquer new habitats. Such a scenario

may have developed in several groups of frogs including both ergotroph with tadpoles and

lecithotroph species, and indeed the possibility of its occurrence in the genus Philautus is

suggested by the topology of the cladogram published by MEEGASKUMBURA et al. (2002a): if

this cladogram was confirmed (but see Dugois, 20044), lecithotrophy would have appeared

independently twice, in two groups of species (the Indonesian-Indochinese, and the Indian-

Sri Lankan, ones) nested within a clade of ergotroph rhacophorids.

If two different developmental programs can indeed be conserved in parallel in the

genome of some species, then this would open the possibility of successful hybridization

between species having different developmental pathways: in the early hybrid embryo, the

regulatory genes of one of both species might “take over” those of the other one, and

“impose” the use of one developmental pathway. At this stage, this suggestion is purely

theoretical, but experimental testing of this possibility, between closely related species having

different developmental modes, might be very rewarding. Given the difficulty to carry out

such hybridization experiments in all rigour (with control crosses, caryological and electro-

phoretic assesssment of the real hybrid, and not gynogenetic, nature of the embryos, etc.; see

Dusois, 1988), such experiments would certainly have more chances to be successful if carried

out with fresh animals just collected in the field, i.e. close to their natural populations in their

native countries.

Should hybridization prove successful, in some cases, between ergotroph and argiotroph

species, this would require, in order to follow both the hybridizability criterion (DuBois, 1988)

and the criterion of similar developmental mode (Dugois, 1987), to place these species in

different subgenera of the same genus. If reversion from one developmental mode to another

occurred independently in several different groups, these groups should be treated as different

subgenera of the same genus, as suggested by DuBois (1987) in the genus Gastrotheca. On the

other hand, in other cases, e.g. in groups where lecithotroph species are not known to have

closely related species, it may be unlikely to ever find ergotroph species that would have

retained the ability to hybridize successfully with them. In such cases, if there is no conflict

with the other criteria suggested (DuBois, 1988: 76-77, 105-108), the two groups should be

recognized as distinct genera, not subgenera.

DETAILED PROPOSALS REGARDING GENERIC TAXONOMY

In zoology, the establishment of supraspecific taxa and of their taxonomic ranks, under

the guidelines suggested above, can rely upon several nonarbitrary criteria. In frogs, among

other criteria, three powerful ones are holophyly of taxa (delimitation criterion), common

development pathways of species (delimitation criterion) and hybridizability between species

(both delimitation and ranking criterion). To make clearer the hierarchical relationships

between these criteria, the hypothetical cladograms presented in fig. 1 can be commented in

Source : MNHN, Paris

Duois 47

some details. AI these cladograms, involving six species, have the same topology, but include

different kinds of information regarding developmental pathways and hybridization. As will

be shown in detail below, in some cases the use of the criteria presented above lead to clear and

unique proposals regarding taxa that should be recognized as genera or subgenera, whereas in

other cases these criteria alone are not enough to decide among several possibilities: in these

latter cases, other criteria must be used to go further, as discussed e.g. by MAYR (1969) or

Dusois (1988), but these further steps won't be considered here.

In the three cladograms of figures 1a-c, no information is available regarding hybridiza-

tion, but some species are known to develop through a free aquatic tadpole stage, whereas

others have leipolecithotroph development, e.g. direct development in eggs deposited in

terrestrial shelters. According to the proposals above, genus rank should be afforded to

groups of species that share a developmental pathway. In order to respect the principles of

cladonomy, i.e. to recognize only holophyletic groups as taxa, this results in a different number

of genera according to the distribution of developmental pathways among the six species.

Note that in the situation of figure la, the use of this criterion alone does not allow to decide

whether a single genus, or a genus with two subgenera, or two distinct genera, should be

recognized among the four species with tadpoles, but in the two other cases no such

uncertainty exists.

In the three cladograms of figures 1d-f, no information is available regarding develop-

mental pathways, but data are available about some pairs of species that are known to be liable

to give birth to viable true diploid adult hybrids. Here also, in some cases the information

provided by hybridizability does not allow to choose between several generic taxonomies, as

hybridizability is only a criterion for inclusion (i.e. for grouping species in a single genus) but

should never be used for exclusion (i.e. for splitting genera). However, in some cases, like that

shown in figure 1f, information on hybridizability of two quite distantly related species may be

enough to stabilize the generic taxonomy of a whole group.

Now, let us consider the consequences of combining information on developmental

pathways and information on hybridizability in a single cladogram. Crossing the three

situations of figures la-c with the three situations of figures Id-f gives nine different situa-

tions, presented in figures 1g-0. Taxonomic decisions in these nine situations must follow a

hierarchy between criteria, as proposed in detail by DuBois (1988: 82-84): according to this

hierarchy, data on hybridizability must be used first, to establish which species cannot be

placed in different genera. This means that, in the hypothetical case (not yet known to be

indeed possible in some groups of amphibians) where species showing different developmen-

tal pathways would be able to give viable true diploid adult hybrids, they should be placed in

the same genus: but then they should be referred to different subgenera. Such hypothetical

situations are shown in figures li, 1k, 11, 1m and In. After the criterion of hybridizability,

developmental data should be used to split further some genera into subgenera (in the

exceptional case just mentioned), or, more frequently, to decide between alternative generic

taxonomies among which the hybridization criterion alone does not allow to choose. Thus, in

the situation of figure Id, hybridization data do not allow to choose between recognizing one,

two or three genera. In figure 1g, developmental data allow to recognize a distinct genus for

the species Sgl and Sg2, but still do not allow to decide between one or two genera for the

species Sg3 to Sg6; this decision will have to rely on other pieces of information. In contrast,

Source : MNHN, Paris

Fig. 1. — Cladograms showing hypothetical relationships between si

species and providing information on their developmental pathways and a

hybridization between them. Abbreviations of taxa include à capital

letter for rank of taxon (S, species: SG, subgenus; G, genus), a different

lower-case letter for each subfigure (a, b, c), a number for each species or ah cb ca

genus, sometimes followed by à letter for subgenera within a genus: thus c dr 1 à

1 indicates species with free aquatic ergotroph tadpoles and the symbol 2 Sbi

indicates species with leipolecithotroph development (direct develop-

ment in eggs deposited in terrestrial shelters). Species liable to give birth

10 viable true diploid adult hybrids are connected by the symbol 3.

Generic and subgeneric taxa recognized on the basis of the information

provided are shown at the top of figures as square brackets that can be

continuous line (in the case of nonambiguous taxonomies; symbol 4 for

genus, symbol 6 for subgenus) or composed of hyphens (in the case of b

several possible alternative taxonom mbol 5 for genus). (a-c) Only

information on developmental pathway is available, but none on hybrid-

ization. (d-1) Only information on hybridization is available, but none on Gel Ge2 Ge Ged Gcs

developmental pathways. (g-0) AI possible combinations of cases (a-c)

and (d-1). with both kinds of information available. E) = = =

Sb2 Sb3 Sb4 SbS Sb6

(T-D) TT SALAIV

Symbols

G sG

æ ® = — En

1 2 3 4 5 6

Source : MNHN, Paris

Dugois

OST ns ES us

ve CT

CO]

est us ë Eus Zu Ius

98SdmpSS pis (is 2is 15S

æ æ €

Lo) 189

se SIS HIS EISm US us

: MNHN, Paris

Source

ALYTES 22 (1-2)

os 91S SIS @em +IS £IS IS és 11S

vzInS |

| RL

u j i n

ETES vue sr n

C j pv

uc u =£ Su +

[

QUS ÇUS pus EUS@mpTUS US

è n D © © je)

quus vus à p ) A

(LD) ao ]

MNHN, Paris

Source

Dusois 51

in figures 1h and li, the combined use of the two criteria here proposed allows to decide

without ambiguity which groups should be recognized as genera, and which as subgenera.

As discussed already in DuBois (1988), supraspecific taxa defined under such guidelines

are likely to be more informative than taxa just recognized by simple “consensus” but without

any clear theoretical background. After a brief period of change, the new taxonomy may

prove more useful both for taxonomists and non-taxonomists and for various kinds of studies

and comparisons. As information on hybridizability and developmental pathways, once

obtained, is not liable to change (in contrast with the topology of cladograms), a generic

taxonomy using these criteria would be more stable in the long run than a generic taxonomy

based on cladistic hypotheses alone, but ignoring these biological criteria.

ACKNOWLEDGMENTS

For comments on previous drafts of this paper, I am grateful to Thierry Lodé, Annemarie Ohler,

Miguel Vences, David B. Wake and an anonymous reviewer, Roger Bour helped for the preparation of the

table, and Annemarie Ohler for that of the figure.

LITERATURE CITED

ASHLOCK, P. D., 1971. - Monophyly and associated terms. Syst. Zool., 20: 63-69.

AVISE, J. C. & JOoHNS, G. C., 1999. — Proposal for a standardized temporal scheme of biological

classification for extant species. Proc. nat. Acad. Sci. USA, 96: 7358-7363.

BLAIR, W. F. (ed.), 1972. - Evolution in the genus Bufo. Austin & London, Univ. Texas Press: i-viii + 1-459,

pl. 1-6.

BossuyT, F. & Duois, A., 2001. — A review of the frog genus Philautus Gistel, 1848 (Amphibia, Anura,

Ranidae, Rhacophorinae). Zeylanica, 6 (1): 1-112.

De Quemoz, K. & GAUTHIER, J., 1992. - Phylogenetic taxonomy. Am. Rer. Ecol. Syst., 23: 449-480

De Po, E. M. 1980. - Morphology of the pouch and incubatory integument in marsupial frogs

(Hylidae). Copeia, 1980: 10-17.

De. PINo, E. M. & EscomaR, B., 1981. - Embryonic stages of Gastrotheca riobambae (Fowler) during

maternal incubation and comparison of development with that of other egg-brooding hylid frogs.

J Morph., 167: 277-295.

Dunois, A., 1978. - Les principaux stades de développement significatifs en écologie et en génétique des

populations des Amphibiens Anoures. La Terre et la Vie, 32: 453-459.

ais 1981. — Hybridation interspécifique et notion du genre en zoologie. C. r Acad. Sci. , (3), 292 (A):

201-203.

CE 1986. - À propos de l'emploi controversé du terme “monophylétique”: nouvelles propositions. Bull.

Soc. linn. Lyon, 55: 248-254.

1987. - Miscellanea taxinomica batrachologica (1). Alvtes, 5 (1-2): 7-95.

1988. - The genus in zoology: a contribution to the theory of evolutionary systematics. Mém. Mus:

nain. Hist. nat, (A), 140: 1-123.

1997. — An evolutionary biologist's view on the science of biology. A/rres, 15 (3): 13

2004a. - Book review. Amphibians of Nepal: a few words of caution. A/ytes, 21 (

2004. - Developmental pathway, speciation and supraspecific taxonomy in amphi

there so many frog species in Sri Lanka? A/vtes, 22 (1-2): 19-37.

136.

73-180.

ans. 1. Why are

Source : MNHN, Paris

52 ALYTES 22 (1-2)

DUuELLMAN, W. E. & TRUEB, L., 1985. — Biology of amphibians. New York, McGraw-Hill, “1986”:

1-670.

Garcia PARIS, M., ALCOBENDAS, M., BUCKLEY, D. & WaKE, D. B., 2003. - Dispersal of viviparity across

contact zones in Iberian populations of fire salamanders (Salamandra) inferred from discordance

of genetic and morphological traits. Evolution, 57 (1): 129-143.

GLAW, FE & VENCES, M. 1994. — 4 fieldguide to the amphibians and reptiles of Madagascar. Secondedition,

including mammals and freshwater fish. Kôln, Vences & Glaw: 1-480, 48 pl.

GolCOECHEA, O., GARRIDO, O. & JORQUERA, B., 1986. — Evidence for a trophic paternal-larval rela-

tionship in the frog Rhinoderma darwinii. J. Herp., 20 (2): 168-178.

GorHaM, S. W., 1966. — Liste der rezenten Amphibien und Reptilien. Ascaphidae, Leiopelmatidea (sic),

Pipidae, Discoglossidae, Pelobatidae, Leptodactylidae, Rhinophrynidae. Das Tierreich, 85: i-xvi +

1-222.

Gosner, K. L., 1960. - A simplified table for staging anuran embryos and larvae with notes on

identification. Herpetologica, 16: 183-190.

Goup, S. J., 1977. - Ontogeny and phylogeny. Cambridge, Mass., Belknap Press: i-xvi + 1-501.

GÜNTHER, R., 1990. — Die Wasserfrôsche Europas (Anura — Froschlurche). Wittenberg Lutherstadt, A.

Ziemsen Verlag: 1-288, pl. 1-2.

Heces, S. B., 1989. - Evolution and biogeography of West Indian frogs of the genus Fleutherodactylus:

slow-evolving loci and the major groups. /n: W. Woops (ed.), Biogeography of the West Indies: past,

present, and future, Gainesville, Sandhill Crane Press: 305-369.

LamorTE, M. & LESCURE, J., 1977. - Tendances adaptatives à l'affranchissement du milieu aquatique chez

les Amphibiens Anoures. La Terre et la Vie, 31 (2): 225-311.

MARTINEZ RICA, PARDO ARA, M. P. & CERVANTES VALLEJOS, 1984. — Hibridaciôn en A/vtes muletensis.

In: H. HEMMER & J. P. ALCOVER (ed.), Histôria biolôgica del ferreret, Mallorca, Moll: 231-233.

Mar, E., 1969. - Principles of systematic zoology. New York, MeGraw-Hill: i-xiii + 1-428.

ee 1981. — Biological classification: toward a synthesis of opposing methodologies. Science, 214:

510-516.

MEEGASKUMBURA, M., BOSSUYT, F., PETHIYAGODA, R., MANAMENDRA-ARACHCHI, K.., BAHIR, M., MILIN-

KOvITCH, M. C. & SCHNEIDER, C. J., 2002. — Sri Lanka: an amphibian hot spot. Science, 298: 379.

MONTALENTI, G., 1938. — L'ibridazione interspecifica degli anfibi anuri. Archivio zoologico, 25, suppl. 4:

157-213.

Moore, J. A., 1955. - Abnormal combinations of nuclear and cytoplasmic systems in frogs and toads.

Adv. Gene! 139-182.

PLATEAUX, L., 1981. - Critère mixiologique et notion de genre. Bull. Soc. zool. France, 106: 513-520.

RAFF, R. A. & KAUFMAN, T. C., 1983. Embryos, genes, and evolution. The developmental-genetic basis of

evolutionary change. New York & London, Macmillan: i-xiii + 1-395.

SCANLAN, B. MaxsoN, L. R. & DUELLMAN, W. E., 1980. - Albumin evolution in marsupial frogs

(Hylidae: Gastrotheca). Evolution, 34: 222-229.

Simpson, G. G., 1951. The species concept. Evolution, 5: 285-298.

me 1961. — Principles of animal taxonomy. New York, Columbia Univ. Press: i-xii + 1-247.

THIBAUDEAU, G. & ALTIG, R., 1999. — Endotrophic anurans. Development and evolution. In: R. W.

McDraRMID & R. ALTIG, (ed.), Tadpoles - The biology of anuran larvae, Chicago & London,

University of Chicago Press: 170-188.

VAN GELDER, R. G., 1977. - Mammalian hybrids and generic limits. American Museum Novitates, 2635:

xix +

H., 1993. - Evolution of oviducal gestation in amphibians. J exp. Zool., 266 (5): 394-413.

R. J. & DUrLLMAN, W. E., 1984. - Oral structures and their development in egg-brooding

hylid frog embryos and larvae: evolutionary and ecological implications. Z Morph., 182: 1-37.

. & GTON, C. R., 1985. - À classification of the Amphibia and Reptilia of New

ealand, Australian Journal of Herpetology, Supplementary Series 1: 62-64.

Waurr, G. S. CiLpers, W. F. & Cno, P. L., 1973. - Allelic expression at enzyme loci in an intertribal

rid sunfish. Journal of Heredity, 64: 54-61.

.. 1981. — Viviparity: the maternal-fetal relationship in fishes. Am. Zool., 21: 473-5

Corresponding editor: Miguel VENCES.

Source : MNHN, Paris

Alptes, 2004, 22 (1-2): 53-64.

Molecular phylogenetic relationships

of Lankanectes corrugatus from Sri Lanka:

endemism of South Asian frogs

and the concept of monophyly

in phylogenetic studies

Magali DELORME*, Alain DUBOIS*, Joachim KosucH** & Miguel VENCES***

* Vertébrés: Reptiles & Amphibiens, USM 0602 Taxonomie & Collection

Département de Systématique & Evolution, Muséum National d'Histoire Naturelle,

25 rue Cuvier, 75005 Paris, France

<delorme@mnhn.fr>, <dubois@mnhn.fr>

#* Institut für Zoologie, Abteilung Okologie, Universität Mainz, SaarstraSSk

**# Zoological Museum, University of Amsterdam, Mauritsl

1092 AD Amsterdam, The Netherlands

99 Mainz, Germany

ade 61,

For more than fifteen years, the frog genus Limnonectes (Ranidae,

Dicroglossinae, Limnonectini) was considered to contain more than 40

South-East Asian species, and a single species from Sri Lanka, L. corruga-

tus. We analysed 1198 base pairs of the mitochondrial 12S and 16S rRNA

genes in L. corrugatus, in representatives of most major subgroups of

Limnonectes, and in several genera thought to be related to this genus. The

data allow to significantly exclude a relationship of the Sri Lankan species to

South-East Asian Limnonectes; instead, it seems clustered with species of

Rana and Nyctibatrachus, which supports the previous recognition of the

monotypic genus Lankanectes for L. corrugatus. The morphological

specializations of this species confirm that it may be the only known

representative of an additional major ranid lineage (Lankanectinae)

endemic to South Asia, an area of high importance as center of basal

diversity and endemism of this family. Our data also suggest some

comments on the generic taxonomy in the Limnonectini tribe of the Dicro-

glossinae. By contradicting previous statements on the monophyly of

Limnonectes, they also point to a general terminological problem in

phylogenetic studies. We propose to use the term homophyletic to refer to

groups in which the available data do not contradict holophyly but in which

taxon sampling is still incomplete or uncertain.

INTRODUCTION

The amphibian fauna of South Asia, that is, India and Sri Lanka, contains an important

number of endemic taxa at deep phylogenetic levels. This distinctness of South Asian frogs

was already highlighted by BLOMMERS-SCHLÔSSER (1993) who erected the new subfamilies

Source : MNHN, Paris

54 ALYTES 22 (1-2)

Indiraninae (now a synonym of Ranixalinae) and Nyctibatrachinae for the endemic Indian

genera Nyctibatrachus and Indirana. The spectacular discovery of the endemic Nasikabatra-

chidae further confirmed the biogeographic importance of this region (Buu & BossuYT,

2003). The Indian plate had been drifting northwards since its separation from Madagascar

88 million years ago (BARRON et al., 1981; Srorey, 1995; SrorEy et al., 1995), and several

lineages of frogs may have dispersed out of India into other regions of Asia (DUELLMAN &

TRUE, 1985; Bossuyr & MiLINKOvVITCH, 2001). However, surprisingly few phylogenetic

studies have focused on South Asian ranids in the past (e.g., BLOMMERS-SCHLÔSSER, 1993), and

only in recent times have some data become available (BossuyT & MILINKOVITCH, 2000, 2001;

VENCES et al., 2000c; KosuCH et al., 2001). ROELANTS et al. (2004) emphasized the deep

evolutionary history of several South Asian lineages in the family Ranidae, many of which

might be crucial to determine relationships in this family and, in a wider context, in the

superfamily Ranoidea. Among the endemic South Asian genera or subgenera which may

yield new insights into ranoid biogeography are the Indian microhylid Melanobatrachus, the

Indian ranids Clinotarsus, Indirana, Micrixalus, Minervarya, Nyctibatrachus and Sphaero-

theca, and the Sri Lankan ranid Nannophrys (Dugois, 1992, 2003; Dugois al., 2001).

Considering the high degree of homoplasic morphological adaptations in frogs, mole-

cular methods have proved to be a useful tool to uncover phylogenetic relationships undis-

turbed from possible convergent similarities (e.g., Hay et al., 1995; VENCES et al., 2000a). Of

the South Asian endemics, so far no published data are available on Clinotarsus, Melanoba-

trachus and Minervarya; the position of /ndirana, Micrixalus and Nyctibatrachus is basically

unsolved (BossuyT & MiziNKovITCH, 2000, 2001; VENCES et al., 2000c; ROELANTS et al.,

2004): and Nannophrys and Sphaerotheca proved to be related to the widely distributed genera

Euphlyctis, Fejervarya and Hoplobatrachus (Bossuyr & MiziNKovirCH, 2000; VENCES et al.,

20004,c; KosucH et al., 2001). However, as ranoïid taxonomy is still largely unsolved, the

generic attribution of South Asian species is not in all cases certain.

Another Sri Lankan species of unclarified phylogenetic relationships was described by

PETERS (1863) as Rana corrugata. This species was included by BOULENGER (1920) in his

section Ranae kuhlianae of the genus Rana, so that DuBois (1981), when he erected Limno-

nectes as a subgenus of Rana, and later (1987, 1992) as a distinct genus, included it in this

group. Since then, the species has been named Limnonectes corrugatus in several works (e.g.,

Durra & MANAMENDRA ARACHCHI, 1996; DurTA, 1997). However, Dusois & OHLER (2001)

pointed to morphological characters that exclude this species from Limnonectes, and erected

for it the monotypic genus Lankanectes.

The genus Limnonectes as currently understood (e.g., OHLER & DuBois, 1999; Dugois &

OHLER, 2000, 2001; EMERSON et al., 2000; DuBois, 2003; Evans et al., 2003) contains a number

of South-East Asian species. Some of these have fangs in the front of their mandibles, so that

these species have been named “fanged frogs”. They served as a model group to understand

the evolution of several traits such as reduction of vocal sacs (EMERSON & Voris, 1992:

EMERSON & BERRIGAN, 1993; EMERSON & WaRD, 1998) and to test biogeographical hypoth-

eses at the interface of the Oriental and Australian zones (EVANS et al., 2003). Limnonectes has

been claimed to constitute a monophyletic group (EMERSON et al., 2000; Evans et al., 2003),

but molecular studies failed to place L. corrugatus in a clade with the South-East Asian L.

kuhlii, type-species of Limnonectes (BossuYT & MiLiNKOVITCH, 2000; VENCES et al., 2000c).

Source : MNHN, Paris

DELORME & AL. 55

Table 1. — Species of Limnonectes and putatively related genera included in this study, their distribution

and their allocation to groups or clades proposed by previous authors. (1) Taxonomic allocation of

“fanged frogs” according to DUBOIS (1992), OHLER & DuBois (1999) and DUBOIS & OHLER

(2000): Æ, subgenus Limnonectes (Elachyglossa); Le, grunniens group of the subgenus

Limnonectes (Limnonectes); Lk, kuhlit group of the subgenus Limnonectes (Limnonectes): Lm,

microdiscus group of the subgenus Limmonectes (Limnonectes); T, genus Taylorana. (2)

Allocation of “fanged frogs” to subclades 1a, 1b, 2, 3 or 4 of the genus Limnonectes according to

EMERSON et al. (2000) and EVANS et al. (2003).

Species Taxonomie allocation (1) |. Cladistic allocation (2) Distribution

Fejervarya cancrivora E e China, Indochina, Indonesia, Malaysia

Fejervarya limnocharis - . |Indochina, Indonesia, Malaysia

Hoplobatrachus chinensis - - China, Indochina, Indonesia, Malaysia

Limnonectes blythit Lg 4 Indochina, Indonesia, Malaysia

Limnonectes gyldenstolpei E a Indochina

Limnonectes kuhlii Lk 1 Indochina, Indonesia, Malaysia

Limnonectes macrocephalus Le 3 Philippines

Limnonectes paramacrodon Lg 4 Indonesia, Malaysia

Limnonectes woodworthi Lm 3 Philippines

Taylorana hascheana T la |Indochina, Indonesia

Lankanectes corrugatus - - [Sri Lanka

Limnonectes is rather species-rich with currently about 50 recognized species but probably

many more indeed (Evans et al., 2003), and several subclades have been identified in this clade

(EMERSON et al., 2000; Evans et al., 2003). However, as these studies did not include L.

corrugatus, the relationships between this Sri Lankan species and the South-East Asian

Limnonectes remained unclarified. Recently, ROELANTS et al. (2004) included Lankanectes

corrugatus and two species of Limnonectes in a molecular phylogenetic analysis, which

supported the exclusion of the former species from Limnonectes.

The aim of this paper is to test more comprehensively if the Sri Lankan species is

phylogenetically related to Limnonectes of South-East Asia or if it may be a representative of

an endemic South Asian lineage, using a larger taxonomic sampling than in ROELANTS et al.

(2004). For this purpose we analyzed mitochondrial DNA sequences of this species and of

representatives of several groups (tab. 1) of Limnonectes sensu DUBoIs & OHLER (2000) and of

three genera, which previously had been included in that genus (Fejervarya, Hoplobatrachus

and Taylorana).

MATERIALS AND METHODS

Tissue samples (muscle or liver: either fresh or preserved in 98 % ethanol) were available

from 25 ranoid species. DNA was extracted using QIAmp tissue extraction kits (Qiagen). We

amplified two fragments of 12S rRNA gene (417 pb and 470 pb). The original couple of

primers are based on the sequence of 12S of Rana catesbeiana (Genbank accession number

MIRC12S): L7 (5 -TTT GGT CCT AGC CTT ATT ATC - 7°) with H424 (5° - GGC ATA

GTG GGG TAT CTA ATC -3'). and L428 (5° -CTT AAA ACC CAA AGG ACT TGA -3')

Source : MNHN, Paris

56 ALYTES 22 (1-2)

Table 2. — Specimens examined in the present study. Collection abbreviations used: FD, Forest

Departement, Bangkok; FMNH, Field Museum, Chicago; KUHE, Graduate School of Human and

Environnemental Studies, Kyoto University Japon; MNHN, Muséum National d'Histoire

Naturelle, Paris; MV, field number of Michael Veith, specimens to be catalogued in the Field

Museum, Chicago; SI, Smithsonian Institution; WHT, Wildlife Heritage Trust, Colombo; ZFMK,

Zoologisches Forschungsinstitut und Museum A. Koenig, Bonn; ZMB, Zoologisches Museum der

Universität, Berlin; ZSM, Zoologische Staatssammlung, München. Genbank accession numbers

marked with an asterisk refer to sequences obtained by other authors.

; a Collection Genbank Collection Genbank

spee unis Ps number 165 [accession 16S] number 12S [accession 128

Buergeria buergeri ë RUHE 26541 KUHE 26541

Bufo melanostictus - AF249061 U52P1

Ceratobatrachus guentheri … | Solomon Islands ZMFK 50484 ZMFK 50484

Chaparana fansipani Sapa, Vietnam MNHN 1999.5818 MNHN 1999.5818

Eleutherodactylus cuneatus - x86310 10944

Euphlyetis eyanophlyetis — |Cochin, India/ Sri Lanka |MNHN2000.650 |AYO14366 |WHT 0043€

Fejervarya cancrivora Sumatra. FMNH256692 |AYO14380 | FMNH 256692

Fejervarya limnocharis Laos / Laos MNHN 19973932 |AF215416 | MNHN 1997.5608

‘Hoplobatrachus chinensis | Laos / Laos MNEN 1997.4900 |AYO14368 | MNHN 1997.5691

Ingerana baluensis Malaysia FMNH 231085 FMNH 231085

Lankanectes corrugatus Sri Lanka WHT 0020C WHT 0020C

Limnonectes blythit Phang Nga, Thailand MNHN 1998.19 MNEHN 1998.19

Limnonectes gvldenstolpei Vietnam MNHN 19984150 MNHN 19984150

Limnonectes kuhli Laos /Phang Nga, Thailand | MNHN 19973904 |AF215415 |FD P921

Limnonectes macrocephalus | Leyte, Philippines M 365 MV 365

Limnonectes woodworthi Leyte, Philippines MNHN 2000.612 MNHN 2000.612

Occidozyga lima Philippines / Laos ZMB 50910 AF215398 MNHN 19996113

Nyctibatrachus sp. Ootacamund, India AF215397 AF215199

Myctibatrachus £ aliciae - AF249018 AF249063

Nyctibatrachus major - AF249017 AF249052

Paa bourreti Sapa, Vietnam MNHN 1999.5861 MNHN 19995861

Polypedates eques Sri Lanka WHT 0036C WHT 0036C

ÆRana catesbeiana - X12841 | MIRCI2S

Rana temporaria Koblenz, Germany / France | ZFMK 69883 AF2413S MNEN 1998.5

[Sphaerotheca pluvialis Myanmar ST 520491 SI 520491

Laylorana hascheana Vietnam MNHN 1997.5355 MNHN 1997.5355

with H898 (5° - ACC ATG TTA CGA CTT GCC TCT - 3). For the 168 rRNA gene, we

amplified one fragment unsing the primers (of PALUMBI et al., 1991) 16SA-L (light chain; 5° —

CGC CTG TTT ATC AAA AAC AT - 3°) and 16SB-H (heavy chain; 5 - CCG GTC TGA

ACT CAG ATC ACGT - 3°). We followed the PCR conditions as given in VENCES et al.

(2000b) and the PCR products were purified and sequenced using automatic sequencers (ABI

377 or CEQ 2000 Beckmann). The sequences (see tab. 2 for Genbank accession numbers) were

aligned using the program Se-Al (RAMBAUT, 1995), and by taking account of the secondary

structure of molecules (KJER, 1995, 1997). Gapped positions were excluded from analysis.

Two outgroups and three ingroup sequences (Eleutherodactylus cuneatus, Bufo melanostictus.

Rana catesbeiana, Nyctibatrachus major, Nyctibatrachus ef. aliciae) from Genbank were

further added to the alignment.

To assess whether the different gene fragments could be submitted to combined analysis,

we tested all possible combinations using the partition homogeneity test (parsimony method

of FARRIS et al., 1995), as implemented in PAUP*, version 4b8 (SwWorFrorD, 2001). Prior to

Source : MNHN, Paris

DELORME & AL. 57

phylogenetic reconstruction, we explored which substitution model fits our sequence data the

best using the program MODELTEST (PosapA & CRANDALL, 1998). The presence of a

significant phylogenetic signal was estimated using the permutation-tailed-probability (PTP)

test with 100 replicates implemented in PAUP*.

Phylogenetic analyses were carried out using PAUP*. We calculated maximum parsi-

mony (MP) and maximun likelihood (ML) trees. In the MP analyses we conducted heuristic

searches with initial trees obtained by simple stepwise addition, followed by branch swapping

using the TBR (tree bisection-reconnection) routine implemented in PAUP*. Ten random

addition sequence replicates were carried out. The ML trees were obtained using heuristic

searches, using the substitution model proposed by MODELTEST.

Following HEDGES (1992), 2000 bootstrap replicates (FELSENSTEIN, 1985) were run in the

MP analysis whereas only 100 (full heuristic) ML bootstrap replicates were possible because

of computational constraints.

Furthermore, we used Bayesian inference in the program MrBayes 2.01 (HUELSENBECK &

Ronquisr, 2001). We run four simultaneous Metropolis-coupled Monte Carlo Markov

chains for 500,000 generations, sampling a tree every ten generations. The initial set of

generations needed before convergence on stable likelihood values (“burnin”) was set at

50,000 (10 %) based on empirical evaluation.

RESULTS

A chi-square test did not contradict homogeneity of base frequencies across taxa (df =

78; P > 0.9). The partition homogeneity test did not reject the null hypothesis of congruence

of the included gene fragments (1000 replicates; P > 0.5), thus not contradicting their

suitability for combination in phylogenetic analysis. The PTP test resulted in a significant

difference (P = 0.01) between the most parsimonious tree and trees generated from random

permutations of the data matrix, demonstrating presence of significant phylogenetic signal.

After exclusion of gapped states, of 1122 characters included in the analysis, 504 were

constant, 179 variable but parsimony-uninformative, and 439 variable and parsimony-

informative. Maximum parsimony analysis found one most parsimonious tree (2422 steps;

consistency index 0.414, retention index 0.412). MODELTEST proposed a Tamura-Nei

substitution model with a gamma shape parameter of 0.433, a proportion of invariable sites

of 0.190, and user-defined substitution rates (A-G, 3.7290; C-T, 7.5587; all other rates, 1) and

base frequencies (A, 0.3857; C, 0.2267; G, 0.1407; T, 0.2469).

The ML analysis using the settings proposed by MODELTEST resulted in the tree

shown in fig. 1. AII species of Limnonectes (excluding L. corrugatus) were grouped as à

homophyletic group, in which Taylorana hascheana was also included. Species of Fejervarya

(once a subgenus of Linmonectes) did not directly cluster with Limnonectes. The included taxa

placed by DuBois (1992) in the Dicroglossinae (a subfamily of the Ranidae) were a homo-

phyletic lineage, which also included the genera Paa and Chaparana placed by the latter

author in the Paini, a tribe then referred to the Raninae but later transferred into the

Dicroglossinae (Dugois et al., 2001; DuBois, 2003: JIANG & ZHOU, in press). Lankanectes

Source : MNHN, Paris

58 ALYTES 22 (1-2)

BAITT j—— Limmonectes paramacrodon

+ D bimnonectes biythi

es/100 — Limnonectes macracephalus

s71100 * Limnonectes woodworthi

8515r—— Limnonectes gyidenstoipei

- Tayiorana hascheana

6860 Limnonectes kunli

ss00 d— Chaparans fansipani

aus + LL Paabouienger

: 2687—— Fojervarya cancrivora

Fe ———_—— Fojerarya lmnochanis

ef sphacrothecs puviais

Euphiycts cyanophiyctis

=. = LL Hoplobatrachus chinensis

ones Nycibatrachus sp.

10000 + L iyyctbatrachus major

= D nycribatrachus cf aiciae

s0/00 Rana temporaria

L_ Rana catesbeiana

pue Te —

se Buergeria buergeri

c —_— Polypedates eques

Dicroglossinae

10087

Ceratobatrachus guentheni

ingerana baluensis

Occidozyga lima

0.05 substiutions / site

Fig. 1.- Maximun likelihood phylogram calculated by PAUP* using a TRN + I + G substitution model

selected by MODELTEST, based on 1198 nucleotides of the mitochondrial 12S and 165 rRNA

genes. Numbers are bootstrap values (in percent; 100 and 2000 replicates) of maximum likelihood

and maximum parsimony analyses. Asterisks mark nodes that received posterior probability

values of 99-100 % in a Bayesian analysis. Values below 50 % are not shown. The arrow marks the

Sri Lankan species Lankanectes corrugatus which previously was considered as member of the

genus Limnonectes in the subfamily Dicroglossinae. Bufo melanostictus and Eleutherodactylus

cuneatus Were used as outgroups (not shown).

corrugatus Was placed as sister group to a clade containing Nyctibatrachus and Rana, the

type-genus of the Raninae. Occidozyga lima was the outgroup to all other ranoids included,

confirmingits strong differentiation in the mitochondrial rRNA genes already emphasized by

MaRMayOU et al. (2000). Most of these groupings were also found in MP and NJ analyses

{not shown) and received moderate to strong bootstrap support (fig. 1).

DISCUSSION

RELATIONSHIPS OF LANKANECTES CORRUGATUS AND ENDEMISM IN SOUTH ASIAN ANURANS

Our results confirm again the existence of a well-defined clade Dicroglossinae among the

Ranidae, and provide support for at least three subclades in this subfamily, which can

Source : MNHN, Paris

DELORME & AL. 59

taxonomically be considered as tribes (Dugois, 2003). The genera included in the present

study were distributed as follows among these lineages: (1) Limnonectini (Limnonectes and

Taylorana), (2) Dicroglossini (Euphlyctis, Fejervarya, Hoplobatrachus and Sphaerotheca); (3)

Paini (Paa and Chaparana).

Our data provide strong evidence that Lankanectes corrugatus does not belong to the

Limnonectini, let alone to the Dicroglossinae. In our analysis this species was instead placed

close to Nyctibatrachus and Rana. However, bootstrap support for this grouping was low.

Weak indications for relationships of Lankanectes to Nyctibatrachus and Rana were also

apparent from the results of BossuyT & MILINKOVITCH (2000) and VENCES et al. (2000c).

However, morphologically Lankanectes is well distinguished from these genera by several

divergent characters such as its forked omosternum (unforked in Rana) or the rare paedo-

morphic presence of a functional lateral-line system in adults (Dugois & OHLER, 2001), a

character shared with the dicroglossine Euphlyctis and the basal genus Occidozyga but absent

in Rana or Nyctibatrachus.

The data set of BossuyT & MILINKOVITCH (2000) contained almost 2700 nucleotides of

mitochondrial and nuclear genes, but their analyses did nevertheless not provide high support

for relationships of Lankanectes to Rana or Nyctibatrachus. Furthermore, no indications of

close relationships of the species to other South Asian endemics (/ndirana, Micrixalus,

Nannophrys) have been found (Bossuyr & MiLiNKoviTCH, 2000; VENCES et al., 2000c).

ROELANTS et al. (2004)'s results, based on a much smaller sample of Limnonectes than ours,

also show that L. corrugatus does not belong in the Dicroglossinae clade and does not have

any close relation with the Raninae. Lankanectes corrugatus would be placed in basal position

of the Ranidae with the genus Nyctibatrachus, but no strong support exists for this relation.

Therefore we are inclined to assume that L. corrugatus is the sole known representative of a

further endemic South Asian ranid lineage. This implies recognition of a new genus for this

species, which may be at least provisionally placed in a subfamily Lankanectinae, of unclear

affinities (Dugois & OHLer, 2001; Dugois, 2003; ROELANTS et al., 2004). These data strongly

confirm the importance of South Asia as a center of endemism of basal ranid lineages

(Bossuyr & MiLiINKOVITCH, 2001; ROELANTS et al., 2004). They also show that much more

remains to be learned on the relationships among basal ranid lineages. Certainly, a much

larger amount of molecular data is needed before a comprehensive scenario of the evolution

of this group can be drawn.

GENERIC TAXONOMY OF LIMNONECTINI

Incidentally, our results provide additional support to previous data regarding rela-

tionships within the South-East Asian Limnonectini clade. All South-East Asian species of

Limnonectes we surveyed were included in a single subclade of the dicroglossine lineage. In

this group, Limnonectes gyldenstolpei (see OnLer & DuBois, 1999) was placed as sister group

of Taylorana hascheana. The topology of our tree, as well as those of other recent studies

(EMERSON et al., 2000; Evans et al., 2003), indicate paraphyly of the genus Limnonectes as

currently understood (DuBois & OHLER, 2001). This does not necessarily imply that Taylo-

rana should be synonymized with Limnonectes. The genus Taylorana is well-defined by

presence of male mating call (absent in Limnonectes) and of direct development (TAYLOR,

Source : MNHN, Paris

60 ALYTES 22 (1-2)

1962; Onrer et al., 1999). This latter character is particularly relevant in anuran generic

taxonomy (Duois, 1987, 1988, 2004). According to the precise suggestions of DuBois (2004),

if confirmed the cladograms referred to would rather suggest that, beside Taylorana, three

genera at least should be recognized in the Limnonectini: (1) a first one, for which the nomen

Elachyglossa Andersson, 1916 is available, including the species listed by OHLer & DuBois

(1999) and possibly others such as Rana laticeps Boulenger, 1882; (2) a second one, that

should retain the nomen Limnonectes Fitzinger, 1843, for L. kuhlii and a few other species; (3)

a third one, including most species of the grunniens and microdiscus groups of DuBois (1987:

63) or of the subclades 2, 3 and 4 of EMERSON et al. (2000) and Evans et al. (2003). No generic

nomen has been associated with the latter group until now, but such a nomen might be

available. Recent re-interpretation of morphological characters of the species originally

described as Rana delacouri by ANGEL (1928) and later placed in the subgenus Chaparana

(Annandia) by Dugois (1992), now suggests that this species may be closer to Limnonectes

blythii than to members of the tribe Paini (Dugois & OHLER, in preparation). As this species

is the type-species of Annandia Dubois, 1992, the latter nomen might be available for the third

genus outlined above. At any rate, until the cladistic relationships of Rana delacouri are

clarified, it would appear better not to create a generic nomen for the latter group.

“FANGED”” FROGS AND THE CONCEPTS OF MONOPHYLY, HOMOPHYLY AND HOLOPHYLY

EMERSON et al. (2000: 136) wrote that “the fanged frogs constitute a monophyletic group”

and that “it seems appropriate, in the future, to refer to these frogs as members of the genus

Limnonectes”. While doing so, however, they did not provide a list of taxa that they referred

to this genus, so that one can infer that they probably adopted DuBois’s (1992) concept of the

latter, thus including the fang-bearing species Lankanectes corrugatus (as Limnonectes corru-

gatus).

However, our data once again show that the latter species is not a member of Limnonec-

tes, and that this genus as it has been understood until the work of DuBois & OHLer (2001) is

not monophyletic. Despite this apparent contradiction, the statement of EMERSON et al.

(2000) regarding monophyly of “fanged” frogs was not incorrect: actually, all species studied

by these authors appeared as a clade in their molecular analysis, and were not para- or

polyphyletic relative to the other taxa studied. This problem is a more general one in

phylogenetic studies, especially those relying on molecular data. In many cases, because of

material limitations, such studi n include only some of the species of the group whose

monophyly is to be tested. However, as noted by Bossuyr & DuBois (2001: 4), the large impact

of species sampling on cladistic analysis should not be underevaluated. This has long been

known for cladistic studies based on morphology: “Ideally, all known taxa of a group should

be included in analysis, since omission can lead to misinterpretation of transformation series

(...) and of relationships (...)” (ARNOLD, 1981: 29).

Part of the confusion is mostly semantic, being rooted in the use of the unclear term

monophyletic. This term was introduced in scientific literature by HAECKEL (1868) as an

antonym to polyphyletic, but HENNIG (1950) redefined it as an antonym to both po/yphyletic

and paraphyletic, a new concept introduced by him. The Hennigian definition of a monophy-

letic group, adopted by many current authors, can be worded as follows: “A group that

Source : MNHN, Paris

DELORME & AL. 61

includes a common ancestor and all of its descendants” (WiLey, 1981: 84). In logical terms,

this means that a monophyletic group has two qualifications, uniqueness (non-polyphyly) and

completeness. Like all double concepts, this can be sorted in two distinct concepts, for which,

in order to avoid the confusions linked to the use of the unclear term monophyletic, two

distinct terms have been proposed: homophyletic (Dusois, 1986, 1988) for unique or non-

polyphyletic, and holophyletic (AsaLocKk, 1971) for unique and complete. Many authors now

use the term monophyletic for the latter concept, but then, if they claim that a group is unique

and complete, they should provide the complete list of included taxa, at least among the taxa

then known and recognized as valid by zoologists in the taxon (family or even higher taxon)

under study.

Because many clades certainly contain extinct species, sometimes in considerable num-

ber, whose fossils will never be found, absolute completeness of sampling of taxa will remain

impossible in many zoological groups. Even the goal of completeness of sampling of extant

taxa is often unrealistic because, despite the ongoing and even accelerating high rate of

discovery of new species, it is clear that many or most extant animal species are not even

known (and certainly not taxonomically described) yet. But a different thing is to realise that,

among the species that we have discovered and described, stating that a group is complete

means that we have identified all those that are members of a given clade. This will be

done only when all species have been properly studied with the techniques (molecular,

morphological or other) that we use to allocate them to clades. The example of Lankanectes

shows that any single species, once seriously studied, may contradict our previous hypotheses.

In this case, one can argue that its strange geographical distribution might have indicated long

ago that L. corrugatus was an intruder in Limnonectes, but this is not always the case. Thus, in

the same frog group, the case of the species Rana delacouri mentioned above, if confirmed,

would illustrate a rather frequent situation in which neither geographical distribution nor

overall morphology had allowed to suspect wrong cladistic allocation of a species: in such

cases, the proper study of a single species may have nomenclatural implications, e.g. if this

species is the type of a nominal genus.

Therefore, in many cladistic analyses, especially molecular, as only a partial list of taxa

has been actually studied, it would be more prudent and exact to state that the group

composed of these studied taxa is somophyletic, i.e. non-polyphyletie, without going further

in inferring the actual cladistic position of taxa whose existence is known but that were not

examined in the study. Only when all known potentially related taxa have been properly

studied and allocated a place in the cladogram is it justified to state that a group appears

non-polyphyletic and complete, i.e. *monophyletic” or, better as fully unambiguous, Aolophy-

letic.

RÉSUMÉ

Depuis plus de 15 ans, le genre Limnonectes (Ranidae, Dicroglossinae, Limnonectini) a

regroupé plusieurs dizaines d'espèces du Sud-Est de l'Asie, ainsi qu’une espèce isolée prove-

nant du Sri Lanka, L. corrugatus. Nous avons analysé 1198 paires de base des gènes ARNr

mitochondriaux 12S et 16S de L. corrugatus, des représentants de tous les principaux

sous-groupes de Limnonectes et de plusieurs genres qui semblent proches. Les données ont

Source : MNHN, Paris

62 ALYTES 22 (1-2)

permis d’exclure clairement l'espèce du Sri Lanka des Limnonectes du Sud-Est de l'Asie. De

plus, celle-ci semble se rapprocher des genres Rana et Nyctibatrachus, ce qui étaye la recon-

naissance récente du genre monotypique Lankanectes pour L. corrugatus. Les spécialisations

morphologiques de cette espèce confirmant qu'elle serait la seule représentante connue d’une

lignée de Ranidés endémique de l’Asie du Sud, une région de grande importance comme

centre de diversité et d’endémisme de cette famille. Nos données suggèrent également quel-

ques commentaires sur la taxinomie générique de la tribu des Limnonectini. En contradiction

avec les précédents résultats sur le monophylétisme de Limnonectes, elles mettent l'accent sur

un problème général de terminologie dans les études phylogénétiques. Nous proposons

d'utiliser le terme homophylétique pour des groupes pour lesquels les données disponibles ne

sont pas contradictoires avec l’hypothèse de monophylétisme, mais dont le contenu est encore

incomplet ou incertain.

ACKNOWLEDGEMENTS

For the help received in Sri Lanka, we thank Rohan Pethiyagoda and Kelum Manamendra-

Arachchi. We are grateful to Annemarie Ohler, Stéphane Grosjean and Renaud Boistel (Paris) and to

Alan Resetar (Chicago) for providing tissues of several Asian and Oriental frog taxa. Franco Andreone

provided useful comments on the manuscript. This work was supported by the PPF “Faune et flore du

sud-est asiatique” of the Paris Museum.

LITERATURE CITED

ANGEL, M. F., 1928. - Sur une espèce nouvelle de grenouille du Tonkin, appartenant au genre Rana. Bull

Mus. natn. Hist. nat., 34: 319-320.

ARNOLD, E. N,, 1981. — Estimating phylogenies at low taxonomic levels, Z. zool. Syst. Evol.-forsch., 19:

1-35.

AsHLock, P. D. 1971. - Monophyly and associated terms. Syst. Zool., 20: 63-69.

BARRON, E. J., HARRISON, C. G. À., SLOAN, 1. L., II & Hay, W. W., 1981. - Paleogeography, 180 million

years ago to the present. Ecol. Geol. Helveticae, 74: 443-470.

Buu, S. D. & Bossuvr, F, 2003. - New frog family from India reveals an ancient biogeographical link with

the Seychelles. Nature, 425: 711-714.

BLOMMERS-SCHLÔSSER, R. M. A., 1993, - Systematic relationships of the Mantellinac Laurent 1946

(Anura Ranoïdea). Ethol. Ecol. Evol., 5: 199-218.

Bossuyr, F. & Durois, À. 2001. — A review of the frog genus Philautus Gistel, 1848) (Amphibia, Anura,

Ranidae, Rhacophorinae). Zeylanica, 6 (1): 1-112.

j M. C:, 2000. - Convergent adaptive radiations in Madagascan and Asian

sindicators of early Tertiary “Out-of-India” dispersal of vertebrates. Science,

A.. 1920. - À monograph of the South Asian, Papuan, Melanesian, and Australian frogs

us Rana. Rec. indian 1 : 1-226.

Duois, A. 1981. — Liste des genres et sous-genres nominaux de Ranoïdea

monde, avec identification de leurs es

(Amphibiens, Anoures) du

ypes: conséquences nomenclaturales, Mont. zool. ital.,

Source : MNHN, Paris

DELORME & AL. 63

ne 1986. - A propos de l'emploi controversé du terme “monophylétique”: nouvelles propositions. Bull.

mens. Soc. linn. Lyon, 55: 248-254.

1987. - Miscellanea taxinomica batrachologica (1). Alytes, 5 (1-2): 7-95.

1988. - The genus in zoology: a contribution to the theory of evolutionary systematics. Mém. Mus.

nain. Hist. nat., (A), 140: 1-123.

ee 1992. — Notes sur la classification des Ranidae (Amphibiens Anoures). Bull. mens. Soc. linn. Lyon, 61:

305-352.

ee 1999. - Miscellanea nomenclatorica batrachologica. 19. Notes on the nomenclature of Ranidae and

related groups. Alyres, 17 (1-2): 81-100.

_— 2003. - True frogs (Ranidae). /n: M. HUTCHINS, W. E. DUELLMAN & N. SCHLAGER (ed.), Grzimek's

animal life encyclopedia, 24 edition, Volume 6, Amphibians, Farmington Hill, Michigan, Gale

Group: 245-264.

ne 2004. - Developmental pathway, speciation and supraspecific taxonomy in amphibians. 2. Develop-

mental pathway, hybridizability and generic taxonomy. A/ytes, 22 (1-2): 38-52.

Dumois, A. & OHLER, À., 2000. - Systematics of Fejervarya limnocharis (Gravenhorst, 1829) (Amphibia,

Anura, Ranidae) and related species. 1. Nomenclatural status and type-specimens of the nominal

species Rana limnocharis Gravenhorst, 1829. Alytes, 18 (1-2): 15-50.

a 2001. - A new genus of Ranidae (Amphibia, Anura) from Sri Lanka. Alytes, 19 (2-4): 81-

106.

Dusois, A., OHLER, A. & Buu, S. D., 2001. - À new genus and species of Ranidae (Amphibia, Anura)

from southern India. A/yres, 19 (2-4): 53-79.

DuELLMAN, W. E. & TRUEB, L., 1985. - Biology of amphibians. New York, McGraw-Hill, “1986”: i-xix +

1-670.

Durra, S. K., 1997. — Amphibians of India and Sri Lanka. (checklist and bibliography). Bhubaneswar,

Odyssey Publishing house: [i-iii] + i-xiii + 1-342 + i-xxii.

Durra, S. K. & MANAMENDRA ARACHCHI, K., 1996. — The amphibians fauna of Sri Lanka. Colombo,

Wildlife Heritage Trust of Sri Lanka: 1-232.

EMERSON, $. B. & BERRIGAN, D., 1993. - Systematics of Southeast Asian ranids: multiple origins of

voicelessness in the subgenus Limnonectes (Fizinger). Herpetologica, 49 (1): 22-31.

EMERSON, S. B., INGER, R. F. & IskANDAR, D., 2000. - Molecular systematies and biogeography of the

fanged frogs of Southeast Asia. Mol. Phyl. Evol., 16: 131-142.

EMERSON, S. & VORIS, H., 1992. - Competing explanations for sexual dimorphism in a voiceless Bornean

frog. Funct. Ecol., 6: 654-660.

EMERSON, S. & WaRD, R., 1998. - Male secondary sexual characters, sexual selection, and molecular

divergence in fanged ranid frogs of Southeast Asia. Zool. J. Linn. Soc., 122: 537-553.

Evans, B. J., BROWN, R. F., MCGUIRE, J. A. SUPRIATNA, J., ANDAYANT, N., DIESMOS, A. ISKANDAR, D.,

MELNick, D. J. & CANNATELLA, D. C., 2003. — Phylogenetics of fanged frogs: testing biogeo-

graphical hypotheses at the interface of the Asian and Australian faunal zones. Syst. Biol., 52 (6):

794-819.

FaRRIs, J. S., KALLERSIO, M., KLUGE, A. G. & BuLr, C., 1995. — Testing significance of incongruence.

Cladisties, 10: 315-319.

., 1985. - Confidence limits on phylogenies: an approach using the bootstrap. Evolution, 39:

El SON. 1995. — Phylogenetic relationships of

amphibian families inferred from DNA sequences of mitochondrial 12S and 16$ ribosomal RNA

genes. Mol. Biol. Evol., 12( 28-937.

HEDGES, S. B., 1992. - The number of replications needed for accurate estimation of the bootstrap P-value

in phylogenetic studies. Mol. Biol. Evol, 9: 366-369.

HENNIG, W., 1950. — Grundzüge einer Theorie der phylogenetischen Systematik. Berlin, Deutscher Zen-

tralverlag: 1-370. A

HUELSENBECK, J. P. & Roxquisr, F.. 2001. - MrBayes: Bayesian inference of phylogenetic trees.

Bioinformaties, 17: 754-755.

JiaNG, J-P. & Zrou, K. Y. in press. — Phylogenetic relationships among Chinese ranids inferred from

sequence data set of 125 and 165 rDNA. Herp. J.

Source : MNHN, Paris

64 ALYTES 22 (1-2)

Kuër, K. M., 1995. - Use of rRNA secondary structure in phylogenetic studies to identify homologous

positions: an example of alignment and data presentation from the frogs. Mol. Phyl. Evol., 4 (3):

314-330.

Den 1997. — An alignment template for amphibian 12S rRNA, domain III: conserved primary and

secondary structural motifs. J Herp., 31: 599-604.

KosucH, J., VENCES, M., DUBOIS, A. OHLER, À. & BÔHME, W., 2001. Out of Asia: mitochondrial DNA

evidence for an Oriental origin of tiger frogs, genus Hoplobatrachus. Mol. Phyl. Evol., 21: 398-407.

MARMAYOU, J., DUBOIS, A., OHLER, A., PASQUET, E. & TILLIER, A., 2000. — Phylogenetic relationships in

the Ranidae. Independent origin of direct development in the genera Philautus and Taylorana. C.

r Acad. Sci. Paris, Sciences de la Vie, 323: 287-297.

OuLer, À. & Duois, A. 1999. - The identity of Elachyglossa gyldenstolpei Andersson, 1916 (Amphibia,

Ranidae), with comments on some aspects of statistical support to taxonomy. Zoologica scripta, 28

(-4): 269-279.

OHLER, A., GROSIEAN, $. & Hoyos, J. M., 1999. - Observation de la construction d’un nid chez Taylorana

hascheana (Anura: Ranidae). Revue fr Aquariol., 26: 67-70.

PALUMBI, S. R., MARTIN, A., ROMANO, S., MCMILLAN, W. O., STICE, L. & GRABOWSKI, G., 1991. - The

simple fool's guide to PCR, Version 2.0. Privately published document compiled by S. Palumbi,

Dept. Zoology, Univ. Hawaï, Honolulu, HI, 96822.

, W, 1863. — Mittheilung über eine neue Schlangengattung, Styporhynchus, und verschiedene

andere Amphibien des zoologischen Museums. Monatsb. kônigl. Akad. Wiss. Berlin, 1863: 399-

413.

PosaDa, D. & CRANDALL, K. A., 1998. - MODELTEST: testing the model of DNA substitution.

Bioinformatics, 14: 817-818.

RAMBAUT, A., 1995. — Se-Al. Sequence Alignment Editor Version 1. d1. Department of Zoology. Univer-

sity of Oxford.

ROELANTS, K., JIANG, J. & BossUYT, F,, 2004. — Endemic ranid (Amphibia: Anura) genera in southern

mountain ranges of the Indian subcontinent represent ancient frog lineages: evidence from

molecular data. Mol. Phyl. Evol., 31: 730-740.

SrorEy, B. C., 1995. — The role of mantle plumes in continental breakup: case histories from Gondwa-

naland. Nature, 377: 301-308.

SToREY, M., MAHONEY, J. J., SAUNDERS, A. D., DUNCAN, R. A., KELLEY, S. P. & CorriN, M. F., 1995. -

Timing of hot spot-related volcanism and the breakup of Madagascar and India. Science, 167:

852-855.

SworrorD, D. L., 2001.- PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods). Version

4. Sunderland, Sinauer Associates.

TAYLOR, E. H., 1962. - The amphibian fauna of Thailand. Univ. Kansas Sci. Bull., 43: 265-599.

Vences, M., GLAW, F., KosucH, J., DAS, L. & Ver, M. 20004. - Polyphyly of Tomopterna (Amphibi

Ranidae) based on sequences of the mitochondrial 16$ and 12S rRNA genes, and ecologic

biogeography of Malagasy relict amphibian groups. /n: W. R. LOURENÇO & S. M. GOODMAN (ed.),

Diversité et endémisme de Madagascar, Paris, Mémoires de la Société de Biogéographie: 229-242.

VENCES, M., KosuCH, J., LÔTTERS, S., WIDMER, A., JUNGFER, K.-H., KÔHLER, J. & VErrH, M., 2000. -

Phylogeny and classification of poison frogs (Amphibia: Dendrobatidae), based on mitochondrial

168 and 12S ribosomal RNA gene sequences. Mol. Phyl. Evol., 14: 34-40.

VENCES, M., WANKE ODtERNA, G., KosuCH, J. & VerrH, M., 2000c. - Molecular and karyological data

on the south Asian ranid genera /ndirana. ctibatrachus and Nannophrys (Anura: Ranidae).

Hamadryad, 25: 75-82.

Wiuev, E. O., 1981. Phylogeneties. The theory and practice of phylogenetic systematies. New York, Wiley:

i-xv + 1-439.

Corresponding editor: Franco ANDREONE.

Source : MNHN, Paris

AINTTES

International Journal of Batrachology

published by ISSCA

EDITORIAL BOARD

Chief Editor: Alain Dupois (Laboratoire des Reptiles et Amphibiens, Muséum national d'Histoire naturelle,

25 rue Cuvier, 75005 Paris, France; <adubois@mnhn.fr>).

Deputy Editor: Thierry Lo (Laboratoire d'Ecologie animale, Université d'Angers, 2 boulevard Lavoisier,

49045 Angers Cedex, France; <thierry.lode@univ-angers.fr>).

Conservation Editor: Stephen J. RicHaRDs (Vertebrates Department, South Australian Museum, North Terrace,

Adelaide, S.A. 5000, Australia; <Richards.Steve@saugov.sa.gov.au>).

Editorial Board: Franco ANDREONE (Torino, Italy); Lauren E. BrowN (Normal, USA): Janalee P. CALDWELL

(Norman, USA): Ulisses CARAMASCHI (Rio de Janeiro, Brazil); Claude CoMBEs (Perpignan, France) : C.

Kenneth Don», Jr. (Gainesville, USA) : Günter GOLLMANN (Wien, Austria): Heinz GRILLITSCH (Wien,

Austria); Tim HALLIDAY (Milton Keynes, United Kingdom): W. Ronald HeyEr (Washington, USA):

Esteban O. LAviLLA (Tucumän, Argentina); Masafumi MatsuI (Kyoto, Japan); Alain PAGANO (Angers,

France); John C. POyNToN (London, England): Miguel VENCES (Konstanz, Germany).

Technical Editorial Team (Paris, France): Alain DuBots (texts); Roger BouR (tables); Annemarie OuLEr (figures).

Book Review Editor: Annemarie OHLER (Paris, France).

SHORT GUIDE FOR AUTHORS

(for more detailed Instructions to Authors, see Alytes, 1997, 14: 175-200)

Alytes publishes original papers in English, French or Spanish, in any discipline dealing with amphibians.

pese actes and hote Os el à OL dE DE ln on let

review articles, book reviews, comments and replies, and to papers based upon original high quality illustrations

{such as colour or black and white photographs), showing beautiful or rare species, interesting behaviours, etc.

The title should be followed by the name(s) and address(es) of the author(s). The text should be typewritten

or printed double-spaced on one side of the paper. The manuscript should be organized as follows: English

abstract, introduction, material and methods, results, discussion, conclusion, French or Spanish abstract,

acknowiedgements, literature cited, appendix.

Figures and tables should be mentioned in the text as follows: fig. 4 or tab. 4. Figures should not exceed 16 x

24 em. The size of the lettering should ensure its legibility after reduction. The legends of figures and tables

should be assembled on a separate sheet. Each figure should be numbered using a pencil.

References in the text are to be written in capital letters (BOURRET, 1942; GRAF & POLLS PELAZ, 1989; INGER

et al., 1974). References in the Literature Cited section should be presented as follows:

BoURRET, R., 1942. — Les batraciens de l'Indochine. Hanoï, Institut Océanographique de l'Indochine: i-x + 1-547,

GRAF, J.-D. & PoLLs PELAZ, M., 1989. - Evolutionary genetics of the Rana esculenta complex. In: R. M. DAWLEY

& J. P. BOGART (ed.), Evolution and ecology of unisexual vertebrates, Albany, The New York State Museum:

289-302.

INGER, R. F,, Voris, H. K. & Voris, H. H., 1974. - Genetic variation and population ecology of some Southeast

Asian frogs of the genera Bufo and Rana. Biochem. Genet., 12: 121-145.

Manuscripts should be submitted in triplicate either to Alain Duois (address above) if dealing with

amphibian morphology, anatomy, systematies, biogeography, evolution, genetics, anomalies or developmental

biology, or to Thierry Loné (address above) if dealing with amphibian population genetics, ecology, ethology or

life history, or to Stephen J. RICHARDS (address above) if dealing with conservation biology, including declining

amphibian populations or pathology. Acceptance for publication will be decided by the editors following review

by at least two referees.

If possible, after acceptance, a copy of the final manuscript on a floppy disk (3 or 5 }4) should be sent 10

the Chief Editor, We welcome the following formats of text processing: (1) preferably, MS Word (1.1 to 6.0, DOS

or Windows), WordPerfect (4.1 to 5.1, DOS or Windows) or WordStar (3.3 to 7.0); (2) less preferably, formated

DOS (ASCII) or DOS-formated MS Word for the Macintosh (on a 3 4 high density 1.44 Mo floppy disk only).

Page charges are requested only from authors having institutional support for this purpose. The publication

of colour photographs is charged. For each published paper, 25 free reprints are offered by ISSCA to the

author(s). Additional reprints may be purchased.

Published with the support of AALRAM

(Association des Amis du Laboratoire des Reptiles et Amphibiens

du Muséum National d'Histoire Naturelle, Paris, France).

Directeur de la Publication: Alain Durois.

Numéro de Commission Paritaire: 64851.