Calluella, Boulenger, 1882

Frost, Darrel R., Grant, Taran, Faivovich, Julián, Bain, Raoul H., Haas, Alexander, De Sá, Célio F. B. Haddad Rafael O., Channing, Alan, Wilkinson, Mark, Donnellan, Stephen C., Raxworthy, Christopher J., Campbell, Jonathan A., Blotto, Boris L., Moler, Paul, Drewes, Robert C., Nussbaum, Ronald A., Lynch, John D., Green, David M. & Wheeler, And Ward C., 2006, The Amphibian Tree Of Life, Bulletin of the American Museum of Natural History 2006 (297), pp. 1-291 : 1-291

Calluella as associated with the direct­devel­

oping Asterophryinae and any similarities

with Dyscophus as reflecting plesiomorphy.

We sampled one species each of the two

nominal genera: Calluella guttulata and Dys­

cophus guineti .

MELANOBATRACHINAE (3 GENERA, 4 SPE­

CIES): On the basis of geography alone (East

Africa [2 genera] and southern India [1 ge­

nus]), one would suspect that this is not a

monophyletic taxon. Nevertheless, the three

genera share an incomplete auditory appa­

ratus (convergent in Balebreviceps [Brevi­

cipitinae]; Largen and Drewes, 1989) and fu­

sion of the sphenethmoid with the paras­

phenoid (Parker, 1934). Savage (1973), fol­

lowed by Laurent (1986) and Dubois (2005),

placed Melanobatrachus in Microhylinae

and retained Hoplobatrachus and Parhoplo­

phryne in Hoplophryninae , but did so only Fig. 33. Trees of Cophylinae suggested by A, by discarding absence of the auditory appa­ Blommers­Schlösser and Blanc (1993), on the ba­ ratus and fusion of the sphenethmoid to the sis of nine morphological transformation series, parasphenoid, as convergences, without ofrooted (by implication) on Dyscophinae and Sca­ fering specific characters that conflicted with phiophryninae. The figure is redrawn with branch­ these as synapomorphies. Although we are es collapsed that were unsupported by evidence suspicious of the monophyly of this taxon, in the original; B, Andreone et al. (2004 ‘‘2005’’), we stick with the most parsimonious hypothbased on 1,173 bp of 12S and 16S rRNA mtDNA. esis (monophyly of Melanobatrachinae , sen­ This tree is redrawn to note only monophyletic su lato) until alternative evidence emerges. genus­group taxa. Alignment was made using the Clustal option in Sequence Navigator (Applied Apparently based on information provided Biosystems), with cost functions for alignment not for Hoplophryne by Barbour and Loveridge provided. All sections that could not be aligned, (1928) and Noble (1929), Parker (1934) genincluding those with three of more gaps in one or eralized that all members of his Melanobamore taxa, were excluded from analysis. Whether trachinae lack a free­swimming tadpole, the gaps were treated as unknown or evidence was larvae with ‘‘metamorphosis taking place on not stated. The Tamura­Nei substitution model land, but not in an egg’’. No reproductive or was selected for maximum­likelihood analysis of developmental data on Parahoplophryne or aligned data. The tree was rooted on Scaphio­ Melanobatrachus have been published (Dalphryne (not shown). Quotation marks around

try and Martin, 1997). Thibaudeau and Altig names denotes nonmonophyly.

(1999) listed Melanobatrachus and Parho­

plophryne as having endotrophic larvae, pre­ (1993) or Andreone et al. (2004 ‘‘2005’’). sumably because of the earlier statement by Species of Cophylinae have nidicolous lar­ Parker (1934). McDiarmid and Altig (1999: vae (Blommers­Schlösser and Blanc, 1991; 13), however, listed Hoplophryne as exo­ Glaw and Vences, 1994). trophic, because Barbour and Loveridge DYSCOPHINAE (2 GENERA, 10 SPECIES): The (1928: 256) reported vegetable matter in the Madagascan Dyscophinae is distinguished guts of larvae and because R. Altig examined from most other microhylid subfamilies by AMNH larvae of Hoplophryne and inferred retaining maxillary and vomerine teeth, oth­ that they could feed (R.W. McDiarmid, pererwise known only in Cophylinae and Gen­ sonal commun.). Laurent (1986) reported the in

his yophryninae, both of which are procoelous taxon ( Parhoplophryne and Hoplophryne rather than diplasiocoelous (Parker, 1934) as his Hoplophrynnae; Melanobatrachus in in Dyscophinae . Savage (1973) had regarded Microhylinae ) as procoelous, unlike most

80 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 297

ed the Old World and New World compo­

nents separately, implying some kind of tax­

onomic division. This was followed, without

discussion, by Dubois (2005), who recog­

nized Gastrophryninae for the New World

component and Microhylinae for the Old

World component. We are not aware of any

evidence in support of this arrangement so

we retain the old taxonomy. Of the 30 nom­

inal genera we were able to sample represen­

tatives of 14: Chaperina fusca , Ctenophryne

geayei, Dasypops schirchi , Dermatonotus

muelleri , Elachistocleis ovalis , Gastrophryne

elegans, G. olivacea , Hamptophryne bolivi­

ana , Kalophrynus pleurostigma , Kaloula

pulchra, Microhyla heymonsi , Microhyla sp. ,

Micryletta inornata , Nelsonophryne aequa­

torialis, Ramanella obscura , and Synaptur­

anus mirandaribeiroi ). We were not able to

sample Adelastes , Altigius , Arcovomer ,

Chiasmocleis , Gastrophrynoides , Glyphog­

lossus, Hyophryne , Hypopachus , Metaphry­ Fig. 34. Tree of New World microhylids by nella, Myersiella , Otophryne , Phrynella , Re­ Wild (1995) based on a parsimony analysis of 14 lictovomer, Stereocyclops , Syncope , and morphological transformations series, outgroups Uperodon . Most of these appear to be clus­ and evidence for ingroup monophyly not speci­ tered with sampled taxa. The exclusion of fied. Otophryne and Uperodon , however, is partic­

ularly regrettable. Our sampling will not alother ranoids so this may also be synapo­ low detailed elucidation of the evolution of morphic. Unfortunately, we were able to life­history strategies. Adelastes , Altigius , sample only Hoplophryne rogersi and so will Gastrophrynoides , Hyophryne , Kalophrynus not be able to comment on the monophyly (nidicolous), Myersiella (direct developof Melanobatrachinae . ment), Phrynella , Synapturanus (nidicolous), MICROHYLINAE (30 GENERA, 133 SPECIES): and Syncope (nidicolous) have endotrophic The American and tropical Asian Microhy­ larvae that exhibit (or are suspected to exlinae have free­swimming tadpoles (except hibit) various degrees of truncation of larval for a few species, such as Myersiella mi­ development (Thibaudeau and Altig, 1999). crops, that have direct development; Izeck­ That we lack representatives of about half of sohn et al., 1971). Although microhylines these is lamentable, but our results will procan be morphologically characterized, they vide an explicit starting point for future, have no known synapomorphies, and their more detailed studies. The remaining genera monophyly is deeply suspect. According to have exotrophic larvae of typical microhylid Parker (1934), maxillary and vomerine teeth morphology (Altig and McDiarmid, 1999). are absent (as in several other extra­Mada­ PHRYNOMERINAE (1 GENUS, 5 SPECIES): The gascar subfamilies); the vomer is much re­ African Phrynomerinae is diagnosable from duced and usually divided; the sphenethmoid Microhylinae solely by possessing intercalais divided or absent; and the vertebrae are ry cartilages between the ultimate and pendiplasiocoelous (or rarely procoelous). Wild ultimate phalanges (Parker, 1934). Like most (1995) provided a cladogram of New World other ranoids it is diplasiocoelous. Of this genera (fig. 34), but this assumed that the small taxon we sampled Phrynomantis bifas­

New World group is monophyletic and was ciatus. Phrynomantis typically has aquatic, unclear about the outgroup(s) used to polar­ exotrophic microhylid larvae (Altig and ize the transformations. Laurent (1986) treat­ McDiarmid, 1999).

2006 FROST ET AL.: AMPHIBIAN TREE OF LIFE

81

‘‘ RANIDAE ’’ (CA. 54 GENERA, 772 SPECIES): Ceratobatrachus ). Ceratobatrachinae repre­ Ranidae is a large ranoid taxon, that is likely sents the direct­developing part of Cornuferparaphyletic with respect to Mantellidae and inae sensu Noble (1931) and Platymantinae Rhacophoridae —at least on the basis of mo­ of later authors (e.g., Savage, 1973; Laurent, lecular evidence (Vences and Glaw, 2001; 1986). Those taxa formerly included in Roelants et al., 2004; Van der Meijden et al., Cornuferinae or Platymantinae that exhibit 2005). Ford and Cannatella (1993; fig. 14 ) unforked omosterna and/or free­living tadsuggested that the group is paraphyletic, or, poles (what are now Amolops , Huia, Merat least, that it does not have recognized syn­ istogenys, Staurois , Hylarana [sensu lato], apomorphies. Nevertheless, Haas (2003; fig. and Micrixalus ) are now placed in Raninae 15) suggested the following to be synapo­ or Micrixalinae. Batrachylodes is inferred to morphies for Ranidae , excluding other ran­ have direct development (Noble, 1931; oids: (1) cartilaginous roofing of the cavum Brown, 1952; Duellman and Trueb, 1986; cranii present as taenia transversalis and me­ Thibaudeau and Altig, 1999), but unlike othdialis; (2) free basihyal present; and (3) fir­ er members of Ceratobatrachinae, Batrachymisterny (convergent elsewhere in Haas’ lodes has an entire omosternum (rather than tree). being forked). Noble (1931) regarded Batra­

Laurent (1986) included the mantellines chylodes as derived from his Cornufer (5 and rhacophorids in his Ranidae , a content Platymantis ) and, by inference, exhibiting dithat allows at least two other characters (dis­ rect development. Because of the character tinctly notched tongue and bony sternal conflict of omosternum shape and life­hisstyle) to be considered as possible synapo­ tory, Brown (1952) regarded Batrachylodes morphies (Ford and Cannatella, 1993). as related either to ‘‘ Hylarana ’’ (exotrophic, (These are, however, incongruent with char­ entire omosternum) or to the Ceratobatraacters suggested by Haas, 2003). chus group (direct­developing, forked omos­

Dubois and coauthors (Dubois, 1992; Du­ ternum). Laurent (1986) treated Batrachylobois and Ohler, 2001; Dubois et al., 2001; des as a member of Raninae, although Bou­ Dubois, 2005) suggested a taxonomy of 11– lenger (1920) had noted the intraspecific 14 subfamilies of uncertain monophyly or re­ plasticity of omosternum shape, the only evlationship with respect to each other. For dis­ idence supporting placement of Batrachylocussion, we recognize Dubois’ subfamilies, des in Raninae. This arrangement was acexcept as noted. As discussed by Inger cepted by Dubois (1987 ‘‘1985’’), although (1996), the diagnostic features supporting subsequently, Dubois (2005) transferred Ba­ Dubois’ (1992) classification at the time of trachylodes out of Raninae and into Ceratothat writing frequently reflected overgener­ batrachinae, presumably on the basis of the alized and postfacto approximations for clus­ direct development. Our analysis should proters that were aggregated with overall simi­ vide more evidence on the placement of this larity, not synapomorphy, as the organizing taxon. principle. The relationships suggested by this Dubois (1992) recognized Ceratobatrachitaxonomy (and Dubois, 2005, as well) can ni within his Dicroglossinae , but later (Dube at variance with evidence of monophyly, bois et al., 2001) considered it to be a subnotably evidence from DNA sequences (Em­ family, of unclear relationship to Dicrogloserson and Berrigan, 1993; Bossuyt and Mil­ sinae. Even later, Dubois (2003) stated, on inkovitch, 2000; Emerson et al., 2000b; Mar­ the basis of unpublished molecular data, that mayou et al., 2000; Biju and Bossuyt, 2003; Ceratobatrachini is a tribe within Dicroglos­ Roelants et al., 2004), so this taxonomy re­ sinae. Van der Meijden et al. (2005) presentquires careful evaluation. ed DNA sequence evidence that Ceratoba­

CERATOBATRACHINAE (6 GENERA, 81 SPE­ trachus is outside of Dicroglossinae , and on CIES): Ceratobatrachinae is composed of di­ that basis (Dubois, 2005) once again emrect­developing species found from western braced the subfamilial rank Ceratobatrachi­ is

China (i.e., Ingerana ) to the Indo­Australian nae. Roelants et al. (2004; fig. 35), in a study archipelago ( Batrachylodes , Discodeles , Pal­ of predominantly Indian taxa, provided momatorappia, Platymantis , and the monotypic lecular evidence that suggest that Ingerana 82 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 297 2006 FROST ET AL.: AMPHIBIAN TREE OF LIFE

83

in Occidozyginae , rather than in Ceratobatra­ neous group of southern African ranoids, inchinae, although Dubois (2005), without dis­ cluding Afrana , Cacosternum, Natalobatracussion , did not accept this. chus, Petropedetes , Pyxicephalus , Strongy­ Of this group we sampled Batrachylodes lopus , and Tomopterna . Kosuch et al. (2001; vertebralis , Discodeles guppyi , Ceratobata­ figs. 38 , 39), on a relatively small amount of chus guentheri , Platymantis pelewensis , P. evidence, had previously placed Conraua alweberi, and Ingerana baluensis . Thus, we ternatively as either the sister taxon of Limonly lack Palmatorappia from this group 19. nonectes (based on 16S alone) or as the sister Although we obviously cannot test the taxon of Tomopterna 1 Cacosternum (based monophyly of these individual genera (ex­ on combined 12S and 16S). The latter result cept Platymantis ), our taxon sampling is ad­ was suggestive of the more complete results equate to test the monophyly of the inclusive of Van der Meijden et al. (2005). Although group. characters have not been suggested that are CONRAUINAE (1 GENUS, 6 SPECIES): Until clearly synapomorphic, the group is morphothe recent publication by Dubois (2005), this logically compact and monophyly is likely. genus ( Conraua ) had been placed on the ba­ Of the six species we sampled two: Conraua sis of overall similarity in a monotypic tribe, robusta and C. goliath . Conrauini , in Dicroglossinae (Dubois, 1992) . DICROGLOSSINAE (12 GENERA, 152 SPECIES): Conrauini was proposed (Dubois, 1992) for Recounting the taxonomic history of Dicrogthe West African genus Conraua , the diag­ lossinae is difficult inasmuch as it was orignostic characters being the retention of a inally formed on the basis of overall similarfree­living tadpole stage (plesiomorphic), ity, and the content has varied widely, even with a larval keratodont formula of 7–8/6– by the same authors. Only recently has its 11 (see Dubois, 1995, for the definition of concept begun to be massaged by phylogekeratodont formula) and lateral line not re­ netic evidence. Dubois (1987 ‘‘1985’’, 1992) tained into adulthood (plesiomorphic). Van diagnosed Dicroglossinae (in the sense of in­ der Meijden et al. (2005; fig. 36 ), on the ba­ cluding Conrauinae and excluding Paini ) as sis of DNA sequence data, showed that Con­ having the omosternum moderately or raua is not close to Dicroglossinae but the strongly bifurcate at the base and the nasals sister taxon to a taxonomically heteroge­ usually large and in contact with each other

and with the frontoparietal, although none of 19 The status of Liurana Dubois, 1987 , is unclear. Du­ these characters is demonstrably synapobois (1987 ‘‘1985’’) named Liurana as a subgenus of morphic. The most recent taxonomy of Di­ Ingerana ( Ceratobatrachinae ) but, without discussing evidence, Dubois (2005: 4) subsequently considered croglossinae (Dubois, 2005) recognized four Liurana to be a synonym of Taylorana (5 Limnonectes , tribes: Dicroglossini (for Euphlyctis , Fejer­ Dicroglossinae ). Similarly, Dubois (2005), with minimal varya, Hoplobatrachus , Minervarya, Nandiscussion , placed Annandia Dubois, 1992 , in his tribe nophrys, and Sphaerotheca ), Limnonectini Limnonectini, although he had named this taxon as a (for Limnonectes , as well as some taxa consubgenus of Paa , in his Paini . Because these statements are not associated with evidence, they do not merit fur­ sidered by most authors to be synonyms of ther discussion. Limnonectes ), Occidozygini (for Occidozyga 1 alignment was subsequently adjusted manually. Sequence segments considered to be ambiguously aligned were excluded from analysis (508 bp). Substitution model assumed for analysis was GTR G1 I. It was not stated whether gaps were treated as missing data or as evidence.

84 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 297 2006 FROST ET AL.: AMPHIBIAN TREE OF LIFE

85

and Phrynoglossus ), and Paini (for Chapar­ have expected given that the species of ana , Nanorana , and Quasipaa ). Sphaerotheca were long placed in Tomopter­

Dicroglossini was diagnosed by Dubois na ( Pyxicephalinae ). Roelants et al. (2004; (1992; in the sense of including Occidozy­ fig. 35) also placed Nannophrys in Dicrogginae) as retaining a free­living tadpole (ple­ lossinae (by implication) on the basis of siomorphic) and having a lateral line system mtDNA and nuDNA evidence, substantiating that usually is retained into adulthood (pre­ the earlier assessment by Kosuch et al. sumably apomorphic, but not present in Oc­ (2001; figs. 38 ) which was made on less evcidozyga, sensu stricto). As conceived by idence. It was previously assigned to Ranix­ Dubois (1992), the taxon contained Euphlyc­ alini by Dubois (1987 ‘‘1985’’) and to Ditis, Occidozyga , and Phrynoglossus . Fei et al. croglossini by Dubois et al. (2001). Dubois (1991 ‘‘1990’’) and, subsequently, Dubois et et al. (2001: 55) implied on the basis of varal. (2001) on the basis of published and un­ ious published and unpublished mtDNA data published molecular evidence (Marmayou et that Euphlyctis (formerly in his Dicroglossial., 2000— fig. 37 ; Kosuch et al., 2001—figs. ni), Fejervarya , Hoplobatrachus , Minervar­ 38; Delorme et al., 2004— fig. 40 ) placed Oc­ ya, Nannophrys , and Sphaerotheca (formerly cidozyga and Phrynoglossus in the subfamily in his Limnonectini) should be included in a Occidozyginae , and transferred without dis­ reconstituted Dicroglossini. cussion into Dicroglossini Fejervarya and Delorme et al. (2004; fig. 40 ) demonstrat­ Hoplobatrachus (from Limnonectini) and ed—as had Roelants et al. (2004; fig. 35)— Sphaerotheca (from Tomopterninae), and that Lankanectes is phylogenetically distant Nannophrys (from Ranixalinae). from Limnonectes .

Grosjean et al. (2004), building on the ear­ Of these taxa we sampled rather broadly: lier work of Kosuch et al. (2001) suggested Euphlyctis cyanophlyctis ; Fejervarya cancrion the basis of several mtDNA and nuDNA vorus, F. kirtisinghei , F. limnocharis , and F. loci that Euphlyctis is the sister taxon of Ho­ syhadrensis ; Hoplobatrachus occipitalis and plobatrachus with Fejervarya , Sphaerothe­ H. rugulosus ; Limnonectes acanthi , L. grunca , Nannophrys , and Limnonectes forming niens, L. heinrichi , L. kuhlii , L. limborgi (formore distant relations, a result that is consis­ merly Taylorana limborgi ), L. poilani , and L. tent with the tree of Roelants et al. (2004; visayanus ; Nannophrys ceylonensis ; Sphaerfig. 35). otheca breviceps and S. pluvialis . On the ba­

Dubois (1992) also recognized a tribe sis of this sampling we should be able to Limnonectini diagnosed nearly identically evaluate the reality of this taxon and, at least with Conrauini (Conrauinae of this review), to some degree, the monophyly of the condiffering only in the larval keratodont for­ tained genera. mula of 1–5/2–5, which is arguably plesiom­ Occidozygini is a tropical Asian group of orphic. Nominal genera contained in this arguable position. Marmayou et al. (2000; group occur from tropical Africa to tropical fig. 40 ) presented mtDNA evidence that Oc­ Asia with most taxonomic diversity being in cidozyga and Phrynoglossus are not within Asia: Hoplobatrachus , Limnonectes , and Fe­ Dicroglossinae but are outside of a clade jervarya (which was considered a subgenus composed of Rhacophoridae and other memof Limnonectes at the time). In addition Mar­ bers of a paraphyletic Ranidae . Fei et al. mayou et al. (2000; fig. 37 ) and Delorme et (1991 ‘‘1990’’) had already transferred Ocal. (2004; fig. 40 ) suggested on the basis of cidozyga (sensu lato) out of Dicroglossinae mtDNA evidence that Sphaerotheca (former­ and into its own subfamily on the basis of ly in Tomopterninae; Dubois, 1987 ‘‘1985’’) larval characters and this evidence supported and Taylorana (now a synonym of Limno­ the view that Dicroglossinae , as previously nectes; originally considered to be a member conceived, is polyphyletic. Roelants et al.’s of Limnonectini [Dubois, 1987 ‘‘1985’’] but (2004) greater sampling of Asian ranoids subsequently transferred to Ceratobatrachi­ suggested that Ingerana (nominally in Certhe

nae by Dubois, 1992) are in Limnonectini. atobatrachinae) is in this clade and together Sphaerotheca , therefore, is likely not to be form the sister taxon of a reformulated Diclosely related to Tomopterna , as one would croglossinae ( fig. 35), which together are 86 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 297

sister taxon of a clade composed of Mantel­ sampled Phrynoglossus baluensis , P. boreallidae , Rhacophoridae , and Raninae. No Af­ is, P. martensii , and Occidozyga lima .

rican taxa were examined by Marmayou et Paini is a montane Asian tribe diagnosed al. (2000; fig. 37 ), Roelants et al. (2004; fig. among ranids by having an unforked omos­ 35), or Delorme et al. (2004; fig. 40 ), so the ternum (and was therefore formerly included relative position and monophyly of Occidoz­ in Raninae by Dubois, 1987 ‘‘1985’’, 1992) yginae and Dicroglossinae needed to be fur­ and males having black, keratinous ventral ther elucidated. This issue was addressed by spines (presumably a synapomorphy with Van der Meijden et al. (2005; fig. 36 ), who Nanorana ; Jiang et al., 2005: 357). Paini acdid analyze Asian and African taxa simulta­ cording to Dubois (1992) was composed of neously and found Occidozygya lima to sit two genera, each with four subgenera: genus within their Dicroglossinae . Dubois (2005), Chaparana with subgenera Annandia , Cha­

on the strength of the evidence produced by parana , Feirana , and Ombrana ; genus Paa Van der Meijden et al. (2005) , returned Oc­ with subgenera Eripaa , Gynandropaa , Paa , cidozyginae to Dicroglossinae as a tribe. We and Quasipaa . Dubois et al. (2001), citing

2006 FROST ET AL.: AMPHIBIAN TREE OF LIFE

87

88 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 297

composed of ‘‘ Chaparana ’’, several species

of ‘‘ Paa ’’, and Nanorana , characterized by

spines forming two patches on the chest (save

C. quadranus , the type of subgenus Feirana ,

which does not have spines on the chest); and

(2) Group 2, composed of ‘‘ Paa ’’ species as­

sociated previously with the subgenera Quas­

ipaa Dubois, 1992 ( P. robertingeri ), and one

species nominal of the genus Chaparana , sub­

genus Feirana Dubois, 1992 ( Paa yei ). The

second group is characterized by having

spines as a single group, more or less over the

entire venter, but this characteristic is suffi­

ciently variable among subgroups as not to be

diagnostic practically except in the not­ Na­

norana group sense. These authors recom­

mended that the generic name Quasipaa be

applied to Group 2, but for unstated reasons Fig. 39. Neighbor­joining tree of ranoid ex­ hesitated to resolve taxonomically the nonemplars of Kosuch et al. (2001). Underlying data monophyly of Chaparana and Paa in their were 16S data (see figure 38 ) and 12S mtDNA Group 1. Nanorana Günther, 1896 , is the old­ (331 bp). Alignment was done manually using Se­ est available name for their first group. quencher (Applied Biosystems). Gaps treated as Three nominal genera are definitely inmissing data. cluded in Paini : ‘‘ Chaparana ’’ (polyphyletic;

see above); Nanorana ; and ‘‘ Paa ’’ (paraphy­

letic with respect to ‘‘ Chaparana ’’ and Nanunpublished DNA sequence, suggested that orana 20). We sampled Nanorana pleskei, Paini be transferred from Raninae to Dicog­ Quasipaa exilispinosa and Q. verrucospinolossinae . Jiang and Zhou (2001, 2005; fig. sa but did not sample ‘‘ Chaparana ’’ or 41), Jiang et al. (2005; fig. 42), Roelants et ‘‘ Paa ’’ (sensu stricto). al. (2004; fig. 35), and Van der Meijden Jiang et al. (2005) did not mention or ad­ (2005; fig. 36 ) on the basis of published dress three supraspecific taxa usually asso­ DNA sequence evidence, suggested that Di­ ciated with Paini . The first is Eripaa Dubois , croglossinae, with a forked omosternum, is 1992, whose type and only species is Rana paraphyletic with respect to Paini , with an fasciculispina Inger, 1970. Eripaa Dubois , unforked omosternum. For this reason Roe­ 1992, was named and is currently treated as lants et al. (2004) and Jiang et al. (2005) a subgenus of Paa . Although Eripaa exhibits transferred Paini out of Raninae and into Di­ spines on the entire chest and throat, such as croglossinae. Larvae in the group are exo­ in group 2 of Jiang et al. (2005), they are trophic and aquatic (Altig and McDiarmid, uniquely distinct from all other ‘‘ Paa ’’, 1999). ‘‘ Chaparana ’’, and Nanorana species in that Jiang et al. (2005) recently provided a phy­ these spines are clustered in groups of 5–10 logenetic study (fig. 42) of Paini on the basis on circular whitish tubercles. We cannot hazof 12S and 16S rRNA fragments. Unfortu­ ard a guess as to how Eripaa is related to nately, that study appeared too late to guide the rest of Paini . The second is Annandia our choice of terminals, but their results are important in helping us interpret our own re­ 20 Without mentioning content, Dubois (2005) recogsults. They found Paa to be paraphyletic with nized three genera: Chaparana , Nanorana , and Quasirespect to Chaparana and Nanorana ; Cha­ paa. In light of the phylogenetic study by Jiang et al. parana to be polyphyletic with the parts im­ (2005), it is not clear how Chaparana and Nanorana

were intended to be delimited or what the content of

bedded within ‘‘ Paa ’’; and Nanorana to be

2006 FROST ET AL.: AMPHIBIAN TREE OF LIFE

89 Dubois, 1992, whose type and only species originally proposed as a subgenus of Chais Rana delacouri Angel, 1928 . Annandia parana . This species also posseses spinules was originally named as a subgenus of Cha­ only around the anus, prompting Dubois parana Bourret, 1939 , but recently, Dubois (1987, ‘‘1986’’) to consider it evidence of a (2005), without discussion of evidence, treat­ unique reproductive mode, and thus a close ed Annandia as a genus in Limnonectini. relative of Annandia delacouri . Unfortunate­ Perhaps this was done because this species ly, we did not sample any of these three taxa, bears a smooth venter, with spinules only so their status will remain questionable.

clustering around the anus (Dubois, 1987 LANKANECTINAE (1 GENUS, 1 SPECIES): This ‘‘1986’’). Regardless, this is a large taxo­ subfamily was named for Lankanectes cornomic change (from Paini to Limnonectini) rugatus of Sri Lanka by Dubois and Ohler and because no evidence was produced or (2001). Its distinguishing features are (1) discussed to justify this change, we must forked omosternum (plesiomorphy); (2) voconsider the status of this taxon questionable. merine teeth present (presumed plesiomor­ en­

The third is Ombrana Dubois, 1992 , whose phy); (3) median lingual process absent (liketype and only species is Rana sikimensis Jer­ ly plesiomorphy); (4) femoral glands absent don, 1870). Ombrana Dubois, 1992 , was (likely plesiomorphy); (5) toe tips not 90 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 297

larged (arguable polarity); (6) tarsal fold pre­ nectes is far from Limnonectes , where it had sent (likely plesiomorphy at this level); and been placed by Dubois (1992). Roelants et (7) lateral line system present in adults (also al. (2004) placed it as the sister taxon of in Phrynoglossus and Euphlyctis , but pre­ Nyctibatrachinae , and Delorme et al. (2004) sumably apomorphic). Roelants et al. (2004; placed it as the sister taxon of Nyctibatra­

fig. 35) and Delorme et al. (2004; fig. 40 ) chinae 1 Raninae. We sampled the sole spesubsequently suggested on the basis of cies, Lankanectes corrugatus .

2006 FROST ET AL.: AMPHIBIAN TREE OF LIFE

91

NYCTIBATRACHINAE (1 GENUS, 12 SPECIES):

Nyctibatrachinae contains the Indian taxon

Nyctibatrachus and is characterized by hav­

ing a forked omosternum (likely plesiom­

orphic), vomerine teeth present, digital discs

present, femoral glands present (shared with

Ranixalinae and some Dicroglossinae ) and

an aquatic tadpole with a keratodont formula

of 0/0 (likely apomorphic; Dubois et al.,

2001). Of this taxon we sampled Nyctibatra­

chus cf. aliciae and N. major .

PETROPEDETINAE (2 GENERA, 10 SPECIES);

PHRYNOBATRACHINAE (4 GENERA, 72 SPECIES)

AND PYXICEPHALINAE (13 GENERA, 57 SPE­

CIES): Until recently, members of Petrope­

detinae and Phrynobatrachinae, as well as

several genera now assigned to Pyxicephali­

nae (e.g., Anhydrophryne , Arthroleptella ,

Cacosternum , Microbatrachella , Nataloba­

trachus, Nothophryne , and Poyntonia ) were

considered members of ‘‘Petropedetidae’’

(sensu lato), aggregated on the basis of over­ Fig. 42. Consensus of four parsimony trees of all similarity, with no evidence for its mono­ Paini by Jiang et al. (2005), based on 796 bp (of phyly ever suggested. Noble (1931) recogwhich 174 were parsimony informative) of the nized his Petropedetinae ( Arthroleptides and 12S and 16S rRNA framents of mtDNA. Se­ Petropedetes ), as united by the possession of quences were aligned using ClustalW (Thompson dermal scutes on the upper surface of each et al., 1994), cost functions not disclosed, with digit and otherwise corresponding osteologsubsequent manual modifications. Gaps and am­ ically and morphologically with Raninae. biguously aligned sequences were excluded from Noble (1931) also recognized Cacosterninae analysis (ci 5 0.584, ri 5 0.571). The trees were

for Cacosternum and Anhydrophryne , united rooted on Hoplobatrachus chinensis and Fejervarya fujianensis . A conclusion of Jiang et al. by lacking a clavicle and having palatal ridg­ (2005) is that their Group 2 was recognized as es. He related the cacosternines to brevicip­ Quasipaa . itines, and the remainder of the genera then

named he allocated to Raninae.

Laurent (1941 ‘‘1940’’) addressed the conical Asian Micrixalus (11 species) is the sole fusion between Arthroleptis and Phrynobamember of this taxon, diagnosed by Dubois trachus and transferred Petropedetes , Anhy­ (2001) as differing from Dicroglossinae in drophryne, Phrynobatrachus (including Nalacking a forked omosternum (possibly apo­ talobatrachus), Dimorphognathus , and Armorphic), lacking vomerine teeth, having throleptella into his Phrynobatrachinae. digital discs (present in some limnonectines Laurent (1941) subsequently provided an an­ and otherwise widespread in Ranoidea ) and atomical characterization of the group. having a larval keratodont formula in its Laurent (1951) transferred Cacosterninae aquatic tadpoles of 1/0 (likely apomorphic) into Ranidae and moved Microbatrachella (Dubois et al., 2001) . On the basis of mtDNA into Cacosterninae . Poynton (1964a) sug­ and nuDNA evidence, Roelants et al. (2004; gested that Phrynobatrachus is deeply parafig. 35) considered Micrixalinae to be the sis­ phyletic with respect to Cacosterninae and ter taxon of Ranixalinae. We were able to therefore considered Laurent’s Phrynobatrasample Micrixalus fuscus and M. kottigehar­ chinae (5 Petropedetinae ) and Cacosterninae g., ensis. Although this provides only a minimal to be synonyms. Subsequent authors (e. test of the monophyly of Micrixalus , it al­ Dubois, 1981; Frost, 1985) uncritically follows us to place the taxon phylogenetically. lowed this unsupported suggestion, although

92 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 297

there have been significant instances of taxon was stated to be Anhydrophryne, Arworkers continuing to recognize Cacosterni­ throleptella, Cacosternum, Microbatrachelnae and Petropedetinae as distinct (e.g., la, Nothophryne , Poyntonia (from Petrope­ Liem, 1970; J.D. Lynch , 1973). detidae), and, possibly Strongylopus and To­

Another morphologically compact African mopterna (from Ranidae ). group was Pyxicephalinae (Dubois, 1992) , Van der Meijden et al. (2005; fig. 36 ) sugcomposed of Pyxicephalus (2 species) and gested Phrynobatrachus to be the sister tax­ Aubria (2 species). The taxon was diagnosed on of Ptychadena . On this basis Dubois by at least four synapomorphies (Clarke, (2005) recognized a ranid subfamily Phry­ 1981): (1) cranial exostosis; (2) occipital ca­ nobatrachinae, containing Phrynobatrachus , nal present in the frontoparietal; (3) zygo­ but also allocated to this subfamily, without matic ramus being much shorter than otic ra­ discussion, Dimorphognathus, Ericabatramus ; and (4) sternal style a long bony ele­ chus, and Phrynodon . Petropedetes and Conment tapering markedly from anterior to pos­ raua formed successively more distant outterior. Dubois’ (1992) reasoning for groups of the southern African clade of Van excluding this taxon from Dicroglossinae is der Meijden et al. (2005), so Dubois (2005) not clear, but presumably had to do with the removed Conrauini ( Conraua ) from Dicogdistinctive appearances of Pyxicephalus and lossinae and placed it in its own subfamily, Aubria . Conrauinae, and recognized Petropedetinae

Dubois (1992) also recognized a subfam­ for Petropedetes , as well as the presumably ily Tomopterninae, for Tomopterna (sensu closely allied Arthroleptides . The southern lato, at the time including Sphaerotheca , now African clade of Van der Meijden et al. in Dicroglossinae , Limnonectini). The diag­ (2005; fig. 36 ) was composed of Cacosternosis provided by Clarke (1981) presumably num (formerly of Petropedetidae ), Afrana applies inasmuch as he examined only Afri­ and Strongylopus (formerly of Raninae), Nacan species ( Tomopterna , sensu stricto), even talobatrachus (formerly of Petropedetidae ), though the optimization of these characters Tomopterna (Tomopterninae) , and Pyxiceon his cladogram may well be contingent on phalus ( Pyxicephalinae ), a group that Dubois being compared only with other African ra­ (2005) allocated to an enlarged Pyxicephalnids: (1) zygomatic ramus much shorter than inae. Aubria was asserted by Dubois (2005) otic ramus; (2) outline of anterior end of cul­ to be in this group because it was grouped triform process pointed, with lateral borders by morphological evidence with Pyxicephaltapering to a point; (3) distal end of the an­ us. Amietia he transferred into the group terior pterygoid ramus overlapping the dorsal without discussion, but presumably because surface of the posterior lateral border of the they appeared to him to be related to Stronpalatine; (4) no overlap of the anterior border gylopus and Afrana . He transferred Arthroof the parasphenoid ala by the medial ramus leptella, Microbatrachella , Nothophryne , and of the pterygoid in the anterior–posterior Poyntonia into Pyxicephalinae , presumably plane; (5) sternal style short, tapering poste­ because he thought that they were more likeriorly; (6) dorsal protuberance of the ilium ly to be here than close to either Petropedenot or only slightly differentiated from the tinae or Phrynobatrachinae. spikelike dorsal prominence; and (7) terminal Of Dubois’ (2005) Petropedetinae (which phalanges of the fingers and toes reduced, presumably is diagnosed as by Noble, 1931) almost conelike. we were able to sample both genera: Arthro­

In 2003 this untidy, but familiar arrange­ leptides sp. and Petropedetes cameronensis , ment began to unravel. Dubois (2003), re­ P. newtoni , P. palmipes , and P. parkeri . moved Cacosterninae from ‘‘Petropedetidae’’ Of the newly constituted Phrynobatrachiwithout discussion, apparently anticipating nae, we were also able to sample species evidence to be published elsewhere, although from three of four genera: Dimorphognathus Kosuch et al. (2001; fig. 38 ) had suggested africanus , Phrynobatrachus auritus , P. cal­

earlier that Cacosternum was more closely caratus, P. dendrobates , P. dispar , P. marelated to Tomopterna and Strongylopus than babiensis, P. natalensis , and Phrynodon sanit was to Petropedetes . The content of this dersoni. We did not sample Ericabatrachus ,

2006 FROST ET AL.: AMPHIBIAN TREE OF LIFE

93

an unfortunate omission, inasmuch as we are The three nominal genera in the taxon are unaware of the evidence for Dubois’ (2005) Ptychadena (47 species), Hildebrandtia (3 association of Ericabatrachus with Phryno­ species), and Lanzarana (1 species) of which batrachinae, other than the statement that it we sampled only Ptychadena anchietae , P. is ‘‘ Phrynobatrachus ­like’’ (Largen, 1991). cooperi , and P. mascareniensis . Because we Phrynobatrachus , at least for the species did not sample Hildebrandtia and Lanzarwhich it is known, have exotrophic larvae. ana , we did not adequately test the mono­ Larvae are unknown in Dimorphognathus phyly of this group. Nevertheless, assuming and Ericabatrachus , and Phrynodon is en­ the group to be monophyletic, our three spedotrophic (Amiet, 1981; Altig and Mc­ cies of Ptychadena allow us to test the place­ Diarmid, 1999). ment of Ptychadeninae within Ranoidea . For Of the reformulated Pyxicephalinae we his analysis Clarke (1981) assumed that Ptywere able to sample Aubria ( Aubria subsi­ chadeninae is imbedded within other African gillata [2 samples 21]) and Pyxicephalus (Py­ ranids, although a lack of comparison with xicephalus edulis ) as well as several of the Asian members of the group makes this astaxa recently transferred into this taxon in­ sumption questionable. Van der Meijden et cluding Anhydrophryne rattrayi , Arthrolep­ al. (2005; fig. 36 ) suggested that Ptychadena tella bicolor , Cacosternum platys , and Na­ is the sister taxon of Phrynobatrachus among talobatrachus bonebergi . We also sampled his exemplars, thereby implying that Ptymembers of Afrana ( A. angolensis and A. chadeninae is the sister taxon of Phrynobafuscigula), Tomopterna ( T. delalandii ), trachinae.

Strongylopus ( S. grayii ), and Amietia (A. ver­ ‘‘ RANINAE ’’ (CA. 8 GENERA, 309 SPECIES): tebralis), but for reasons having to do with ‘‘Raninae’’ is a catch­all largely Holarctic the evidentiary basis and history of taxono­ and tropical Asian taxon united because the my in Raninae, considerable discussion of members do not fit into the remaining subthese genera is presented there. We did not families and have unforked omosterna. Until sample Microbatrachella , Nothophryne , or recently, ‘‘Raninae’’ included two tribes: Pai­ Poyntonia . Pyxicephalines have exotrophic ni and Ranini (Dubois, 1992). However, Pailarvae, with the exception of Anhydrophryne ni and Nanorana of Ranini were transferred and Arthroleptella , which are endotrophic; to Dicroglossinae on the basis of mtDNA and unknown in Nothophryne (Hewitt, 1919; nuDNA evidence (Roelants et al., 2004—fig. Procter, 1925; DeVilliers, 1929; Altig and 35; Jiang et al., 2005—fig. 42), so Raninae, McDiarmid, 1999). This selection should al­ as we use it, is coextensive with Ranini of low us to test the phylogenetic results of Van Dubois (1992), itself dubiously monophylet­ der Meijden et al. (2005). ic 22.

PTYCHADENINAE (3 GENERA, 51 SPECIES): ‘‘Raninae’’ is distributed on the planet co­ Ptychadeninae is a morphologically compact extensively with the family and is united by group of sub­Saharan ranids diagnosed the lack of putative apomorphies, either in (Clarke, 1981; Dubois, 1987 ‘‘1985’’, 1992) the adult or in the larvae. There does not by having: (1) an otic plate of the squamosal appear to be any reason to suggest that this covering the crista parotica in dorsal view nominal taxon is monophyletic.

and extending mesially to overlap the otoc­ The starting point of any discussion of cipital; (2) palatines absent; (3) clavicles re­ Ranini must be Dubois (1992), who provided duced; (4) sternal style a short compact ele­ an extensive, and controversial, taxonomy. ment tapering anteriorly to posteriorly; (5) Because the distinction between ranks (seceighth presacral vertebra fused with sacral

vertebra; and (6) the dorsal protuberance of 22 Van der Meijden et al. (2005; fig. 36 ), provided evilium smooth­surfaced and not prominent. idence from DNA sequences that suggests strongly that ‘‘Raninae’’ is polyphyletic, with at least Afrana and Strongylopus in a southern African clade (along with 21 We included two specimens of Aubria subsigillata Pyxicephalus , Tomopterna , Natalobatrachus , and Cain as separate terminals in the analysis because the identity costernum), far from other ranines, and in Pyxicephaliof one of the specimens was not determined conclusive­ nae of Dubois (2005). We therefore treat ‘‘Raninae’’ ly until after the analyses were complete. the following discussion as dubiously monophyletic.

94 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 297

tion, subsection, genus, and subgenus) in Du­ pole morphology (presence of a raised, bois’ system appears to rest primarily on sub­ sharply defined abdominal sucker). Like othjective perceptions of similarity and differ­ er cascade­dwelling taxa, larvae of Amolops ence, the evidentiary basis of this taxonomy (sensu lato) all share high numbers of kerais unclear, even though we accepted his sys­ todont rows. Subsequently, Yang (1991b) tem as a set of bold phylogenetic hypotheses. recognized two other genera from within Nevertheless, most of these taxa are imper­ Amolops : Meristogenys and Huia . Amolops fectly or incompletely diagnosed and to lay (sensu stricto) has one possible synapomorthe foundation for our results and concomi­ phy (short first metacarpal, also found in tant taxonomic remedies, we discuss this tax­ Huia ), and three synapomorphies joining onomy in greater depth than we do most of Huia and Meristogenys to the exclusion of the remainder of current amphibian taxono­ Amolops (lateral glands present in larvae; my. Suffice it to say that we think that we four or more uninterrupted lower labial kersampled ‘‘ Rana ’’ diversity sufficiently to atodont rows; and longer legs). provide at least a rudimentary phylogenetic Subsequently, Dubois (1992) treated Merunderstanding of the taxon as a starting point istogenys and Huia as subgenera of Amolops , for future, more densely sampled studies. and added a fourth subgenus, Amo (including

Within his Ranini, Dubois (1992) recog­ only Amolops larutensis ). Amo was diagnized six genera: Amolops , Batrachylodes , nosed (Boulenger, 1918) as having a digital Nanorana , Micrixalus , Rana , and Staurois disc structure similar to species of Staurois (table 4). Of these, two continue to be placed (i.e., having a transverse groove or ridge on in this taxon ( Amolops and Rana [sensu the posteroventral side of the disc continuous lato]) (Dubois, 2005). Staurois , Nanorana with a circummarginal groove to define a and Micrixalus have subsequently been hemisphere; Boulenger, 1918) and as having transferred out of Ranini, Staurois to a new axillary glands (after Yang, 1991b) that are tribe, Stauroini (Dubois, 2005), Nanorana to otherwise unknown in Amolops . Dicroglossidae (Roelants et al., 2004; fig. Although Dubois (1992) considered Amo­ 35), and Micrixalus to a distant Micrixalinae lops (sensu stricto), Amo , Huia , and Meris­ (Dubois et al., 2001). Batrachylodes was pro­ togenys to be subgeneric parts of a monovisionally transferred, without substantial phyletic genus Amolops , other authors (e.g., discussion, by Dubois (2005) to Ceratobatra­ Yang, 1991b) considered at least Amolops , chinae. Huia , and Meristogenys as genera. For con­

Within both Amolops and Rana, Dubois sistency we treat as genera Amo , Amolops , recognized several subgenera, that other au­ Huia , and Meristogenys . Our samples were thors (e.g., Yang, 1991b) considered to be Amolops ( A. chapaensis , A. hongkongensis ), genera, as we do, although we arrange the Huia ( H. nasica ), and Meristogenys ( M. ordiscussions by Dubois’ genera and subgen­ phocnemis). We were unable to sample Amo era. Dubois (2003) arranged Raninae into larutensis . two tribes (Amolopini for the taxa with cas­ Staurois : The definition of Staurois (digicade­adapted tadpoles, i.e., Amo , Amolops , tal discs broader than long; T­shaped termi­ Huia , Meristogenys , Chalcorana , Eburana , nal phalanges in which the horizontal part of Odorrana ) and Ranini (for everything else). the T is longer than the longitudinal part; This system represents typical nonevolution­ outer metatarsals separated to base but joined ary A and not­A groupings, although Amo­ by webbing; small nasals separated from lopini in this form is testable. Dubois (2005) each other and frontoparietal; omosternal subsequently did not embrace Amolopini, style not forked [Boulenger, 1918]) has also because it was too poorly understood, but he been used to define Hylarana (Boulenger, did erect Stauroini for Staurois , because Roe­ 1920; see below). Although some larval lants et al. (2004) placed Staurois as the pu­ characters are shared among species of Stautative sister taxon of other ranines. rois (deep, cup­like oral disc in the tadpole,

Amolops , Amo , Huia , and Meristogenys : no glands or abdominal disc in tadpole; In­ Amolops has been recognized in some form ger, 1966), the diagnostic value of these charsince Inger (1966) noted the distinctive tad­ acters is unknown due to the large number

2006 FROST ET AL.: AMPHIBIAN TREE OF LIFE

95

of ranid species whose adults are morpho­ (Inger, 1996) that are not based on any comlogically similar to those of Staurois , but prehensive comparative study of either interwhose larvae remain undescribed. Our single nal or external morphology. For instance, larexemplar of Staurois , S. tuberilinguis , is not vae may have dorsal dermal glands, lateral sufficient to test the monophyly of the genus. dermal glands, or ventral dermal glands in Although no one has suggested that Staurois various combinations (e.g., Yang, 1991b). is polyphyletic, or that it is paraphyletic with These characters have become larval dermal respect to any other group, both of these re­ glands present or absent in Dubois’ (1992) main untested possibilities. Roelants et al. diagnoses, thereby conflating the positional (2004; fig. 35) provided evidence that Stau­ homology of these features. Although we adrois is the sister taxon of remaining ranines. dress deficiencies here and in the Taxonomy Rana (sensu Dubois, 1992) 23: Rana of Du­ section, for other critiques see Emerson and bois (1992) is diagnostically coextensive Berrigan (1993), Matsui (1994), Matsui et al. with his Ranini (our ‘‘Raninae’’), and no fea­ (1995), Inger (1996), Bain et al. (2003), and tures provided in his paper exclude ‘‘ Rana ’’ Matsui et al. (2005).

from being paraphyletic with respect to Stau­ As noted earlier, several, if not most taxa rois, Amolops (sensu Dubois, 1992) , or Ba­ recognized by Dubois within his ‘‘ Rana ’’ are trachylodes. So, as we discuss the internal effectively undiagnosed in a utilitarian sense taxonomy of ‘‘ Rana ’’ as provided by Dubois, (i.e., they are diagnosed sufficiently only to readers should bear in mind that Amolops make the names available under the Inter­ (sensu lato), Batrachylodes , and Staurois , as national Code; ICZN, 1999). In addition, discussed by Dubois (1992), must be regard­ several are demonstrably nonmonophyletic ed as potential members of all infrageneric (Matsui, 1994; Matsui et al., 1995; Inger, taxa that do not have characters that specif­ 1996; Tanaka­Ueno et al., 1998a; Emerson et ically exclude them. (And, at least with re­ al., 2000a; Marmayou et al., 2000; Vences et spect to Dubois’, 1992, Rana subgenera, al., 2000a; B.J. Evans et al., 2003; Roelants Strongylopus and Afrana, DNA sequence et al., 2004; Jiang and Zhou, 2005). Unlike data have been published that suggest that the superficially similar situation in Eleuththey have little relationship with other rani­ erodatylus (sensu lato) where it is straightnes [Van der Meijden et al., 2005; fig. 36 ].) forward to get specific information on indi­ With respect to ‘‘ Rana ’’ specifically, Dubois vidual species and where the nominal sub­ (1992) provided a system of sections, sub­ genera and most related genera, even if they sections, and subgenera that has posed seri­ do not rise to the level of synapomorphy ous challenges for us: Rather than a syna­ schemes, have been diagnosed largely compomorphy scheme, or even a system of care­ paratively, the subgeneric (and generic, in fully­evaluated characteristics, the various part) diagnoses of ranids are not comparable, taxa appear to represent postfacto character and the purported differentiating characters justifications of decidedly nonphylogenetic frequently do not bear up to specimen ex­ and subjectively arrived­at groups. We found amination (e.g., Tschudi, 1838; Boulenger, Dubois’ (1987 ‘‘1985’’, 1992) arrangement 1920; Yang, 1991b; Fei et al., 1991 ‘‘1990’’; to be inconsistent with the preponderance of Dubois, 1992).

evidence in certain instances (see the discus­ Historically, taxonomists approached sion of inclusion of Aquarana in his section Rana (sensu lato) as being composed of two Pelophylax , below) and the underlying di­ very poorly defined similarity groupings: (1) agnostic basis of the system to contain over­ those that have expanded toe tips (likely plely­generalized statements from the literature siomorphic) that at one time or another have been covered by the name Hylarana ; and (2) 23 Although Afrana , Amietia , and Strongylopus (now those that lack expanded toe tips, and that in Pyxicephalinae ), Batrachylodes (now in Ceratoba­ have more­or­less always been associated trachidae), Micrixalus (now in Micrixalinae), and Na­ with the generic name Rana . Most authors norana (now in Dicroglossinae ) have been transferred since Boulenger (1920) recognized the lack out of Raninae, we address them as part of the general discussion of ranine systematics prior to 2004. (See table of definitive ‘‘breaks’’ between the two 4) groups, and Dubois was the first to attempt

96 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 297 2006 FROST ET AL.: AMPHIBIAN TREE OF LIFE

97

to summarize the relevant taxonomic literature and to divide Rana (sensu lato) into enough groups to allow some illumination of the problem. Our issue with his system is that it is impossible to tell from the relevant publication (Dubois, 1992) which species have actually been evaluated for characters and which have merely been aggregated on the basis of overall similarity or erected on the basis of specially­favored characters.

Dubois’ primary division of Rana was into eight sections of arguable phylogenetic propinquity to each other or to other ranine genera (see table 4). We discuss these with reference to his diagnoses and other literature relevant to their recognition:

(1) Section Amerana . Dubois (1992) erect­ ed his subgenera Amerana and Aurorana for parts of the Rana boylii group of Zweifel (1955), which he placed in their own section, Amerana . Most previous work (e.g., Case, 1978; Farris et al., 1979; Post and Uzzell, 1981; Farris et al., 1982b; Uzzell and Post, 1986) had placed these frogs from western North American close to, or within, the Eurasian Rana temporaria group. Nevertheless, section Amerana was recognized by Dubois

Fig. 43. Restriction­site tree of exemplars of (1992) on the basis of a combination of char­

Holarctic Rana of Hillis and Davis (1986). Unacters, none unique but corresponding to the derlying data were restriction sites of the nuclear Rana boylii group identified by ribosomal rDNA gene; presence was considered to be evidata by Hillis and Davis (1986; fig. 43). This dence of relationship, absence was not. The tree group had been suggested by Hillis and Da­ was rooted on Pyxicephalus and Pelophylax (as vis (1986) to be in a polytomy with what Rana ridibunda ). The original figure treated all Dubois regarded as his section Rana (R. tem­ species, save Pyxicephalus , as members of Rana . poraria and R. sylvatica were the exemplar We have noted on the right the nominal subgenera species in their analysis), a group composed of Dubois (1992; which we have treated as genof a part of Dubois’ section Pelophyax era), to clarify discussion. ( Aquarana ), and his sections Lithobates and Pantherana . Moreover, Hillis and Davis’ monophyletic group with Rana temporaria , (1986; fig. 43) results suggested that neither to the exclusion of all other North American of the groups subsequently identified by Du­ Rana , inasmuch as this was an assumption of bois (1992) as the subgenera Aurorana and their analysis, based on earlier work (e.g., Amerana are monophyletic. Subsequent Case, 1978). work (Hillis and Wilcox, 2005; fig. 44 ) has Dubois (1992) provided no unique morprovided substantial amounts of evidence in phological features to diagnose section support of the nominal subgenus Aurorana Amerana , and because of his use of presentbeing polyphyletic, and the subgenus Amer­ or­absent as a characteristic, the characters ana being paraphyletic. Hillis and Wilcox provided in his table 1 fail to rigorously dis­ (2005) used the section Amerana 1 Rana tinguish section Amerana from sections Hytemporaria to root the remainder of their larana, Lithobates , Pelophylax , Rana , or the

, tree, so their overall tree cannot be taken as Strongylopus (now in Pyxicephalinae on additional evidence of evolutionary propin­ basis of DNA sequence evidence—Dubois quity of the section Amerana being in a 2005; Van der Meijden et al., 2005). Within

98 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 297 2006 FROST ET AL.: AMPHIBIAN TREE OF LIFE

99

Amerana, Dubois recognized two subgenera, their taxonomy was presented for only the Amerana and Aurorana , differing in the ex­ Chinese fauna, so the wider implication of pansion of toe tips (mildly expanded in this action is not known. Of this section we Amerana ; not expanded in Aurorana ), rows sampled no member of the subgenus Babina , of larval keratodonts (4–7/ 4–6 in Amerana ; although we did sample Nidirana adenopleu­ 2–3/ 3–4 in Aurorana ) karyotype (derived in ra and N. chapaensis . Babina and Nidirana Amerana ; primitive in Aurorana ). This sub­ have also been associated with ‘‘ Hylarana ’’ generic distinction is not phylogenetically (see below), so Dubois’ (1992) reason for consistent with the results of Hillis and Davis recognizing this as a section distinct from (1986; fig. 43), who presented evidence sug­ section Hylarana is unclear. gesting that Dubois’ Aurorana is paraphylet­ (4) Section Lithobates . This section is not ic with respect to his Amerana (making one rigorously diagnosable by the features prewonder what the purpose was in naming two sented by Dubois’ (1992: his table 1) from subgenera). Macey et al. (2001) subsequently sections Amerana , Hylarana , Rana , or Stronprovided additional molecular evidence for gylopus. However, Lithobates is consistent paraphyly of Aurorana with respect to Amer­ with the phylogenetic tree of American Rana ana . Examples of this section in our analysis provided by Hillis and Davis (1986; fig. 43), are Amerana muscosa and Aurorana aurora presumably the source of the concept of this (see table 4). section. Hillis and Davis placed this taxon, (2) Section Amietia (including a single on the basis of DNA substitutions, as the sissubgenus, Amietia , for two species in the Le­ ter taxon of part of Dubois’ section Pelophysotho Highlands of southern Africa). The lax, the subgenus Pantherana . Within section sole synapomorphy of Amietia is the umbra­ Lithobates, Dubois recognized four subgenculum over the eye in the larva. The diag­

era: Lithobates ( Rana palmipes group), Siernosis of section Amietia is otherwise phylo­

rana ( Rana maculata group), Trypheropsis genetically indistinguishable on the basis of

( Rana warszewitschii group), and Zweifelia the table of characters provided by Dubois

( Rana tarahumarae group). All of them are (1992), from Amerana , Hylarana , Lithoba­

consistent with the tree provided by Hillis tes, Rana , or Strongylopus . We sampled

and Davis (1986). Dubois (1992) offered the Amietia vertebralis . Amietia was transferred

following morphological characters which into Pyxicephalinae by Dubois (2005) on the

may be synapomorphies: Lithobates differs apparent but undiscussed assumption that it

from other members of the section by having is closely related to Strongylopus , which was

tympanum diameter larger or equal to the diplaced by Van der Meijden et al. (2005) in that group on the basis of DNA sequence ev­ ameter of the eye; Sierrana without diagnosidence. tic characters that differentiate it from the (3) Section Babina (for the Rana holsti section diagnosis; Trypheropsis by having an and Rana adenopleura groups). The unique outer metatarsal tubercle (unusual in Amerisynapomorphy for this group is a large ‘‘su­ can ranids); and Zweifelia with sacrum not prabrachial’’ gland (sensu Dubois, 1992) on fused with presacral vertebrae. Hillis and the sides of reproductive males (which can Wilcox (2005; fig. 44 ) presented evidence be difficult to assess in nonreproductive an­ that suggests that section Lithobates of Duimals). The diagnosis of section Babina does bois (1992) is paraphyletic, with part of Dunot otherwise allow it to be practically sep­ bois’ subgenera Sierrana ( R. maculata ), and arated from the sections Amerana , Hylarana , all of his subgenera Trypheropsis , and Lith­ Lithobates , Pelophylax , Rana , or Strongylo­ obates falling within one monophyletic pus. Within section Babina, Dubois recog­ group, but Zweifelia (the Rana tarahumarae nized two subgenera, Babina (with a large group) and another part of Sierrana ( R. sierfingerlike prepollical spine, an apomorphy) ramadrensis) forming the sister taxon of Du­ and Nidirana (members of the Babina sec­ bois’ subgenus Pantherana , the Rana pipiens

tion lacking the apomorphy of the subgenus group of Hillis and Wilcox (2001). Babina ). Fei et al. (2005) considered Nidi­ Our exemplars of this section are Lithobrana to be a subgenus of their Hylarana , but ates palmipes , Sierrana maculata , and Try­

100 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY NO. 297

Figures

Fig. 14. Narrative tree of relevant anuran taxa by Ford and Cannatella (1993). A branch subtending Hylidae 1 Pseudidae in the original figure is collapsed per errata distributed with reprint. An asterisk was used by these authors to denote a metataxon, and quotation marks to denote nonmonophyly.

Fig. 38. Neighbor­joining tree of ranoid exemplars of Kosuch et al. (2001), which ‘‘agreed well’’ with the consensus of four equally parsimonious trees (ci 5 0.51). Underlying data were 572 bp of aligned 16S mtDNA sequences of which 221 are parsimony­informative. Alignment was done manually using Sequencher (Applied Biosystems). Indels were treated as missing data. Taxon assignments on the right reflect the taxonomy as it existed at the time.

Fig. 36. Maximum likelihood tree of exemplars of Ranoidea, with a focus on African taxa, by Van der Meijden et al. (2005), based on mt DNA (12S and 16S rRNA) and nu DNA (RAG­1, RAG­2, rhodopsin), for 2,995 bp of sequence. Alignment was made using ClustalW (Thompson et al., 1994), with costs not disclosed and gaps and highly variable sites excluded from analysis. The model assumed for maximum­likelihood analysis was TrN 1 I 1 G. The tree was rooted on an hierarchical outgroups (not shown in original) composed of Latimeria, Homo, Gallus, Lyciasalamandra, Alytes (2 spp.), Agalychnis, and Litoria. The ‘‘southern African clade’’ represents Pyxicephalinae as subsequently redelimited by Dubois (2005).

Fig. 37. Consensus of two equally parsimonious trees from Marmayou et al. (2000) of exemplars of Ranidae and Rhacophoridae (Ranidae: Rhacophorinae in their usage) based on 305 bp (151 informative sites) of 12S mtDNA, aligned using the program MUST (Philippe, 1993) and subsequently manually modified with reference to secondary structure models. Cost functions for alignment were not stated, nor whether gaps were treated as missing data or as evidence (ci 5 0.382, ri 5 0.429). Tree rooted on Eleutherodactylus cuneatus (5 Euhyas cuneata).

Fig. 40. Maximum­likelihood tree of ranoids of Delorme et al. (2004), based on sequences from 12S and 16S rRNA for a total of 1198 bp. Alignment was made using the program Se­Al (Rambaut, 1995; cost functions not provided) and by comparison with models of secondary structure. Gaps were treated as missing data. The maximum­likelihood nucleotide substitution model accepted was TrN 1 I 1 G.

Fig. 44. Maximum­likelihood tree of Holarctic Rana of Hillis and Wilcox (2005). The underlying data are ca. 2kb of mtDNA of the 12S–16S region (spanning the tRNAVal gene). Sequence alignment was done initially using Clustal W (Thompson et al., 1994), costs not disclosed, and manually adjusted, guided by assumed secondary structure, ambiguously aligned sequences discarded. It was not stated whether gaps were treated as data, but we presume not. Substitution model GTR 1 G1 PINVAR was assumed for the maximum­likelihood analysis. On the basis of previous research, the root was assumed