BREVIORA

Museum of Comparative Lioology

US ISSN 0006-9698

CAMBRIDGE, MAss.

24 JuLy 2008

NuMBER 513

DISTINGUISHING FEATURES OF THE SUB-SAHARAN FROG GENERA ARTHROLEPTIS AND PHRYNOBATRACHUS: A SHORT GUIDE FOR FIELD AND MUSEUM RESEARCHERS

BREDA M. ZimMKus!* AND Davip C. BLACKBURN!

ABSTRACT.

Typically small body size and similar external anatomy have made it difficult for researchers to

distinguish between species of two distantly related, but often sympatric, African ranoid frog genera: Arthroleptis (Arthroleptidae) and Phrynobatrachus (Phrynobatrachidae). We define a suite of external morphological characters, extending beyond the traditionally used secondary sexual characteristics, that can be used to definitively distinguish between adult Arthroleptis and Phrynobatrachus. Photographs comparing absence, presence, and variation of morphological characters are included for clarification. Significant differences between body proportions (head width/snout-vent length) are also observed between these genera. Although smaller species might be more difficult to identify, larger species (> 30 mm snout—vent length) of Arthroleptis can be distinguished from Phrynobatrachus because of their relatively wider heads. We intend this to serve as a heuristic guide for both field- and museum-based

researchers.

KEY WORDs:

INTRODUCTION

Nearly every herpetologist is familiar with the overwhelming diversity of “little brown frogs.” Many leaf litter anurans are small, generally drab in color, and morphologically similar, and this has hampered both system-

‘Museum of Comparative Zoology, and Department of

Organismic and Evolutionary Biology, Harvard Univer- sity, Cambridge, Massachusetts 02138, U.S.A.; *Corresponding author. E-mail: bzimkus@oeb.harvard. edu.

Africa; Amphibia; Arthroleptidae; identification; Ranoidea; Phrynobatrachidae

atic research and identification in the field and in museum collections. Within sub- Saharan Africa, two such genera, Arthrolep- tis Smith 1849 and Phrynobatrachus Gunther 1862, have been confused for more than a century (i.e., Boulenger, 1882). On the basis of studies of skeletal morphology (e.g.. Laurent, 1940, 194la, b; Scott, 2005) and the results of molecular phylogenetic analy- ses (e.g., Bossuyt et al., 2006; Frost et al, 2006; Roelants et a/., 2007), it is now very clear that these genera, although externally similar, are only distantly related. In our

2 BREVIORA

experience, many researchers are unaware of the numerous external morphological char- acters that can be used to distinguish between each genus. Arthroleptis and Phrynobatrachus continue to be confused in the field and in museum collections, which can then confound further research such as molecular phylogenetic analyses (see Discussion for an example). Thus the external morphological differences between these two genera warrant a thorough treat- ment to improve the efficacy of identifica- tion.

Squeakers (4rthroleptis sensu Frost, 2007) are generally small (16-54 mm snout—vent length), terrestrial, leaf litter anurans, and all of the approximately 37 species are believed to have lost the free-living, feeding tadpole stage (i.e., direct development). The smallest Arthroleptis species were previously placed in Schoutedenella (Laurent, 1954); Frost er al. (2006) synonymized Schoutedenella with Ar- throleptis on the basis of the results of their molecular phylogenetic analysis. An analysis of the relationships within Arthroleptis and its putative sister taxon Cardioglossa 1s currently in preparation (Blackburn, unpub- lished).

Puddle frogs (Phrynobatrachus) comprise a lineage of approximately 75 species found in diverse terrestrial habitats across sub-Sa- haran Africa (Frost, 2007). Phrynobatrachus has been a long-standing source of confusion to systematists because of extensive geo- graphic and intrapopulation variation ac- companied by only slight morphological differences among species (Largen, 2001; Rodel, 2000; Stewart, 1974). Molecular data from mitochondrial and nuclear markers indicate that both Dimorphognathus and Phrynodon might be embedded within Phry- nobatrachus, thus rendering the genus para- phyletic (Frost et al., 2006; Scott, 2005). Both Scott (2005) and Frost et al. (2006) argued for synonymizing these genera with

No. 513

Phrynobatrachus. Whereas Dimorphognathus africanus is embedded deeply within Phryno- batrachus, supporting the synonymization of this monotypic genus, the placement of Phrynodon sandersoni within Phrynobatra- chus is more controversial because of its basal position within the lineage; the synon- ymy of Phrynodon with Phrynobatrachus 1s currently being examined as part of a phylogenetic study of the Phrynobatrachidae (sensu Frost, 2007; Zimkus, unpublished).

The long taxonomic history linking Ar- throleptis and Phrynobatrachus contributes to the present difficulties resolving the identifications of these distantly related frog genera (e.g., Boulenger, 1882). Many species originally described as Arthroleptis have been since transferred to other genera, including 21 species that are currently considered valid species of Phrynobatrachus (Frost, 2007). In part on the basis of the work of Deckert (1938), Laurent (1940, 1941la, b) used oste- ological characters to separate species into two genera, Arthroleptis and Phrynobatra- chus, which he believed were not closely related. The morphological differentiation of these genera is further supported by recent molecular phylogenetic studies (e.g., Bossuyt et al., 2006; Frost et al., 2006; Roelants et al., 2007), which show that Arthroleptis and Phrynobatrachus belong to different radia- tions within ranoid frogs.

The purpose of this study is to clearly determine those external morphological characters that can be used to discrimi- nate between Arthroleptis and Phrynobatra- chus. We also document characters that could be useful for identification but vary within these genera. Because we have found that written descriptions of these characters are often unclear, we provide photographs to illustrate the more important characters. Although this information might not be entirely novel for tenured African amphibian biologists, we hope that those not familiar

2008

with these genera will find this to be a useful guide.

METHODOLOGY

Using the taxonomic, systematic, and faunal literature as our guide, we examined museum specimens for characteristics report- edly useful in differentiating between Arthro- leptis and Phrynobatrachus. We focused solely on external characteristics to make our observations equally useful to research- ers conducting field surveys or sorting specimens in museum collections. A total of 15 morphological characters were investigat- ed; six of these are male secondary sexual characters. Characters are provided in Ap- pendix 1, and Appendix 2 comprises a data matrix in which all species examined are coded for these characters.

The results of this work are based on our study of specimens from museum collections, as well as our respective experiences in the field. Museum abbreviations correspond to those of Leviton et al. (1985), with the exception of MMB (Museums of Malawi. Blantyre). The allocation of species to either Arthroleptis or Phrynobatrachus follows Frost (2007). High-resolution images were obtained with a JVC 3-CCD digital camera using AutoMontage Pro 5.0 (Synoptics). Images were saved as TIFF files, cropped, and contrast-adjusted with Adobe Photo- shop 7.0 for Macintosh. We gathered meris- tic data from 280 adult specimens represent- ing 27 species of Arthroleptis and 246 adult specimens representing 31 species of Phryno- batrachus (Online Supplement: Appendix 1).

Arthroleptis and Phrynobatrachus appear to generally exhibit different body shapes, with Arthroleptis having relatively wider heads. To test whether this difference is significant, we took measurements (+ 0.1 mm) of snout— vent length (SVL) and head width (HdWd) with digital calipers (Online Supplement:

DIAGNOSING ARTHROLEPTIS AND PHRYNOBATRACHUS 3

Appendix 1). We used an analysis of covari- ance to test for significant difference in the slope of the regression lines in which HdWd is the dependent variable and SVL is indepen- dent. To take into account variation in both HdWd and SVL, reduced major axis regres- sion was used. Statistical significance was evaluated for « = 0.05.

RESULTS

The characters examined and evaluated in this study are summarized in Table 1. Nearly all characters reported to differentiate Ar- throleptis and Phrynobatrachus were found to vary both within and between species in each genus (Table 1). Only two external morphological characters definitively differ- entiate these genera:

1. the presence of a tarsal tubercle and

2. the presence of an outer metatarsal tubercle.

However, among African anurans, only the presence of a tarsal tubercle is unique to Phrynobatrachus.

In general, Arthroleptis have relatively wider heads than Phrynobatrachus. The coefficients (B) of linear regression of HdWd on SVL are significantly different between these two genera (Arthroleptis: B = 0.461, SE = 0.006; Phrynobatrachus: B = 0.348, SE = 0.005; P < 0.0001, F = 64.39). Although highly significant, in practice it is very difficult to discriminate between these genera on the basis only of these data if specimens are less. than 30mm, SVL (Pic. 1). _ The difference in relative head width is only obvious for specimens more than 30 mm SVL (HdWd/SVL: 40% vs. 34%).

DISCUSSION

In this study, we examined a suite of morphological characters believed to be useful in distinguishing between adults of Arthroleptis and Phrynobatrachus. The most

4 BREVIORA

No. 513

EXTERNAL MORPHOLOGICAL CHARACTERS USEFUL FOR DIFFERENTIATING BETWEEN A RTHROLEPTIS

AND PHRYNOBATRACHUS.

TABLE l. Character Arthroleptis Tarsal tubercle absent Outer metatarsal tubercle absent Heel tubercle (located on the proximal absent end of the tarsus) Circummarginal groove on terminal absent phalanx Pedal webbing absent Median dorsal skin raphe present* Hourglass or triple diadem pattern on variable dorsum Chevron-shaped glands in scapular region absent Eyelid cornicle or spur absent Sexually mature males Elongate third finger variable Dermal digital spines variable Inguinal spines variable Nuptial excrescence (thickened pad of absent skin) on first finger Femoral gland absent Lateral vocal folds absent

Phrynobatrachus References

present Blackburn, 2005; Channing and Boycott, 1989; Drewes and Perret, 2000: Scott, 2005

present Deckert, 1938

variable Scott, 2005

variable Blackburn, 2005; Scott, 2005

variable Schmidt and Inger, 1959; Stewart, 1967

absent Drewes and Perret, 2000; Laurent, 1957; Scott, 2005

absent Scott, 2005

variable Scott, 2005; Stewart, 1974

variable Perret, 1988; Rodel, 2000

absent Blackburn, 2005; Noble, 1931

absent Noble, 1931; Schmidt and Inger, 1959

absent Blackburn, unpublished

variable Scott, 2005; Stewart, 1967

variable Blackburn, 2005: Parker, 1935: Stewart, 1967

variable Stewart, 1967

*Variation is the result of preservation, desiccation, or both.

definitive characteristic that can be used to differentiate between these two genera 1s the presence of both an outer metatarsal tubercle and a tarsal tubercle in Phrynobatrachus (including the previously synonymized Di- morphognathus and Phrynodon) (Fig. 2B). Arthroleptis (including Schoutedenella) only exhibits an inner metatarsal tubercle, which is also found in Phrynobatrachus. Although Drewes and Perret (2000) found Phrynoba- trachus (Dimorphognathus) africanus and Phrynobatrachus (Phrynodon) sandersoni to lack a tarsal tubercle, we found the tarsal tubercle present in both taxa, albeit rather reduced in size in the latter species. Numerous morphological characters ex- amined in this study are present in other

° Arthroleptis o Phrynobatrachus

E £ Tv = Be} ir SVL (mm) Figure 1. Head width (HdWd) plotted against

snout—vent length (SVL) for Arthroleptis (squares) and Phrynobatrachus (circles).

2008

African frogs, but the presence or combina- tion of particular characters can be useful in distinguishing between adult Phrynobatra- chus and Arthroleptis. Expansion of the terminal phalanx of manual and pedal digits is found in both genera, but circummarginal grooves are found only in some species of Phrynobatrachus (Blackburn, 2005; Scott, 2005; Fig. 2E). Phrynobatrachus differs from Arthroleptis by the presence of pedal web- bing in many species of the former (Schmidt and Inger, 1959). Webbing ranges from absent or rudimentary in numerous smaller species such as P. minutus, P. parvulus, and P. ukingensis to extensive in the majority of larger, rheophilic species such as P. acutiros- tris, P. krefftii, and P. versicolor (Fig. 2C, D). Five species of Phrynobatrachus, includ- ing P. annulatus, P. calcaratus, P. cornutus, P. taiensis, and P. villiersi, possess a single spinelike dermal tubercle on the upper eyelid (Rodel, 2000), a character that can be used to distinguish Phrynobatrachus from Arthrolep- tis (Perret, 1988), as well as other African anurans.

A number of external morphological characters vary both within and between species in each genus, and this variation is due to either natural variation within or between populations or the condition of preserved specimens. Chevron-shaped glands in the scapular region of Phrynobatrachus and a dorsal triple diadem pattern in Arthroleptis can both be used to identify species of these genera (Scott, 2005; Stewart, 1974). However, these are not always prom- inent in all members of a species or may be completely lacking in some species. The size and shape of the chevron-shaped glands of Phrynobatrachus are quite variable and can originate and terminate in the scapular region or instead extend almost the entire length of the body (Fig. 2H). The median dorsal skin raphe of Arthroleptis can be a difficult character to visualize, especially in

DIAGNOSING ARTHROLEPTIS AND PHRYNOBATRACHUS

n

poorly preserved specimens (Fig. 2G); how- ever, the presence of this character can be used to accurately identify Arthroleptis. Finally, some species of Phrynobatrachus exhibit circummarginal grooves on the man- ual or pedal digit tips; in some species, these furrows are found only on the longest digits. In addition, the presence of circummarginal grooves could be difficult to determine because of desiccation of preserved speci- mens. However, if circummarginal grooves are present, this character can be used to differentiate Phrynohatrachus from Arthro- leptis, because it is lacking in the latter, but not necessarily from other African frog genera (Blackburn, 2005; Scott, 2005; Fig. 2E).

The presence of male secondary sexual characters can lead to the most straightfor- ward identifications of specimens as Arthro- leptis or Phrynobatrachus. An elongate third finger is found in many male Arthroleptis, as well as most species of Cardioglossa, the putative sister genus of Arthroleptis (Black- burn, in press; Fig. 3A, B). This sexual dimorphism is not found in Phrynobatrachus or any other anurans. Males of nearly all Arthroleptis species have dermal spines lining the medial surface of the elongate third finger and are also sometimes found on the lateral, medial, or both surfaces of the second finger (Fig. 3B). In some species, the presence of spines can vary seasonally (e.g., Schmidt and Inger, 1959). Males of some species of Phrynobatrachus possess a nuptial excrescence or thickened pad of skin on the medial and dorsal surface of the first finger, which was hypothesized by Parker (1940) to be an adaptation to aquatic amplexus. In P. africanus, this pad 1s greatly hypertrophied and covers much of the dorsal surface of the hand (Fig. 3C). The presence of an elongate ovoid femoral gland, which is most evident in life because it often can be hardened and yellow in color, allows the

BREVIORA No. 513

Figure 2. Morphological characters useful in distinguishing between Arthroleptis and Phrynobatrachus. A, Left foot of A. stenodactylus (MCZ A-137060) in ventral view illustrating inner metatarsal tubercle and lack of webbing. B, Left foot of P. natalensis (MCZ A-138084) in ventral view illustrating inner metatarsal tubercle (a), outer metatarsal tubercle (b), and tarsal tubercle (c). C, Left foot of P. parvulus (MCZ A-137121) in ventral view

illustrating absent or rudimentary webbing. D, Right foot of P. steindachneri (MCZ A-136907) in ventral view

2008

identification of breeding Phrynobatrachus males (Blackburn, 2005; Stewart, 1967). However, in most species, this gland might be difficult to identify after preservation. Breeding Phrynobatrachus males have a single subgular vocal sac, which, when not distended, forms one or multiple folds, roughly parallel to the lower jaw, on the lateral margins of the throat (Stewart, 1967: Fig. 3E). Although the gular region of Arthroleptis can be distended and wrinkled, prominent gular folds are not present (Fig. 3D).

For many field biologists that regularly collect Arthroleptis and Phrynobatrachus, it is obvious that the body proportions of these frogs generally differ. Indeed, we find that Arthroleptis have heads that are relatively wider than Phrynobatrachus. Meristic data collected in this study illustrate that the linear regression coefficients differ signifi- cantly between these two genera. However, in general, the difference in body proportions is only obvious in large specimens (> 30 mm SVL). Thus, at least for larger specimens, the relative width of the head could be a useful quick diagnostic feature, especially if used in concert with other morphological characters discussed herein (Table 1).

It is also important to note that numerous nonmorphological characters such as call structure, breeding biology, and _ habitat preference can also be used in the field to assist in the identification of Arthroleptis and Phrynobatrachus. All Arthroleptis are be- lieved to have direct development, in which

DIAGNOSING ARTHROLEPTIS AND PHRYNOBATRACHUS 7

hatch from a small clutch of terrestrially deposited eggs. In contrast, most Phrynobatrachus species deposit hundreds to thousands of eggs in ponds, streams, or pools, and a small number of species deposit small clutches of eggs in stagnant water found in tree holes, in empty fruit capsules, within snail shells, or terrestrially (Rédel, 1998; Rodel and Ernst, 2002). Species exhibiting these alternative reproductive modes include P. dendrobates, P. guineensis, P. krefftii, P. phyllophilus, P. sandersoni, and P. tokba, although all have free-living tadpoles (Amiet, 1981; Rédel, 1998: Rédel and Ernst, 2002). Although few advertise- ment calls of these genera are published, those currently available will undoubtedly assist in distinguishing Phrynobatrachus from Arthroleptis (e.g., Drewes and Perret, 2000; Rodel, 2000; Schiotz, 1964).

Although we demonstrate in this study that it is possible to distinguish adult Phrynobatrachus and Arthroleptis with the use of morphological characters, the identi- fication of juvenile and subadult Arthroleptis and Phrynobatrachus continues to be chal- lenging, even occasionally for the authors of this study. Fortunately, the identification of numerous taxa has been greatly facilitated by the recent use of DNA barcoding (e.g., Hebert et al, 2003; Moritz and Cicero, 2004; Vences et al., 2005). The use of the mitochondrial 16S ribosomal RNA (rRNA) gene has proven to be particularly successful in amplification of amphibian DNA (Vences et al., 2005). We have found that the

froglets

illustrating extensive webbing. E, Tips of left pedal digits II and IV of P. tokba (MCZ A-26905) illustrating

expansion of digit tip and circummarginal grooves. F, Dorsal surface of A.

stenodactylus (MCZ A-137060)

illustrating the hourglass (i.e., triple diadem) pattern. G, Dorsal surface of A. sp. nov. (MCZ A-137978), in life, illustrating the median dorsal skin raphe. H, Dorsal surface of P. steindachneri (MCZ A-136907) on left and P. auritus (MCZ A-138095) on right illustrating variability in chevron-shaped glands. Scale bar | mm in 2A—D, F-H;

0.1 mm in 2E.

orc

M¢

Z A-13

v1IeW

lishing between Arthroleptis and Phrynoba-

showing elongate digit III. B, Fingers I lermal spines. C, Right hand of male P a anus (MCZ A-136757), below. f male A. schubotzi (CAS 201717), left, and

y dermal

Ids. E, Throat of male P. auritus (MCZ A-

ible vocal folds. Scale bar 1 mm

2008

amplification of this gene is effective in identifying tadpoles, juveniles, and subadults of numerous African ranoid frogs, including Phrynobatrachus and Arthroleptis. However, molecular data and analyses can be con- founded by the misidentification of voucher specimens. For instance, nucleotide BLAST searches against GenBank sequence data for the Arthroleptis specimens for which Vences et al. (2003) amplified the 16S rRNA gene reveals that these specimens are misidentified (GenBank sequences AF215139-40). AI- though Vences et al. (2003) attribute the unusual paraphyly of Arthroleptis 16S genes to uncertainties in their sequence data, this result is easily explained: These sequences have high similarity to Phrynobatrachus and thus are misidentified at the genus level. We advocate DNA identification of specimens only to the level of genus because of the difficulty in accurately identifying species of either genus. Thorough taxonomic study in combination with molecular data will be necessary before DNA identification can be used confidently to identify species of these two genera.

ACKNOWLEDGMENTS

This study was supported by the Depart- ment of Organismic and Evolutionary Biol- ogy and Museum of Comparative Zoology (Harvard University) and NSF grant EF- 0334939 (AmphibiaTree) to J. Hanken. Many thanks to the following individuals who made specimens under their care avail- able for study: M. Burger, J. Campbell. D. Cannatella, B. Clarke, R. Drewes, R. Gun- ther, V. Gvodzik, T. Harstell, T. LaDuc, L. Lawson, S. Loader, P. Malonza, L. Mazi- buko, J. McGuire, J. Measey. A. Ohler, A. Resetar. J. Rosado, A. Schmitz, E. Scott, J. Simmons, L. Trueb. J. Vindum, D. Wake, A. Wynn. J. Hanken provided useful comments on a draft of the manuscript.

DIAGNOSING ARTHROLEPTIS AND PHRYNOBATRACHUS 3,

APPENDIX 1. MORPHOLOGICAL CHARACTERS EXAMINED.

External Morphology

1. Tarsal tubercle: (0) absent; (1) present.

2. Outer metatarsal tubercle: (0) absent: (1) present.

3. Heel tubercle (located at the proximal end of the tarsus): (0) absent; (1) present.

4. Circummarginal groove at manual and/ or pedal digit tips: (0) absent: (1) present. In some species, this groove is present only on the tips of the longest digits. Variation in museum specimens might be the result of preservation, desiccation, or both.

5. Pedal webbing: (0) absent or rudimen- tary with more than two distalmost phalan- ges of digit IV unwebbed: (1) moderate to extensive with no more than the two distal- most phalanges of digit IV unwebbed.

6. Median dorsal skin raphe: (0) absent: (1) present.

7. Hourglass or triple diadem pattern on dorsum: (0) absent; (1) present.

8. Chevron-shaped glands in scapular region: (0) absent: (1) present. These ridges of skin can be short or can extend the entire length of the body when present.

9. Eyelid cornicle or spur: (0) absent; (1) present.

Secondary Sexual Characteristics

10. Sexually mature males, third finger relatively longer than in females: (0) absent: (1) present.

11. Sexually mature males, dermal spines on fingers: (0) absent; (1) present.

12. Sexually mature males, inguinal spines: (0) absent; (1) present.

13. Sexually mature males, nuptial excres- cence (thickened pad of skin that can appear velvety) on first finger: (0) absent: (1) present.

14. Sexually mature males, femoral glands: (0) absent: (1) present. Most often the

10

femoral glands of Phrynobatrachus are |\o- cated on the posterior thigh. However.

BREVIORA

No. 513

15. Sexually mature males, lateral vocal folds: (0) absent: (1) present. When pres-

glands can be situated closer to either the ent. folds run roughly parallel to the knee or vent. These glands are most easily lower jaw, at the lateral margins of seen in living specimens because they are the throat. and form one or multiple usually bright yellow. creases.

APPENDIX 2.

DISTRIBUTION OF MORPHOLOGICAL CHARACTERS AMONG SPECIES OF ARTHROLEPTIS AND PHRYNOBATRACAUS. ALL CHARACTER STATES ARE BINARY (0. 1). MISSING DATA ARE CODED AS *?”:; 0,1 DENOTES POLYMOR-

PHISM. REFER TO APPENDIX | FOR CHARACTER DESCRIPTIONS.

| Paar. 3 4A. 5° 6 7 8 9 10. fh 2 13 14 15 Arthroleptis adelphus 0 O 0 Or at 1 0.1 0 0 1 1 1 0 0 0 A. adolfifriederici O> Oy S07 “0% to at 0 OO 1 1 1 0 0 0 A. affinis 5°00 OF tise e031 Oren O ] ] 0 0 0 0 A. brevipes OO Os 20): dinmec 0 or 0 ? ? ? 0 0 0 A. crusculum Dano 0 Te i! 1 0.1 0 0 1 1 1 0 0 0 A. francei 0 160250 7 05 ff 7 os OTTO 1 | 1 0 0 0 A_ hematogaster Ge FeO. ey O ez, si 0 O57: O72 4 ? 0 0 0 A. lameerei 0 O 0 0 1 LO 0 0 1 0 0 0 0 0 A. nikeae 0 O 0 Oe vie ed 0 0 0 ? v4 ? 0 0 0 A. nimbaensis 0 O 0 UR Fe aa 1 21 | 0 0 1 1 1 0 0 0 A. “poecilonotus™ Gc 0.540" 0) ots iy SOs OF O'S E 1 1 0 0 0 A. pyrrhoscelis O00) 0 ss 0 0 0°. 1 0 0 0 0 0 A. reichei OF05 0) On tte 6-1 0” 40 ] 1 1 0 0 0 A. schubotzi io 0" 02a tt Ot 0 O 1 1 | 0 0 0 A. stenodactylus OO: 120) Orn le 401 0 O 1 1 1 0 0 0 A. sylvaticus Ome O02 120 Certs SiO 0 O east 1 0 0 0 A. taeniatus 0 O 0 Oa yt 1 0 0 0 1 1 1 ) 0 0 A. tanneri O76) 6. OL tie 0 Gr 0 1 1 1 0 0 0 A. troglodytes OF 70a 0770 ot ie 041 0 oO ? 1 1 0 0 0 A. tuberosus O70? Cie nOa ek 1051 OP 0 1 1 ] 0 0 0 A. variabilis 0 a0e O00" te Os 0 O 1 1 1 0 0 0 A. wahlbergii O00 5:0 St 3 01 0 O 1 1 1 0 0 0 A. xenochirus 0 O 0 Oe be cit 0.1 0 0 1 1 1 0 0 0 A. xenodactyloides OO OO, Ae Ost 0 O 1 1 1 0 0 0 A. xenodactylus OOO ees et a 0 0 O 1 1 1 0 0 0 A. zimmeri 0 O 0 Oye sis kz ? 0 0 ? ? ? 0 0 0 Arthroleptis sp. nov. 0-0") 0) OG. 1 A 0 0: 0, ~2 ? ? 0 0 0 Phrynobatrachus acridoides 5 On eal ie 0 0 1 0 O 0 0 1 1 1 P. acutirostris 1 1 1 | io 0 l 0 0 0 0 1 0 ] P. africanus PPA Oy Te 201 0 0 1 0 O 0 0 1 0 1 P. auritus 1 1 1 1 Me MG 0 1 0 0 0 0 0 1 1 P_ batesii ees | 0 1 0 O 0 ] 0 O 0 0 ? ? ? P. bequaerti isp On 10 Oe 0 0 ] OF 20 0 0 1 0 0 P. bullans tL 4 ] Oa 0 0 1 0 O 0 0 1 1 1 P. calcaratus 1 i 0a 20020 0 1 1 0 0 0 0 0 1

2008 DIAGNOSING ARTHROLEPTIS AND PHRYNOBATRACHUS 11 APPENDIX 2. Continued.

iC 2 3, 4.57 46 7 S. Serie iy] 2 13 14 15 P. cornutus 1 ] 0 i 0 6 0 ] 1 0 0 0 ? 2 ? P. cricogaster ] ] ] ] 1 O 0 1 0 0 0 0 ] 0 0 P. cryptotis Pei (Os Os 0 0 0 ] 0 O 0 0 0 0 1 P. dendrobates ] l 0 L076 0 1 0 0 0 0 | 0 1 P. francisci 1 1 OF 2:0) Ail 20 0 ] 0 O 0 0 ? ? ? P. inexpectatus heel OF 0 OL 30 0 G7 09 56 0 0 0 ] ] P. keniensis 1 1 0 OF~ 04.20 0 1 0 0 0 0 1 0 1 P. kinangopensis 19 oh 0:16 SO aes 0 0 1 0 O 0 0 ] 0 ] P. krefftii ee | 0 Pst 2D 0 1 0 O 0 0 1 1 1 P. mababiensis fe i O81 0; 0" 50 0 ] > san0 0 0 ? ? ? P. manengoubensis | 0 1200 0 ] VucO 0 0 0 ] 1 P. minutus ie 00" 7.0'470 0 | Of>.0 0 0 ? ? ? P. natalensis ] Por) Sil =p 0 0.1 O 0 0 0 1 1 1 P. pakenhami Os ae, eee att 0) 038" 20 0 0 0 1 ] P. parvulus Pad ee OS 0S a 0 is=_ 05 =70 0 0 ] ] 1 P. plicatus bee te (0-2 Lr SO 0 ft 0 20 0 0 0 0 ] P. rungwensis i as | 1 10: <0 0 1 0 O 0 0 1 1 1 P. sandersoni 1 1 1 1 0 0 1 0 0 0 0 1 1 1 P. scheffleri i, ot 1 0% 202740 0 1 0 O 0 0 ? ? ? P. steindachneri i. AO Pe tes SO 0 ] 0 0 0 0 1 1 0 P. stewartae 1s 5 a4 | Gr 0 ] 0 0 0 0 1 1 1 P. ukingensis ee els OES Os SD 0 0.1 0 O 0 0 1 1 1 P. versicolor 1 1 0 ] tea0 0 ] 0 0 0 0 1 0 1

LITERATURE CITED

AmicT, J.-L. 1981. Ecologie, éthologie et développement de Phrynodon sandersoni Parker, 1939 (Amphibia. Anura, Ranidae). Amphibia-Reptilia. 2: 1-13.

BLACKBURN, D. C. 2005. Cardioglossa liberiensis Bar- bour & Loveridge 1927 is a junior synonym of Phrynobatrachus fraterculus (Chabanaud 1921). African Journal of Herpetology. 54: 171-179.

—. A new species of Cardioglossa (Amphibia: Anura: Arthroleptidae) endemic to Mount Manen- gouba in the Republic of Cameroon. with an analysis of morphological diversity in the genus. Zoological Journal of the Linnean Society. In press.

Bossuyt, F.. R. M. Brown, D. M. Hiiuis. D. C. CANNATELLA, AND M. C. Miutinxovitcu. 2006. Phylogeny and biogeography of a cosmopolitan frog radiation: Late Cretaceous diversification resulted in continent-scale endemism in the family Ranidae. Systematic Biology. 55: 579-594.

BouLENGER, G. A. 1882. Catalogue of the Batrachia Salientia s. Ecaudata in the Collection of the British Museum. Second edition. London, British Muse- um, 503 pp. + 30 pls.

CHANNING, A., AND R. C. Boycott. 1989. A new frog genus and species from the mountains of the

southwestern Cape. South Africa (Anura: Rani- dae). Copeia, 1989: 467-471.

DeEcKERT. K. 1938. Beitrage zur osteologie und system- atic ranider froschlurche. Sitzungsberichte Gesell- schaft Naturforschender Freunde Berlin, 1938: 127-184.

Drewes, R. C., AND J.-L. PERRET. 2000. A new species of giant, montane Phrynobatrachus (Anura: Rani- dae) from the central mountains of Kenya. Pro- ceedings of the California Academy of Sciences. 52: 55-64.

Frost, D. R. 2007. Amphibian Species of the World: An Online Reference. Version 5.0. New York, Amer- ican Museum of Natural History. Available from: http://research.amnh.org/herpetology/amphibia/index. php.

. T. Grant, J. Fatvovicn, R. H. Bary, A. Haas, C. F. B. Happap, R. O. de Sa, A. CHANNING, M. WILKInson, S. C. DONNELLAN, C. J. RAxwortny, J. A. CampsBett, B. L. Biotro, P. Mo rer, R. C. Drewes, R. A. Nusssaum, J. D. Lyncn, D. M. GREEN, AND W. C. WHEELER. 2006. The amphibian tree of life. Bulletin of the American Museum of Natural History, 297: 8-291.

Hesert, P. D. N., A. Cywinska, S. L. BALL, AND J. R. DE Waarpb. 2003. Biological identification through

DNA barcodes. Proceedings of the Royal Society of London Series B, 270: 313-321.

LarGEN, M. J. 2001. The status of the genus Phrynoba- trachus Gunther 1862 in Ethiopia and Eritrea, including description of a new species (Amphibia Anura Ranidae). Tropical Zoology, 14: 287-306.

LaureENT, R. F. 1940. Contribution a Postéologie et a la systématique des ranides africains, Premiere Note. Revue de Zoologique Botanique Africaine, 34: 74-97, 3 pls.

———. 194la. Contribution a Vostéologie et a la systématique des rhacophorides africains, Premiére Note. Revue de Zoologique Botanique Africaine, 35: 85-110, 7 pls.

1941b. Contribution a lostéologie et a la systématique des ranides africains, Deuxi¢me Note. Revue de Zoologique Botanique Africaine, 35: 192-234, 2 pls.

———. 1954. Remarques sur le genre Schoutedenella Witte. Annales du Musée du Congo, Tervuren, in 4°, Zoologie, 1: 34-40.

. 1957. Notes sur les Hyperoliidae. Revue de Zoologique Botanique Africaine, 56: 274-282. Leviton, A. E., R. H. Grsss, Jr., E. HEAL, AND C. E. Dawson. 1985. Standards in herpetology and ichthyology: part I. Standard symbolic codes for institutional resource collections in herpetology and

ichthyology. Copeia, 1985: 802-832.

Moritz, C., AND C. Cicero. 2004. DNA barcoding: promise and pitfalls. PLoS Biology, 2: 1529-1531.

Nosie, G. K. 1931. Biology of the Amphibia. New York, Dover Publications. x11 + 577 pp.

Parker, H. W. 1935. A new genus of frogs from the Cameroons. Annals and Magazine of Natural History, 10: 401-404.

1940. The Australian frogs of the family Leptodactylidae. Novitates Zoologicae, 42(1): 1-106.

Perret, J.-L. 1988. Les especes de Phrynobatrachus (Anura, Ranidae) a éperon palpébral. Archives des Sciences (Geneve), 41: 275-294.

Ropet, M.-O. 1998. A reproductive mode so far unknown in African ranids: Phrynobatrachus guineensis Guibé and Lamotte, 1961 breeds in tree holes (Anura: Ranidae). Herpetozoa, I1: 19-26.

12 BREVIORA

No. 513

. 2000. Herpetofauna of West Africa. Volume I. Amphibians of the West African Savanna. Frank- furt, Edition Chimaira, 332 pp.

. AND R. Ernst. 2002. A new reproductive mode

for the genus Phrynobatrachus: Phrynobatrachus

alticola has nonfeeding, nonhatching tadpoles.

Journal of Herpetology, 36: 121-125.

, AND M. A. BANGourA. 2004. A conservation assessment of amphibians in the Foret Classée du Pic de Fon, Simandou Range, southeastern Repub- lic of Guinea, with the description of a new Amnirana species (Amphibia Anura_ Ranidae). Tropical Zoology, 17: 201-232,

RoeLaAnts, K., D. J. Gower, M. WILkinson, S. P. Loaper, S. D. Buu, K. GuILLAUME, L. Mortau, AND F. Bossuyt. 2007. Global patterns of diversi- fication in the history of modern amphibians. Proceedings of the National Academy of Sciences, 104: 887-892.

Scutotz, A. 1964. The voices of some West African amphibians. Videnskabelige Meddelelser fra Dansk Naturhistorisk Forening, 127: 35-83,

ScumipTt, K., AND R. F. INGER. 1959. Amphibians, exclusive of the genera Afrixalus and Hyperolius. Exploration du Pare National de /!Upemba, Mis- sion G. F. de Witte, Institut des Parcs Nationaux du Congo Belge, 56: 1-264, 9 pls.

Scott, E. 2005. A phylogeny of ranid frogs (Anura: Ranoidea: Ranidae), based on a simultaneous analysis of morphological and molecular data. Cladistics, 21: 507-574.

Stewart, M. M. 1967. Amphibians of Malawi. New York, State University of New York Press. ix + 163 pp.

1974. Parallel pattern polymorphism in the genus Phrynobatrachus (Amphibia: Ranidae). Co- peia, 1974: 823-832.

Vences, M., J. Kosucu, F. GLAw, W. BOHME, AND M. VeitH. 2003. Molecular phylogeny of hyperoliid treefrogs: biogeographic origin of Malagasy and Seychellean taxa and re-analysis of familial para- phyly. Journal of Zoology, Systematics, and Evo- lutionary Research, 41: 205-215.

———, M. Tuomas, A. VAN DER MEIJDEN, Y. CHIARI, AND D.R. Vertes. 2005. Comparative performance of the 16S rRNA gene in DNA barcoding of amphibians. Frontiers in Zoology, 2: 5.