Ebenavia

Abstract: Aim We studied the gecko genus Ebenavia to reconstruct its colonization history, test for anthropogenic versus natural dispersal out of Madagascar, and correlate divergence date estimates of our phylogeny with geological age estimates of islands in the region. Location Madagascar and surrounding islands of the Western Indian Ocean (Comoros, Mayotte, Mauritius, Pemba). Methods We reconstructed the phylogeny of Ebenavia covering its entire geographical range using a molecular data set of three mitochondrial and two nuclear markers. We estimated divergence times based on calibrations using (1) previously calculated mutation rates of mitochondrial markers, (2) a combination of these rates with old or (3) young geological age estimates for some of the islands inhabited by the genus, and (4) an independent data set with fossil outgroup calibration points. Results Ebenavia inunguis, one of two recognized species of the genus, comprises multiple ancient evolutionary lineages. The earliest divergence within this complex (Miocene, 13–20 Ma; 95% credibility interval [CI]: 4–29 Ma) separates the population of the Comoros Islands, excluding Mayotte, from all other lineages. The age estimates for island lineages coincide with the geological age estimates of the islands except for Grand Comoro, where the age of the local clade (3–5 Ma; 95% CI: 2–7 Ma) significantly predates the estimated island age (0·5 Ma). A clade from north Madagascar + Mayotte + Pemba is estimated to have diverged from an eastern Malagasy clade in the Miocene. Main Conclusions Our results suggest that Grand Comoro Island is geologically older than previously estimated. The islands of the Comoros and Pemba were probably colonized via natural dispersal out of Madagascar (> 1000 km in the case of Pemba). Mauritius was most likely colonized only recently from eastern Madagascar via human translocation.

Abstract: The gecko genus Ebenavia Boettger is rediagnosed as phyllodactyle gekkonines with mostly multicarinate scales and claws restricted to the digits of the pes of females. We describe a second, diminutive species of Ebenavia from Lac Tsimanampetsotsa on the southwestern coast of Madagascar. This new species differs from the type species, E. inunguis, in its smaller size, darker and more uniform coloration, presence of multicarinate postrostral scales, and a relatively larger and differently shaped first supralabial scale. The phylogenetic affinities of Ebenavia are unknown, but the two species share similarities with Phylloclactylus peringueyi of South Africa, and the possibility that the three species form a monophyletic group should be explored. The discovery of this new species adds to the growing list of Madagascan species-pairs that have east-west disjunctions, with western species adapted to xeric conditions and eastern species found in rainforest. Key wordts: Reptilia; Squamata; Gekkonidae; Gekkoninae; Ebenavia; Systematics; New species; Madagascar THE monotypic gekkonine genus Ebenavia Boettger is confined to the islands of the western Indian Ocean, where it is recorded from Madagascar, the Comoro Islands, and Mauritius. Pakenham (1947) 1 PRESENT ADDRESS: CERC, Columbia University, 1200 Amsterdam Ave., MC 5557, New York, NY 10027, USA. described eggs and hatchlings of Ebenavia cf. inunguis from Pemba Island, and Loveridge (1957) confirmed the identifications, but there have been no subsequent reports of the genus on Pemba. Warman and Todd (1977) reported Ebenavia inunguis on Aride Island in the granitic Seychelles, but this was based on a specimen of Urocotyledon inexpectata (Nussbaum, This content downloaded from 157.55.39.144 on Thu, 22 Sep 2016 06:57:52 UTC All use subject to http://about.jstor.org/terms March 1998] HERPETOLOGICA 19 unpublished data; see Cheke, 1984). Early reports of Ebenavia in western Australia are based on Ebenavia horni Lucas and Frost, 1895, a species subsequently synonymized with Diplodactylus ocellatus and placed in a new monotypic genus, Crenadactylus, by Dixon and Kluge (1964). Ebenavia was established on dubious grounds. Boettger (1878) diagnosed it essentially as equivalent to Phyllodactylus, but without claws. Subsequent authors, including Boulenger (1885), Mocquard (1909), Angel (1942), and Russell (1972), repeated this diagnosis without modification. Russell (1972), however, doubted whether clawlessness was sufficient grounds for separating Ebenavia from Phyllodactylus, and stated that the internal and functional anatomy of the digits of E. inunguis are identical to Phyllodactylus. Recently, several studies (e.g., Dixon and Kroll, 1974; Kluge, 1983) have shown that "Phyllodactylus" is an heterogeneous assemblage of geckos, and many species have been transferred to other genera. Therefore, Boettger's original diagnosis of Ebenavia as a clawless Phyllodactylus, sensu lato, has become meaningless. Furthermore, our observations indicate, contrary to the literature, that some specimens of Ebenavia inunguis have small claws on some digits. These circumstances mean that even though E. inunguis is a distinctive, easily recognizable species, the genus Ebenavia is of uncertain status. It lacks a formal diagnosis by which it can be distinguished from other gekkonine genera. In Madagascar, Ebenavia inunguis is primarily a species of the eastern, low elevation rainforests and adjacent degraded habitats where it is normally found on vegetation (trunks, stems, leaf axils) and only infrequently on rocks. We recently discovered a second, partially rock-dwelling, dwarf species of Ebenavia in the desert regions of southwestern Madagascar. The discovery of this new species is important, because it helps to resolve the status of the genus Ebenavia, identifies Madagascar as the center of evolution of the genus, and establishes another link in the emerging pattern of east-west vicariance in the herpetofauna of Madagascar. In this paper, we review the nomenclatural history of Ebenavia, suggest a new diagnosis for the genus, and describe a second species. METHODS AND MATERIALS Specimens were anesthetized by injection with chlorobutanol, fixed in 10% buffered formalin, soaked in water to remove the formalin, and stored in a final solution of 75% ethanol. Snout-vent and tail lengths were measured with a ruler to the nearest 1.0 mm; other measurements were taken with dial calipers and recorded to the nearest 0.1 mm. Specimens are identified by catalog numbers of the Museum of Zoology, The University of Michigan (UMMZ) and by field numbers (RAN). Other abbreviations used are BMNH (British Museum) and SMF (Natur-Museums und Forschungs-Institutes Senckenberg). Throughout this paper, we use "apotypic", "plesiotypic", and derivitives thereof rather than "apomorphic", etc., following the logic of Tuomikoski (1967) and L0vtrup (1977). NOMENCLATURAL HISTORY Boettger (1878) established Ebenavia and described the only known species, E. inunguis, based on a single specimen collected on Nosy Be, Madagascar. Boettger originally diagnosed Ebenavia as a clawless Phyllodactylus lacking enlarged scales on the undersurface of the tail and without large chin shields. Boulenger (1885:96) diagnosed Ebenavia as differing "from Phyllodactylus in having all the digits destitute of claws" and repeated, in abbreviated form, Boettger's description of the holotype of E. inunguis. He also described E. boettgeri from "Madagascar" as a new, second species of the genus. It is unclear why Boulenger thought E. boettgeri was distinct from E. inunguis. Only the holotypes of the two species were known, and Boulenger had not examined the holotype of E. inunguis. The most striking differences, according to Boulenger's descriptions, are in dorsal coloration: olive-brown in E. inunguis and reddish brown in E. boettgeri. After examining the holotype of E. inunguis, Boulenger (1887:482) decided E. boettgeri was not valid, and all subsequent This content downloaded from 157.55.39.144 on Thu, 22 Sep 2016 06:57:52 UTC All use subject to http://about.jstor.org/terms 20 HERPETOLOGICA [Vol. 54, No. 1 authors (Mocquard, 1909; Angel, 1942) have considered E. boettgeri to be a junior synonym of E. inunguis, a conclusion with which we concur. Angel (1942) also viewed Ebenavia inunguis as a clawless Phyllodactylus, and he listed 11 species of Phyllodactylus for Madagascar and the Comoros. The heterogeneous Madagascan "Phyllodactylus" has since been reduced through taxonomic revision to a single species, P. brevipes, of doubtful generic assignment. Dumeril and Bibron (1836) and Boulenger (1885) reported that Phyllodactylus porphyreus, a well known South African species, also occurs in "Madagascar", and Angel (1936) described P. barbouri on the basis of a specimen in the Paris Museum that had been recorded as P. porphyreus from "Madagascar." Kluge (1965) believed that the holotype of P. barbouri is a specimen of Diplodactylus vittatus, a species otherwise known only from Australia, and he concluded that the locality for the holotype of P. barbouri was wrongly recorded. However, Henle (in Brygoo, 1991:33) indicated that P. barbouri is synonymous with a different Australian species, D. granariensis. Guibe (1956) also thought that P. porphyreus occurs in Madagascar, but this was based on his belief that Mocquard's (1900) P. brevipes is a junior synonym of P. porphyreus. Phyllodactylus brevipes is, however, a distinctive species endemic to Madagascar, which only superficially resembles P. porphyreus (Nussbaum and Raxworthy, unpublished data). We conclude that there is no evidence that P. porphyreus occurs in Madagascar. Dixon and Kroll (1974) placed the remaining species of Madagascan "Phyllodactylus" in the resurrected genus Paroedura, leaving "Phyllodactylus" brevipes as the single putative species of Phyllodactylus in Madagascar. Because Ebenavia was originally diagnosed with very few characters and only by comparison to Phyllodactylus, and because the concept of Phyllodactylus has changed considerably, the taxonomic status and diagnostic characteristics of Ebenavia are uncertain. Recent phylogenetic analyses of AfroMadagascan geckos that include Ebenavia B FIG. 1-.Dorsal views of snouts of (A) Ebenavia inunguis (UMMZ 207477) and (B) E. maintimainty (UMMZ 211208), illustrating differences in keeling patterns of postrostral scales and relative size of prenasal scale. Scale = 2 mm. (Bauer, 1990a; Kluge and Nussbaum, 1995) suggested various relationships with other geckos, but they did not lead to a formal diagnosis of Ebenavia. CHARACTERISTICS OF EBENAVIA The discovery of a new gecko in southwestern Madagascar, similar in overall body form to Ebenavia inunguis, led to a search for synapotypic character states that would verify their sister-species status. The most striking feature of the two species is the multicarinate condition of most dorsolateral scales (Figs. 1-3). Of the dorsolateral head scales, only the rostral, nasals, some internasals and postrostrals, and the anterior supralabials are acarinate. All dorsolateral body scales are multicarinate, as are the dorsal and preand postaxial limb scales. Dorsolateral caudal scales may or may not be multicarinate, depending on species and whether or not the tail has been regenerated. The pattern of keeling on the ventral scales is also variable. The This content downloaded from 157.55.39.144 on Thu, 22 Sep 2016 06:57:52 UTC All use subject to http://about.jstor.org/terms March 1998] HERPETOLOGICA 21

Abstract: Cryptic species have been detected in many groups of organisms and must be assumed to make up a significant portion of global biodiversity. We study geckos of the Ebenavia inunguis complex from Madagascar and surrounding islands and use species delimitation algorithms (GMYC, BOLD, BPP), COI barcode divergence, diagnostic codon indels in the nuclear marker PRLR, diagnostic categorical morphological characters, and significant differences in continuous morphological characters for its taxonomic revision. BPP yielded ≥ 10 operational taxonomic units, whereas GMYC (≥ 27) and BOLD (26) suggested substantial oversplitting. In consequnce, we resurrect Ebenavia boettgeri Boulenger 1885 and describe Ebenavia tuelinae sp. nov., Ebenavia safari sp. nov., and Ebenavia robusta sp. nov., increasing the number of recognised species in Ebenavia from two to six. Further lineages of Ebenavia retrieved by BPP may warrant species or subspecies status, but further taxonomic conclusions are postponed until more data become available. Finally, we present an identification key to the genus Ebenavia, provide an updated distribution map, and discuss the diagnostic values of computational species delimitation as well as morphological and molecular diagnostic characters.