Eleutherodactylinae, Lutz, 1954

Padial, José M., Grant, Taran & Frost, Darrel R., 2014, Molecular systematics of terraranas (Anura: Brachycephaloidea) with an assessment of the effects of alignment and optimality criteria, Zootaxa 3825 (1), pp. 1-132 : 59

Eleutherodactylinae

This clade includes the genera Diasporus and Eleutherodactylus as sister groups and was previously recovered by Hedges et al. (2008a), Pyron & Wiens (2011), and Canedo & Haddad (2012). Within Eleutherodactylus, Hedges et al. (2008a) recognized the subgenera Eleutherodactylus , Euhyas, Pelorius , and Syrrhophus , with the subgenus Schwartzius as the sister group of them all. We recover these groups as monophyletic and continue to recognize all of these subgenera, although we note that the inclusion of the wildcard Eleutherodactylus diasporus in Euhyas is assumed and must be tested by adding the DNA sequences that are missing for this taxon. All of these subgenera except the Hispaniolan endemic Schwartzius were recognized by Hedges (1989) on the basis of allozymic evidence, and were treated by Frost et al. (2006) as genera. Although this arrangement has not been followed by most authors, we expect that in the future all of these will be recognized as genera. Distinct characteristics of liver shape have been proposed as synapomorphic for each of the subgenera, as well as for Diasporus (Hedges, 1989; Hedges et al., 2008a).

Among the species series recognized by Hedges et al. (2008a) for the subgenera of Eleutherodactylus , all but the Eleutherodactylus martinicensis series are monophyletic in our preferred topologies. Therefore, we restrict the E. martinicensis series to the monophyletic E. martinicensis group of Hedges et al. (2008a) and, accordingly, remove the E. antillensis group from the E. martinicensis series and place it in its own species series. Similarly, the E. antillensis subgroup is paraphyletic with respect to the E. gryllus group, so we combine the species of these two subgroups into a single monophyletic E. antillensis subgroup. Also, the E. bakeri group of Euhyas is paraphyletic because E. glanduliferoides is the sister of E. jugans . We therefore transfer E. glanduliferoides to the E. jugans group to restore the monophyly of both the E. bakeri and E. jugans species groups.

The most problematic taxa are subgroups recognized within the E. luteolus species group within Euhyas, with all the groups we tested being para- or polyphyletic. Therefore, we reject these subgroups and refer species only to the more inclusive E. luteolus species group, which is endemic to Jamaica. Hedges et al. ' s (2008a) analyses did not resolve the relationships among subgroups within this clade, although the subgroups are partially recovered by protein variation, albumin immunology, and karyology and are all morphologically distinctive according to Hedges et al. (2008a).

Within Euhyas, Rodríguez et al. (2013) inferred new phylogenetic relationships for the E. limbatus group that partially conflict with Hedges et al. (2008a) and with our TA + PA and SA + PA results, although not with our SA + ML results. Their clade-Bayes tree (see Wheeler and Pickett, 2008) shows E. etheridgei as sister of all other species of the E. limbatus group and E. cubanus as sister of a clade containing E. orientalis as the sister group of a clade with two subclades, one with E. limbatus and E. jaumei as sister taxa, and another with E. iberia and an unnamed species as sister taxa. None of these conflicts require taxonomic changes.

Morphological evidence previously supported the monophyly of a mainly Antillean clade of Eleutherodactylus ( Lynch, 1971) , although some of its species were considered to be allied with mainland forms (e.g., Sminthillus [now Eleutherodactylus ] limbatus and Eleutherodactylus auriculatus ; Lynch & Duellman, 1997). Also, Lynch’s (1971) alpha division of Eleutherodactylini, based on the fusion of frontoparietals and proötic, grouped Syrrhophus , Tomodactylus (now in the synonymy of Syrrhophus ), Pelorius ( E. inoptatus ), and the subgenus Eleutherodactylus ( E. karlschmidti ), although Sminthillus (= Eleutherodactylus limbatus ) was considered distinct from the alpha and beta divisions. Heyer (1975) and Ardila-Robayo (1979) suggested a monophyletic group composed of Eleutherodactylus , Syrrhophus , and Tomodactylus based on morphological characteristics, although none of these are apparently synapomorphic.

The monophyly of Diasporus is corroborated in all our analyses. Hedges et al. (2008a) only sampled D. diastema and allocated seven other species to the genus based on the presence of a pointed tip of the ungual flap, an oval palmar tubercle, and prominent vomerine teeth. Although the pointed tip of the ungual flap characteristic of Diasporus is also found in Pristimantis chalceus ( Lynch, 2001) , that species groups with other species of Pristimantis . Our analyses recovered D. diastema as the sister taxon of D. quidditus and D. citrinobapheus . Hertz et al (2012) purported to test the monophyly of Diasporus and sampled 535 bp of 16S mtDNA for five nominal species, including a new taxon ( D. citrinobapheus ). However, their outgroup sampling consisted of only one species of Pristimantis ( P. ridens ), so the monophyly of Diasporus was an assumption of their analysis and not the result of a test. Populations referred to D. citrinobapheus were found to form the sister taxon of another clade composed of allopatric populations of D. aff. diastema from an area relatively near to the type locality (50 km airline distance) of D. citrinobapheus and showed divergences of 1.8%. We therefore assign those populations to D. citrinobapheus and, accordingly, the sequences from GenBank we analyzed (deposited by Crawford et al. 2010a) as D. aff. diastema should be reidentified as D. citrinobapheus . Nevertheless, the taxonomic status of D. citrinobapheus is problematic because no morphological character reported by Hertz et al. (2012) allows it to be distinguished from D. tigrillo ( Savage, 1997) and sequence data are currently not available for the latter species. Although populations assigned to these two species are separated by ca. 200 km, they both occur at the same elevations and in the same habitat. Hertz et al. (2012) also sampled D. diastema (paraphyletic in their analysis), D. hylaeformis (paraphyletic in their analysis), D. quidditus , and D. vocator . Pinto-Sánchez et al. (2012) included sequences of two additional species of Diasporus ( D. hylaeformis and D. vocator ) in their outgroup sample, but they were not available in time to be included in the present study.