Craugastorinae, Hedges, Duellman & Heinicke, 2008

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 : 51-53

Craugastorinae

Pyron & Wiens (2011) proposed Craugastorinae for Haddadus and Craugastor and placed their sister genus Strabomantis in a monotypic Strabomantinae (originally proposed to accommodate nearly 500 species in 10 genera; Hedges et al., 2008a). Although we recover the same relationships among the three genera found by Pyron & Wiens (2011), we reformulate Craugastorinae to include Strabomantis and treat the otherwise monotypic

6. We presume that those authors will soon provide a solution for this taxonomic problem. subfamily Strabomantinae as a junior synonym of Craugastorinae . Only this study and that by Pyron & Wiens (2011), the two studies of brachycephaloids with the largest taxon and character sampling, have recovered the monophyly of Craugastorinae . Hedges et al. (2008a) and Canedo & Haddad (2012) found Strabomantis to be associated with taxa now referred to Holoadeninae .

The relationship between Craugastor and Strabomantis should come as no surprise after a long history of putative relationship among species now placed in these genera. This Central American clade of terraranas with some South American components has long been recognized on the basis of morphology. Lynch's (1971, 1975a) Beta division of Eleutherodactylini (characterized by the frontoparietals and proötic not being fused and the pterygoids overlapping the parasphenoid alae) approximated what is now considered Craugastorinae . Also, Lynch (1986a) identified a condition of the m. levator mandibulae externus (muscle undivided, mandibular ramus of trigeminal nerve passing medial to muscle, the so-called “E” condition; Haas, 2001) that turned out to be a synapomorphy for Craugastor ( Crawford & Smith 2005; Heinicke et al., 2007; Hedges et al., 2008a; this study). Savage (1987) also suggested previously that the ca. 70 Central American species of the subgenus Craugastor lack the state of the m. depressor mandibulae present in all other subgenera of (then) Eleutherodactylus . Although Lynch (1993) disputed in some detail that claim, the group is maintained on the basis of molecular evidence, so it is worth revisiting the character-states proposed by Savage (1987) as they may also be synapomorphic.

Nevertheless, the subgenus Craugastor as it was then conceived is not identical to the current delimitation, containing at various times species now referred to Strabomantis . In fact, Strabomantis roughly coincides with what Lynch (1976a) considered the Eleutherodactylus biporcatus group, which was transferred to Craugastor by Crawford & Smith (2005). Several species of the broad-headed " Eleutherodactylus " in the C. bufoniformis and C. biporcatus species groups of the subgenus Craugastor were transferred to Limnophys by Heinicke et al. (2007) and later to Strabomantis , while others remained within Craugastor ( Hedges et al., 2008a). Interestingly, Heinicke et al. (2007, 2009) and Hedges et al. (2008a) discarded the possibility of a close relationship among the species they transferred to Strabomantis and Craugastor and, thus, rejected previous morphological evidence of evolutionary propinquity. In addition, the only derived morphological condition suggested by Hedges et al. (2008a) for Craugastoridae was the first finger longer than second, a condition that is shared by Craugastor , Haddadus , and Strabomantis . Furthermore, the “E” condition of the m. levator mandibulae has so far been only reported for Craugastor and Strabomantis . Within Strabomantis, Lynch (1986 a, 1993 ) recorded the "E" condition in S. anatipes , S. anomalus , S. biporcatus , S. bufoniformis , S. ingeri , S. necerus , S. ruizi , S. sernai , S. sulcatus , and S. zygodactylus , while he recorded the S condition for S. cerastes and S. cornutus —although Lynch (1997) synonymized under S. cerastes a species ( S. sernai ) that he diagnosed with the “E” condition ( Lynch, 1986a). Only the “S” condition (single muscle, mandibular ramus of trigeminal nerve passing lateral to muscle) has been reported for Haddadus binotatus (by Lynch, 1986a) as well as for H. aramunha ( Cassimiro et al., 2008) . Because the “E” condition has proved difficult to assess in some cases and some variants have been discovered (see comments on Yunganastes below), a reevaluation of the condition in the Craugastorinae is needed.

The dissonant note within Craugastorinae is Haddadus , a taxon historically considered to be related to what is now Ischnocnema , although that concept was based solely on biogeographic grounds and not synapomorphy. Nevertheless, Haddadus also presents an external morphology barely distinguishable from those of Strabomantis or broad-headed Craugastor . In fact, Strabomantis aramunha ( Cassimiro et al., 2008) from southeastern Brazil was placed in Strabomantis mainly on the basis of the presence of cranial crests, but is now placed in Haddadus on the basis of molecular evidence ( Amaro et al., 2013).

Craugastor .—The 60 Craugastor species we analyzed form a clade divided into the monophyletic subgenera Campbellius , Craugastor , and Hylactophryne 7. The internal relationships are largely congruent with those of Hedges et al. (2008a), who rearranged species diversity in putatively monophyletic species series. A detailed analysis of the correspondences between the groups and those of Savage (1987, 2002), Lynch & Duellman (1997), and Lynch (2000) was provided by Hedges et al. (2008a).

The strict consensus of optimal TA + PA trees collapses all structure within Hylactophryne . As such, we do not recognize any species series within this subgenus. Within the subgenus Craugastor , we recognize the C. mexicanus and C. podiciferus series. We found C. megacephalus (previously referred to the C. gulosus series) to be nested

7. We expect that Campbellius , Craugastor , and Hylactophryne will ultimately be considered genera, but because we cannot diagnose them morphologically, except by enumeration, we refrain from doing so here.

within the C. punctariolus series, making the C. punctariolus series paraphyletic and, therefore, merge both series into an expanded C. punctariolus series. The paraphyly of the C. punctariolus series was anticipated by Hedges et al. (2008a), who noted that Crawford & Smith (2005) had found C. megacephalus as the sister of a species of the C. punctariolus series, namely C. ranoides , and warned that the distinction between these two series might not hold.

We found Campbellius to be the monophyletic sister group of all other species of the genus Craugastor . Lynch (2000) hypothesized that C. greggi and C. daryi are sister species based on the shared fusion of the sacrum and the last presacral vertebra, but Hedges et al. (2008a) placed C. greggi in the Craugastor (Craugastor) laticeps species series. Craugastor (Campbellius) greggi has not yet been included in any molecular phylogenetic analysis.