Carphodactylidae

Members of the family now known as Carphodactylidae were first grouped together in Kluge (1967)’s reconfiguration of gecko systematics. He named a tribe, Carphodactylini, which included Carphodactylus, Nephrurus, and Phyllurus ("padless carphodactylins") as close relatives in a subgroup. The tribe also included Pseudothecadactylus and all New Zealand and New Caledonian geckos ("padded carphodactylins") in another subgroup. Carphodactylini was diagnosed by a large patch of preanal pores formed by multiple rows, though these pores were noted as vestigial in Nephrurus and absent in Phyllurus.[8]

Under Kluge’s classification scheme, Carphodactylini was the sister group to Diplodactylini, which contained other Australian legged geckos. Both tribes formed the subfamily Diplodactylinae within the family Pygopodidae, alongside the legless lizards of the subfamily Pygopodinae. "Padless carphodactylins" (particularly Carphodactylus) were considered the most primitive and generalized members of Carphodactylini, and by extension Diplodactylinae.[8] A morphological phylogenetic analysis by Bauer (1990) generally supported Kluge’s classification scheme, though the internal relationships of carphodactylin species were very different.[9]

However, the monophyly and structure of Carphodactylini (as defined by morphological traits) started to become unclear once molecular phylogenies were developed. King (1987) found that Nephrurus and Phyllurus apparently retained an ancestral chromosome structure, but that Carphodactylus, "padded carphodactylins", and several species of Oedura (a supposed diplodactylin) shared a derived structure, necessitating a major revision of Kluge’s structure.[10] However, there is much debate over whether chromosomes are useful for inferring gecko systematics.[9][11][12] Donnellan et al. (1999) estimated gecko relationships using C-mos, a nuclear gene conserved throughout squamates. Their results considered Carphodactylini to be a paraphyletic assemblage, with Rhacodactylus and Pseudothecadactylus incrementally closer to diplodactylins (including Oedura).[13]

Simplified results of Donnellan et al., 1999, focusing on Australasian geckos:[13]

Carphodactylidae, in its present-day usage, was codified by Han et al. (2004). They expanded on the study of Donnellan et al. (1999), and found that Carphodactylini and Diplodactylinae, as previously used, were polyphyletic. "Padless carphodactylins" formed a clade sister to pygopodines, while "padded carphodactylins" formed a grade of geckos basal to diplodactylins. This prompted several renames of major clades. The clade of "padless carphodactylins" was given the new family Carphodactylidae, while the family Diplodactylidae was erected to encompass traditional diplodactylins and "padded carphodactylins". The family Pygopodidae was basically restricted to the legless pygopodines, and subsequent studies would rename the Australasian radiation to Pygopodoidea or Diplodactyloidea.[6]

Simplified results of Han et al., 2004, using new names as defined in the study:[6]

Although placing carphodactylids sister to pygopodids was an unusual result, nearly all subsequent molecular phylogenies have supported it.[2][3] Shared morphological traits are scant, but the two families both have an unusually complex squamosal with its lower end wrapping around the rear branch of the parietal.[14] Oliver & Bauer (2011) looked into carphodactylid ingroup relationships in more depth using an analysis based on C-mos along with RAG1 (a nuclear gene) and ND2 (a mitochondrial gene). They had difficulty with resolving the base of Carphodactylidae, but strongly supported the erection of a new genus, Uvidicolus.[2] Skipwith et al. (2019) found a different result while running phylogenetic analyses on 4268 ultra-conserved elements (UCEs) in diplodactyloid geckos. Their analysis placed Orraya as the earliest-diverging carphodactylid, suggesting that the ancestral habit for the family may have been similar to the cryptic leaf-tailed genera (Orraya, Phyllurus, Saltuarius).[3]