There are some potential problems too with Moriarty and Cannatella's mtDNA data. They may have chosen a member of the ingroup as an outgroup. All of the species of Hyla found in the United States, according to immunological data, form a clade along with Acris, Pseudacris, L. ocularis and the Eurasian Hyla arborea species group. Taking both biogeography and morphology into consideration, it would appear that Hyla eximia is the ancestral species from which all other US species of Hyla evolved, given the fact that the bulk of its range is south of the border, the fact that the genus Hyla and the Hyla eximia species group is of neotropical origin, and the fact that practically all of the US species of Hyla plus Hyla arborea can be derived morphologically from an eximia-like ancestor. The "outgroup" chosen by Moriarty and Cannatella, namely Hyla chrysoscelis, Hyla andersonii and Hyla eximia, are all part of the same clade to which Pseudacris, H. regilla, H. crucifer, H. cadaverina and L. ocularis all belong. Their tree shows Hyla chrysoscelis, which is a derived member of this clade, to be the most basal! There is therefore the strong possibility that some of the apomorphs found in H. chrysoscelis have been mistaken as plesiomorphs for this clade.

Hence their analysis is far from the final word on the relationships among members within this group, especially since immunological data shows that there was a veritable adaptive radiation very early in the history of this group's existence, as Maxson's 1976 paper shows. Moriarty and Cannatella could have, and should have, chosen as outgroup another species that is more distantly related to the North American species of Hyla than H. chrysoscelis and Hyla andersonii. Smilisca baudini or Osteopilus septentrionalis, for example, may have been better choices as outgroup than Hyla andersonii and Hyla chrysoscelis, although I would have left Hyla eximia in the analysis, given the possibility that it is the ancestral species of all US Hyla and their closest relatives.

[ Reply To This Message ] [ Subscribe to this Thread ] [ Show Entire Thread ]