Yes Moriarity, Cannatella and you differ from I in taxonomic practice. Specifically the three of you are following Hennig and are intolearant of paraphyletic taxa. But as I pointed out, whether or not one includes Hyla regilla, Hyla crucifer and Hyla cadaverina in Pseudacris, Hyla will be paraphyletic, as Duellman pointed out. Pseudacris is not paraphyletic even if these species are excluded. Although Hedges thought that these species must be included to maintain a holophyletic Pseudacris, and it was indeed the reason for his original taxonomic proposal, both da Silva and Moriarty and Cannatella's data suggest that Hyla regilla and Hyla cadaverina can be excluded without affecting the holophyly of Pseudacris. Therefore it boils down to whether one should include these species in Pseudacris because "most checklists and field guides" included these species, as Moriarty and Cannatella stated. As I have already stated, Hyla arborea is closely related to Hyla regilla Hyla eximia and Hyla cinerea, so close that if Hyla regilla is included in Pseudacris, Hyla arborea may need to be included to maintain the holophyly of Pseudacris. But if Hyla arborea is included, then Pseudacris is a junior synonym of Hyla since Hyla has priority over Pseudacris, and since Hyla arborea is the type species of Hyla. Further, why does one recognize Pseudacris in the first place? Pseudacris has been recognized traditionally because it is a monophyletic group and because it is a group of degenerate, secondarily terrestrial hylids with reduced toe pads. The inclusion of Hyla regilla and Hyla cadaverina in Pseudacris destroys this definition and makes Pseudacris undiagnosable. Hedges, in redefining Pseudacris, found that his "Pseudacris" are frogs with a "cold-weather breeding season, a round or ovoid testis, and a black pigment covering on the testis." As I pointed out, breeding season is not a taxonomic character and the morphology of the testis apply only to the males of these species. It turns out that Hedges' redefinition is unnecessary according to da Silva's data; Pseudacris can indeed be a holophyletic group even if Hyla regilla and Hyla cadaverina are excluded. So, why insist on including these species, if their inclusion creates a heterogeneous Pseudacris and at the same time opens the possibility that Pseudacris will become a junior synonym of Hyla? Moriarty and Cannatella appears to be attempting to maintain some semblance of taxonomic stability in their decision to leave Hyla regilla and Hyla cadaverina in Pseudacris. By doing so, however, Moriarty and Cannatella may have opened the door to the inclusion of Hyla arborea in Pseudacris; they may have unwittingly opened a can of taxonomic worms.

Your claim that immunological data has less resolving power than newer techniques is certainly true. mtDNA data does have a lot more resolving power. However, Moriarty and Cannatella's parsimony tree still shows an unresolved polytomy, suggesting that there may have been an adaptive radiation. Immunological data suggests the same in the early history of this group of holarctic Hyla. Besides, despite the higher resolving power of mtDNA, mtDNA can still be less informative if one is not careful about the choice of outgroup. Moriarty and Cannatella's choice of outgroup species (Hyla chrysoscelis) is analogous to choosing Pan as the "outgroup" when one is analyzing the relationships among Gorilla, Homo, and Pongo. They are in fact using a derived member of the ingroup as the "outgroup" species.

Moriarty and Cannatella's likelihood tree, as I said, requires a reversal. Either Hyla crucifer re-evolved large toe pads from an ancestor with reduced toe pads, which is unparsimonious, or as you suggest "multiple loss of toe pads." That means the species traditionally classified under Pseudacris independently lost their toe pads. This scenario is even less parsimonious since Pseudacris is a monophyletic group. It is true that multiple independent loss or reacquisition of a lost character are both possible, but they are less probable than the alternative hypothesis, especially since immunological data and da Silva's data both support the basal position of Hyla crucifer relative to Pseudacris, making the ad hoc hypothesis of multiple independent loss or reacquistion unnecessary.

In sum, there are at least four reasons for excluding Hyla regilla, Hyla cadaverina and Hyla crucifer from Pseudacris:

1. problems with Moriarty and Cannatella's analysis because of their unfortunate choice of outgroup species,

2. the availability of more than one set of data suggesting that Hyla regilla, Hyla cadaverina and Hyla crucifer can be excluded from Pseudacris without affecting its holophyly

3. inclusion of these species creates a heterogeneous Pseudacris

4. inclusion of these species may endanger the validity of Pseudacris because Hyla arborea (the type species of Hyla) is closely related to Hyla eximia, Hyla regilla, Hyla chrysoscelis et al.

[ Reply To This Message ] [ Subscribe to this Thread ] [ Hide Replies ]