Is The amphibian tree of life really fatally flawed?

Abstract

Wiens (2007, Q. Rev. Biol. 82, 55-56) recently published a severe critique of Frost et al.'s (2006, Bull. Am. Mus. Nat. Hist. 297, 1-370) monographic study of amphibian systematics, concluding that it is a disaster and recommending that readers simply ignore this study. Beyond the hyperbole, Wiens raised four general objections that he regarded as fatal flaws: (1) the sampling design was insufficient for the generic changes made and taxonomic changes were made without including all type species; (2) the nuclear gene most commonly used in amphibian phylogenetics, RAG-1, was not included, nor were the morphological characters that had justified the older taxonomy; (3) the analytical method employed is questionable because equally weighted parsimony assumes that all characters are evolving at equal rates; and (4) the results were at times clearly erroneous, as evidenced by the inferred non-monophyly of marsupial frogs. In this paper we respond to these criticisms. In brief: (1) the study of Frost et al. did not exist in a vacuum and we discussed our evidence and evidence previously obtained by others that documented the non-monophyletic taxa that we corrected. Beyond that, we agree that all type species should ideally be included, but inclusion of all potentially relevant type species is not feasible in a study of the magnitude of Frost et al. and we contend that this should not prevent progress in the formulation of phylogenetic hypotheses or their application outside of systematics. (2) Rhodopsin, a gene included by Frost et al. is the nuclear gene that is most commonly used in amphibian systematics, not RAG-1. Regardless, ignoring a study because of the absence of a single locus strikes us as unsound practice. With respect to previously hypothesized morphological synapomorphies, Frost et al. provided a lengthy review of the published evidence for all groups, and this was used to inform taxonomic decisions. We noted that confirming and reconciling all morphological transformation series published among previous studies needed to be done, and we included evidence from the only published data set at that time to explicitly code morphological characters (including a number of traditionally applied synapomorphies from adult morphology) across the bulk of the diversity of amphibians (Haas, 2003, Cladistics 19, 23-90). Moreover, the phylogenetic results of the Frost et al. study were largely consistent with previous morphological and molecular studies and where they differed, this was discussed with reference to the weight of evidence. (3) The claim that equally weighted parsimony assumes that all characters are evolving at equal rates has been shown to be false in both analytical and simulation studies. (4) The claimed strong support for marsupial frog monophyly is questionable. Several studies have also found marsupial frogs to be non-monophyletic. Wiens et al. (2005, Syst. Biol. 54, 719-748) recovered marsupial frogs as monophyletic, but that result was strongly supported only by Bayesian clade confidence values (which are known to overestimate support) and bootstrap support in his parsimony analysis was < 50%. Further, in a more recent parsimony analysis of an expanded data set that included RAG-1 and the three traditional morphological synapomorphies of marsupial frogs, Wiens et al. (2006, Am. Nat. 168, 579-596) also found them to be non-monophyletic.Although we attempted to apply the rule of monophyly to the naming of taxonomic groups, our phylogenetic results are largely consistent with conventional views even if not wth the taxonomy current at the time of our writing. Most of our taxonomic changes addressed examples of non-monophyly that had previously been known or suspected (e.g., the non-monophyly of traditional Hyperoliidae, Microhylidae, Hemiphractinae, Leptodactylidae, Phrynobatrachus, Ranidae, Rana, Bufo; and the placement of Brachycephalus within Eleutherodactylus, and Lineatriton within Pseudoeurycea), and it is troubling that Wiens and others, as evidenced by recent publications, continue to perpetuate recognition of non-monophyletic taxonomic groups that so profoundly misrepresent what is known about amphibian phylogeny. (C) The Willi Hennig Society 2007.