BERGMANN, P.J.*; RUSSELL, A.P.; University of Calgary; University of Calgary: Application of Bayesian inference in the phylogenetic analysis of multiple data sources: Intraspecific systematics of the turnip-tailed gecko.

Bayesian inference is used to implement a maximum likelihood approach to analysis of morphological, morphometric, and mtDNA sequence data obtained from samples of the neotropical gecko, Thecadactylus rapicauda. Bayesian analysis yields results congruent with those obtained using maximum parsimony analysis. However, Bayesian inference obtains more robust support than bootstrapping or Bremer values for observed clades using the posterior probability. Furthermore, it results in an analysis that is more easily interpreted, and less likely to mislead than maximum parsimony. With the current example, a previously assumed monotypic genus is found to consist of at least two species that are geographically distinct and supported, to varying degrees, by all data sets and analyses. Thecadactylus is distributed from southern Mexico, south to northern Bolivia and east to the mouth of the Amazon and on the Lesser Antilles. Those populations from western Amazonia are highly genetically divergent from the remainder of the nominate taxon. These populations are also distinguishable on the basis of morphological data, and, to a lesser degree, on morphometric grounds.