Snake phylogeny: evidence from nuclear and mitochondrial genes.

We constructed phylogenies of snakes from the c-mos and cytochrome b genes using conventional phylogenetic methods as well as the relatively new method of Bayesian inference. For all methods, there was excellent congruence between the c-mos and cytochrome b genes, implying a high level of support for the shared clades. Our results agree with previous studies in two important respects: first, that the scolecophidians and alethinophidians are monophyletic sister clades; and second, that the Colubroidea is a monophyletic group with the Acrochordidae as its sister clade.

Is the PTP Test Useful?

We show empirically that the PTP test has very little discriminatory power, with highly significant PTP test probabilities often being associated with parsimony data that produce trees with low confidence (as measured by bootstrapping) and resolution. Because of this, we argue that the PTP test is useful only in the following, very limited way: if a data set fails the PTP test, it should not be used in a phylogenetic analysis. More conservative methods of measuring confidence such as the bootstrap or decay index are preferable.

ADAPTIVE RADIATION AND THE TOPOLOGY OF LARGE PHYLOGENIES.

The idea that some organisms possess adaptive features that make them more likely to speciate and/or less likely to go extinct than closely related groups, suggests that large phylogenetic trees should be unbalanced (more species should occur in the group possessing the adaptive features than in the sister group lacking such features). Several methods have been used to document this type of adaptive radiation. One problem with these attempts is that evolutionary biologists may overlook balanced phylogenies while focusing on a few impressively unbalanced ones.

COMPARISONS OF OBSERVED PHYLOGENETIC TOPOLOGIES WITH NULL EXPECTATIONS AMONG THREE MONOPHYLETIC LINEAGES.

Three null models have been proposed to predict the relative frequencies of topologies of phylogenetic trees. One null model assumes each distinguishable n-member tree is equally likely (proportional-to-distinguishable-arrangements model). A second model assumes that each topological type is equally likely (equiprobable model).