Figure 3
From: Molecular evolution and the latitudinal biodiversity gradient
Evidence of variation in the rate of molecular evolution has come from two main sampling strategies that use DNA sequences from extant species (usually represented by a single individual) to infer rate differences among lineages. Possible instances of punctuational evolution at nodes (grey boxes), extinct lineages (dashed lines), pruned or un-sampled lineages (X) are shown. (a) The sister species comparison. Sister species are chosen to represent contrasting distributions; one at low latitude and one at high latitude (temperate versus tropical as in Gillman et al., 2009). The placement of the outgroup in the phylogenetic reconstruction provides the crucial evidence for rate variation between the two ingroup lineages. By sampling only one individual per lineage, this method avoids the node-density effect that might influence estimates of molecular evolution rate (Hugall and Lee, 2007). Punctuated equilibrium (PE) predicts that acceleration in rates of molecular evolution is associated with speciation and therefore occurs at nodes. If so, observed differences in rates (different branch lengths) cannot result from PE unless speciation (PE2) and extinction have occurred only on the branch leading to the tropical species. Hidden (or incipient) speciation (PE3) could also explain the observed faster rate of molecular evolution if it was occurring more frequently in the tropical lineages than the temperate lineages. (b) Pruning large data sets. Species are selected from sister lineages with different species numbers to examine whether or not lineages with higher cladogenesis have faster rates (for example, Lanfear et al., 2010a). Relatively large data sets are pruned (X) to a pair of contrasting lineages, if more sequence data are available for one family than another, to avoid the node density effect. (c) Pruning large data sets with extinction. Lineage extinction (dashed lines) undermines confidence in inferences about the directionality of effects. Surviving species only provide estimates of net-diversification (speciation−extinction), not rates of speciation per se. Insufficient knowledge of extant lineages produces the same effect as extinction suggesting that this method would only work for well-studied groups.
