Molecular phylogeny and historical biogeography of the cosmopolitan parasitic wasp subfamily Doryctinae (Hymenoptera�:�Braconidae)

The recent development of Bayesian phylogenetic inference using Markov chain Monte Carlo (MCMC) techniques has facilitated the exploration of parameter-rich evolutionary models. At the same time, stochastic models have become more realistic (and complex) and have been extended to new types of data, such as morphology. Based on this foundation, we developed a Bayesian MCMC approach to the analysis of combined data sets and explored its utility in inferring relationships among gall wasps based on data from morphology and four genes (nuclear and mitochondrial, ribosomal and protein coding). Examined models range in complexity from those recognizing only a morphological and a molecular partition to those having complex substitution models with independent parameters for each gene. Bayesian MCMC analysis deals efficiently with complex models: convergence occurs faster and more predictably for complex models, mixing is adequate for all parameters even under very complex models, and the parameter update cycle is virtually unaffected by model partitioning across sites. Morphology contributed only 5% of the characters in the data set but nevertheless influenced the combined-data tree, supporting the utility of morphological data in multigene analyses. We used Bayesian criteria (Bayes factors) to show that process heterogeneity across data partitions is a significant model component, although not as important as among-site rate variation. More complex evolutionary models are associated with more topological uncertainty and less conflict between morphology and molecules. Bayes factors sometimes favor simpler models over considerably more parameter-rich models, but the best model overall is also the most complex and Bayes factors do not support exclusion of apparently weak parameters from this model. Thus, Bayes factors appear to be useful for selecting among complex models, but it is still unclear whether their use strikes a reasonable balance between model complexity and error in parameter estimates.

Rates of molecular evolution vary widely between lineages, but quantification of how rates change has proven difficult. Recently proposed estimation procedures have mainly adopted highly parametric approaches that model rate evolution explicitly. In this study, a semiparametric smoothing method is developed using penalized likelihood. A saturated model in which every lineage has a separate rate is combined with a roughness penalty that discourages rates from varying too much across a phylogeny. A data-driven cross-validation criterion is then used to determine an optimal level of smoothing. This criterion is based on an estimate of the average prediction error associated with pruning lineages from the tree. The methods are applied to three data sets of six genes across a sample of land plants. Optimally smoothed estimates of absolute rates entailed 2- to 10-fold variation across lineages.

The Southern Hemisphere has traditionally been considered as having a fundamentally vicariant history. The common trans-Pacific disjunctions are usually explained by the sequential breakup of the supercontinent Gondwana during the last 165 million years, causing successive division of an ancestral biota. However, recent biogeographic studies, based on molecular estimates and more accurate paleogeographic reconstructions, indicate that dispersal may have been more important than traditionally assumed. We examined the relative roles played by vicariance and dispersal in shaping Southern Hemisphere biotas by analyzing a large data set of 54 animal and 19 plant phylogenies, including marsupials, ratites, and southern beeches (1,393 terminals). Parsimony-based tree fitting in conjunction with permutation tests was used to examine to what extent Southern Hemisphere biogeographic patterns fit the breakup sequence of Gondwana and to identify concordant dispersal patterns. Consistent with other studies, the animal data are congruent with the geological sequence of Gondwana breakup: (Africa(New Zealand(southern South America, Australia))). Trans-Antarctic dispersal (Australia southern South America) is also significantly more frequent than any other dispersal event in animals, which may be explained by the long period of geological contact between Australia and South America via Antarctica. In contrast, the dominant pattern in plants, (southern South America(Australia, New Zealand)), is better explained by dispersal, particularly the prevalence of trans-Tasman dispersal between New Zealand and Australia. Our results also confirm the hybrid origin of the South American biota: there has been surprisingly little biotic exchange between the northern tropical and the southern temperate regions of South America, especially for animals.