Dados do Trabalho

TÍTULO

EVOLUTION OF SKULL MORPHOLOGY IN AKODONTINE RODENTS

Resumo

We explored the mode of evolution of cranial shape and size of the Akodontini, a group of highly diversified Sigmodontinae rodents, using branch patterns and diet to test different evolutionary scenarios. The shape of the skull was assessed by geometric morphometrics of the lateral view of the skull of 607 adult specimens belonging to 59 species, using the ‘geomorph’ package. To investigate the evolution of cranial size, the logarithm of centroid size was used. For shape, the first three axes of greatest variation of the PCA were selected, using the Broken Stick model. We used a published dated Sigmodontinae phylogenetic tree, pruning terminals to match our morphological sampling (approximately 70% of tribal diversity). The two data sets were submitted to a model fit test, using the ‘mvMORPH’ package. All analyses were performed in the R environment. We defined 17 models to be tested, using 100 replicates, evaluating the continuous character evolution processes Brownian Motion (BM), Ornstein-Uhlenbeck (OU) and Early Burst (EB). For the first two processes, we defined models with one to four evolutionary regimes (BM1, BMM2, BMM3, BMM4, OU1, OUM2, OUM3, OUM4) according to the four main akodontine clades, in order to test the influence of historical factors related to tree structure on the evolutionary dynamics of shape and size. We further established two models to evaluate the influence of diet on rates and evolutionary regimes (D1_BMM2, D1_OUM2, D2_BMM2, D2_OUM2, D3_BMM2, D3_OUM2, D4_BMM3, D4_OUM3), using different combination of three food categories (insectivory, herbivory, and omnivory). The model fit to the data was evaluated via AICc, taking into account the relative contribution of each model in relation to others (wAICc). The analysis of skull size revealed that the model that best fits our data is BMM4 (wAICc, 0.72 +/- 0.02), indicating that size evolved by a Brownian stochastic process, but via different regimes in the four main clades, suggesting a prevailing influence of historical factors in the distribution of evolutionary rates related to cranial size. The Brownian model is generally associated with a drift process, but may also reflect directional selection with frequent temporal oscillations. Analyzing cranial shape, we found that the best model is OU1 (wAICc, 0.62 +/- 0.006), suggesting that a single optimum may be responsible for rates of cranial shape evolution, which may indicate a stabilizing selection process or an evolutionary constraint. However, the second-best model, OUM2, was moderately supported (wAICc, 0.31 +/- 0.003), whereas other models had low relative support (<0.10). OUM2 would allow inferences similar to the evolutionary process that generated the patterns in question, but in this case with two optima, one for each of the two clades branching from the Akodontini ancestor. We document a decoupling of the evolutionary regimes for shape and size of the skull of akodontine rodents, as well as the greater influence of historical factors (tree structure) over the ecological factors (diet) evaluated here. 

Palavras-chave

macroevolution, comparative methods, shape, Sigmodontinae.

Financiamento

CAPES