Bayesian morphological clock versus parsimony: An insight into the relationships and dispersal events of postvacuum Cricetidae (Rodentia, Mammalia)

Establishing an evolutionary timescale is fundamental for tackling a great variety of topics in evolutionary biology, including the reconstruction of patterns of historical biogeography, coevolution and diversification. However, the tree of life is pruned by extinction and very generally molecular data cannot be gathered for extinct lineages. Methodological challenges have prevented until recently the application of tip-dating Bayesian approaches in morphology-based fossil-only datasets. Herein we present a morphological dataset for a group of cricetid rodents to which we apply a battery of methods fairly new in palaeontology that can be used by palaeontologists for the analysis of entirely extinct clades. We compare the tree topologies obtained by traditional parsimony, Bayesian dated and undated phylogenetic approaches and calculate stratigraphic congruence indices for each. Bayesian tip-dated clock methods seem to outperform parsimony in the case of our dataset, which includes highly homoplastic morphological characters. Regardless, all three topologies support the monophyly of Megacricetodontinae, Democricetodontinae and Cricetodontinae. Dispersal and speciation events inferred through Bayesian Binary Markov chain Monte Carlo and biodiversity analyses provide evidence for a correlation between biogeographic events, climatic changes and diversification in cricetids.