Disentangling the influence of climatic and geological changes on species radiations

Abstract

Aim Our aim was to seek explanations for the differences in the diversity among the austral continents by comparing the diversification rates and patterns in the grass subfamily Danthonioideae. We asked specifically whether diversification is density dependent, whether it is different for each continent, and whether immigration rates impact on diversification rates. We attempted to account for intercontinental differences by comparing the Pleistocene climatic and Neogene geomorphological histories with the inferred diversification rates. Location Mainly the Southern Hemisphere, treated as four areas for the analyses: Africa, Australia, New Zealand and South America. Methods We based our analyses on a densely sampled, dated phylogeny for the grass subfamily Danthonioideae. We compared 24 diversification models for these continental radiations, taking into account speciation models, and extinction and dispersal rates. We used available distribution data to infer the climates under which danthonioids are found, and used these to estimate the change in area and location of suitable habitats between contemporary and Last Glacial Maximum climates. We inferred the geomorphological history from the literature. Results We show that long‐distance dispersal led to parallel radiations, which more than doubled the final standing diversity in the subfamily. Diversification models with the strongest support included separate time‐varying diversification processes for each major geographical region. Pleistocene climatic fluctuation did not explain the intercontinental differences in diversification patterns. Main conclusions Although our results are consistent with density‐dependent diversification, this explanation is not consistent with the time of arrival of danthonioids on each continent. The diversification patterns on the four major Southern Hemisphere landmasses are idiosyncratic. The two most important predictors of diversity may be the lineage‐specific effect of time, and the general effect of topographical complexity and orogenesis.