Angiosperm growth habit,dispersal and diversification reconsidered

Summary Previous studies have sought to elucidate the relationship between dispersal mode (biotic versus abiotic) and the taxonomic diversification of angiosperm families, but with ambiguous results. In this study, we propose the hypothesis that the combination of (1) the large seed size required of plants germinating in closed, light-poor environments and (2) the necessity to move disseminules away from the maternal plant in order to avoid intraspecific competition, predation and pathogens should favour biotically-dispersed relative to abiotically-dispersed woody arborescent angiosperms, resulting in higher diversification of the former. In this paper, we seek patterns of diversification that support this hypothesis. We examine the association between dispersal mode, growth habit and taxonomic richness of monocotyledon and dicotyledon families using (1) contingency table analyses to detect the effect of dispersal mode on the relative abundances and diversification of woody versus herbaceous taxa and (2) non-parametric analyses of variance to detect the statistical effect of dispersal mode on taxonomic diversification (mean number of species per genus, genera per family and species per family) in monocot and dicot families dominated by biotic or abiotic dispersal. We found a clear statistical effect of dispersal mode on diversification. Among families of woody dicots, dispersal by vertebrates is associated with significantly higher levels of species per genus, genera per family and species per family than is abiotic dispersal. The same pattern is observed among woody monocots, but is not significant at the 0.05 level. Among families of herbaceous monocots and dicots, the situation is reversed, with abiotically-dispersed families exhibiting higher levels of diversification than vertebrate-dispersed families. When woody and herbaceous families are pooled, there is no association between dispersal mode and diversification. These data coincide with evidence from the fossil record to suggest vertebrate dispersal has positively contributed to the diversification of woody angiosperms. We suggest that vertebrate dispersal may have promoted the diversity of extant taxa by reducing the probability of extinction over evolutionary time, rather than by elevating speciation rates. Our results suggest vertebrate dispersal has contributed to, but does not explainin toto, the diversity of living angiosperms.