| Past studies have revealed an apparent link between monoecy and dioecy in dimorphic angiosperm groups. Furthermore, comparisons of species richness between monoecious and dioecious sister groups have detected relatively low species richness amongst dioecious clades. Observations of low relative dioecious species richness has led to the hypothesis that dioecy is an evolutionary dead-end. To test the dead-end hypothesis, this study focused on the ancestral state reconstruction of breeding characters over nuclear and chloroplast phylogenies for the genus Amaranthus. Maximum parsimony (MP) and maximum-likelihood (ML) reconstructions were performed for both phylogenies. Continuous rate Markov modeling (Mk) and Binary Speciation and Extinction (BISSE) statistical analyses were applied to the ML reconstructions to test the rates forward and reverse transition, diversification, speciation, and extinction. The results of both the MP and ML analysis indicate numerous reversions from the derived dioecious to the ancestral monoecious state. Mk analyses of the ML reconstruction indicate that character state transition occur at much greater rates in the reverse direction. Thus, the hypothesis that dioecy is an irreversible, evolutionary dead-end, is not supported by the Amaranthus phylogeny. Rather, this study concludes that low species richness amongst dioecious clades is likely due to high rates of reverse character state transitions which bias the distribution in favor of monoecy. |
