Hypotheses for the evolution of dioecy in seed plants

Over the last decade, new hypotheses have been proposed for the evolution of dioecy in plants. Most of the selective mechanisms invoked have been suggested and supported by phylogenetic correlations. Here we review (1) the validity of the correlations (especially in light of recent critiques of the comparative method), and (2) the conformity of the proposed mechanisms to empirical data. None of the hypotheses can be flatly rejected on existing evidence, but the strength of their support varies. Future correlational studies must explicitly consider phylogeny; more importantly, such broad studies should also be supplemented by detailed studies of particular transitions to dioecy (e.g. within genera) - studies of the sort that have clarified analogous issues such as heterostyly.