Breeding systems: Evolution of androdioecy

A primary goal of evolutionary biology is to understand the genetic and ecological forces controlling adaptive evolution. Because an individual's reproductive strategy has a direct effect on fitness, the forces controlling breeding system evolution are readily tractable both theoretically and empirically, making it an important model for the study of evolution. When breeding systems arise that appear to contradict the general assumptions of breeding system theory, they are worth examining in detail.

Androdioecy is a rare and unusual breeding system in which populations consist of separate male and hermaphrodite individuals. Functional androdioecy was at one time thought to be non-existent, due to the stringent and unlikely conditions required for its evolution and maintenance (D. Charlesworth 1984). Yet androdioecy and near-androdioecy have now been identified in several plant and animal species. These species belong to phylogenetically divergent groups (clam shrimp, barnacles, and several diverse plant families), indicating that androdioecy has evolved independently on several occasions. Because breeding system theory predicts that androdioecy is unlikely to evolve, the very existence of androdioecy represents a challenge to the assumptions that underlie this body of theory. Reconciliation of theory with empirical observations necessitates a thorough exploration of the conditions that may have allowed androdioecy to evolve in these diverse species.

The majority of flowering plants are hermaphroditic, which is thought to be the ancestral form (Darwin 1877, Lewis 1942, Bawa 1980, D. Charlesworth 1985). This is likely the reason why previous attempts to model the evolution of androdioecy have assumed that it arose from hermaphroditism after the invasion of a female-sterile mutant conferring maleness (Ross & Weir 1976, Charlesworth & Charlesworth 1978, Charlesworth 1984). For a male to invade a population of hermaphrodites via selection, the new mutant must achieve at least twice as many outcross fertilizations as hermaphrodites, and even more if hermaphrodites self-fertilize any of their ovules (Lloyd 1975, Charlesworth & Charlesworth 1978, Charlesworth 1984). These authors agreed that a new mutation eliminating an individual's female function was unlikely to double its pollen production, and even less likely to double its fertilization success. Therefore androdioecy was considered unlikely to evolve.

The goal of our work on androdioecy is to understand both the genetic changes that allowed androdioecy to arise, as well as the ecological forces which promoted the evolution of this unusual breeding system.

On the genetic front, we are studying the genetics of sex determination in the androdioecious species, Datisca glomerata, and its dioecious (males and females) sister species, D. cannabina (Wolf et al. 1997, 2000). We believe that androdioecy is derived from dioecy in this group, so hermaphrodites are essentially females with pollen production. Comparisons between the androdioecious and dioecious species should help us understand the genetic changes necessary for hermphrodites to arise. An interesting result of this work is the discovery of some heritable sex ratio element in the dioecious species, possibly meiotic drive, which causes female-biased sex ratios (Wolf et al. 2000). We are currently trying to determine if a sex ratio element could have played a role in the evolution of androdioecy.

On the ecological front, we are investigating the relative siring ability of males and hermaphrodites, to determine how males are able to gain sufficient fitness for coexistence with hermaphrodites. Further, we are investigating the possible role of pollen limitation on the evolution of androdioecy. The basic idea is that if females in the ancestral dioecious population were unable to obtain enough pollen to fertilize all of their ovules, their reproductive output is pollen limited, rather than resource limited. A female that allocates some reproductive resources to making a small amount of pollen (i.e. a hermaphrodite), should be able to fertilize some of its own seeds, and have higher overall fitness than ordinary females. Indeed, this can promote the evolution of androdioecy if hermaphrodites resemble females, having high seed production and low pollen fitness, and if the selfing rate is not too high (Wolf and Takebayashi 2004). The next step is to determine if pollen limitation is likely to occur in natural populations of Datisca or other androdioecious species.