Predicting the risk of extinction through hybridization

Natural hybridization threatens a substantial number of plant and animal sp ecies with extinction, but extinction risk bas been difficult to evaluate i n the absence of a quantitative assessment of risk factors. We investigated a number of ecological parameters likely to affect extinction risk, throug h an individual-based model simulating the life cycle of two hybridizing an nual plant species. All parameters tested, ranging from population size to variance in pollen-tube growth rates, affected extinction risk. The sensiti vity of each parameter varied dramatically across parameter sets, but, over all, the competitive ability, initial frequency, and selfing rate of the na tive taxon had the strongest effect on extinction. In addition, prezygotic reproductive barriers bad a stronger influence on extinction rates than did postzygotic barriers. A stable hybrid zone was possible only when habitat differentiation was included In the model. When there was no habitat differ entiation, either one of the parental species or the hybrids eventually dis placed the other two taxa. Tbe simulations demonstrated that hybridization is perhaps the most rapidly acting genetic threat to endangered species, wi th extinction often taking place in less than five generations. The simulat ion model was also applied to naturally hybridizing species pairs for which considerable genetic and ecological information is available. The predicti ons from these "worked examples" are in close agreement with observed outco mes and further suggest that an endemic coragrass species is threatened by hybridization. These simulations provide guidance concerning the kinds of d ata required to evaluate extinction risk and possible conservation strategi es.