The relative importance of reproductive assurance and automatic selection as hypotheses for the evolution of self-fertilization

Abstract

Background: The field of plant mating-system evolution has long been interested in understanding why selfing evolves from outcrossing. Many possible mechanisms drive this evolutionary trend, but most research has focused upon the transmission advantage of selfing and its ability to provide reproductive assurance when cross-pollination is uncertain. We discuss the shared conceptual framework of these ideas and their empirical support that is emerging from tests of their predictions over the last 25 years.

Scope: These two hypotheses are derived from the same strategic framework. The transmission advantage hypothesis involves purely gene-level selection, with reproductive assurance involving an added component of individual-level selection. Support for both of these ideas has been garnered from population-genetic tests of their predictions. Studies in natural populations often show that selfing increases seed production, but it is not clear if this benefit is sufficient to favour the evolution of selfing, and the ecological agents limiting outcross pollen are often not identified. Pollen discounting appears to be highly variable and important in systems where selfing involves multiple floral adaptations, yet seed discounting has rarely been investigated. Although reproductive assurance appears likely as a leading factor facilitating the evolution of selfing, studies must account for both seed and pollen discounting to adequately test this hypothesis.

Conclusions: The transmission advantage and reproductive assurance ideas describe components of gene transmission that favour selfing. Future work should move beyond their dichotomous presentation and focus upon understanding whether selection through pollen, seed or both explains the spread of selfing-rate modifiers in plant populations.

Fig. 1.

The conceptual framework shared by two hypotheses for the evolution of selfing. Selfing will evolve by its gene-level transmission advantage if it is not erased by pollen discounting (p_x – p_s) or inbreeding depression (δ). If selfing increases seed production, selfing will be selected if the number of alleles passed on by plants capable of selfing (2y + x_s + p_s) is greater than the number transmitted by outcrossers (x_x + p_x), and exceeds a threshold inbreeding depression. Pollen and seed discounting (x_x – x_s) may both counter selection of selfing in this case.