The Haig-Westoby model revisited.

The Haig-Westoby model predicted that seed set in flowering plants would be equally limited by both pollen capture and resource supply because the optimal level of pollinator attraction should garner just the number of ovule fertilizations needed to consume the available seed-provisioning resources. Variability in the underlying resource and fertilization functions can disrupt this predicted optimum, a point made but only briefly explored by Haig and Westoby. Here I incorporate stochastic variation in both ovule fertilization and resource availability into the Haig-Westoby model and show that the modified model makes two noteworthy predictions: (1) pollen limitation of seed set (as measured by the response to supplemental pollen) should be common, and (2) the degree of pollen limitation may be greater when plants are more attractive to pollinators. The first prediction accords with recent meta-analyses of pollen limitation; the second remains to be examined.