EFFECTS OF FLORAL TRAITS, POLLINATOR VISITATION, AND PLANT SIZE ON IPOMOPSIS-AGGREGATA FRUIT PRODUCTION

Current understanding of reproduction for animal-pollinated plant species is based on the premise that traits of flowers and inflorescences influence reproductive success by affecting pollinator behavior. This implies that within-population variation in such traits should affect reproductive success. I used path analysis and structural equation modeling to test this hypothesis for two populations of the monocarpic herb Ipomopsis aggregata (Polemoniaceae). The basic hypothesis proposes that pollinator behavior may be affected by plant traits (including floral nectar production rate, corolla size, number of open flowers, and inflorescence height) and that behavior, along with plant size, may influence reproductive success through female function (proportion fruit set and total fruit production). This hypothesis was consistent with the data in both populations, although the specific details of the appropriate formulation varied. For both populations, the basic hypothesis is significantly better as a description of the interactions than alternative hypotheses that ignore the effects of plant traits on pollinator behavior or plant size on reproduction. In one population, disallowing the effects of pollinator behavior on reproduction significantly decreased the fit of model to data, and in the other, the decrease was almost significant (P = .06). Overall, the fundamental causal relationships proposed in the basic hypothesis were confirmed.