Pollen and resource limitation of compensation to herbivory in scarlet gilia, Ipomopsis aggregata

This study was motivated by an interest in herbivore-pollinator interactions, and in the potential of pollination ecology and the physical environment to influence levels of tolerance to herbivore damage in scarlet gilia, Ipomopsis aggregata. To investigate these potential interactions, we performed a factorial combination of clipping treatments, fertilizer additions, and hand pollinations in a natural population of scarlet gilia in 1994 and 1995. Clipping imposed the strongest treatment effect on plants and acted by delaying phenology, altering plant architecture, and reducing plant fitness in both years. Hand pollinations increased the production of fruits and seeds in 1995 only, suggesting that our population was pollen limited in that year. Clipped plants could compensate for damage, but only under a restrictive set of environmental conditions, including fertilizer and hand pollinations in 1994, and hand pollinations in 1995. We did not detect significant overcompensation in either year of study, even under conditions releasing plants from nutrient and pollen limitation. A phenotypic selection analysis was conducted in 1995 to investigate several traits thought to influence fitness in the presence and absence of herbivore damage. Clipped plants were under strong selection for early flowering and increased plant height. We did not detect a significant association between branch production or plant size and fitness in grazed plants, suggesting that other factors determined the ability of plants to compensate for damage in 1995. We suggest that clipping-induced changes in plant architecture and phenology altered interactions between damaged plants and pollinators. We offer several hypotheses to explain these patterns and highlight the complexity of pollinator- and herbivore-mediated selection in natural plant populations.