Date of this Version
Insect herbivory is common, but the conditions under which it reduces the fitness and population size of plants remain poorly understood. We quantified populationlevel impacts of floral herbivory by specialized insects on cobweb thistle (Cirsium occidentale var. occidentale) in a California coastal dune ecosystem, and then compared these demographic effects to those published for Platte thistle (Cirsium canescens) in similarly designed exclusion experiments in the continental sand dune ecosystem in the Great Plains. As a separate test of the strength of the seed-to-seedling linkage, we quantified seedling establishment rates in seed addition plots and compared these results to those obtained in herbivore exclusion experiments. This is the first test of direct vs. indirect methods of evaluating the potential impact of seed loss in plant dynamics. Floral herbivory on cobweb thistle in coastal dunes substantially reduced key components of plant fitness. Reduction of insect feeding within flower heads with insecticide increased mean per capita seed production by 144–316% and led to a 130–196% increase in cumulative seedling recruitment in the next generation, depending on dune habitat. Juvenile plant densities around insecticide-treated plants subsequently increased by >50%. Both seed and flower head addition experiments corroborated the seed limitation of recruitment and juvenile plant establishment for cobweb thistle; exclusion of postdispersal seed predators did not substantially alter this relationship. While seed addition results were qualitatively similar to herbivore exclusion results, seed addition overestimated the seed–seedling transition. These results are broadly parallel to those found previously for Platte thistle in continental dunes, where reduction of insect feeding in Platte thistle flower heads increased mean per capita seed production by 37–240%, depending on year and microhabitat. This reduction in seed translated into a 33–580% increase in cumulative seedling recruitment and juvenile plant density. Thus in both systems, insect exclusions demonstrated that floral herbivory limited numbers of seeds and seedlings and population density of juvenile plants. This unique comparison of effects of insect herbivory on plant dynamics suggests that for related plants, the character and outcome of such interactions are likely to be functionally similar in structurally convergent ecosystems.