Dynamic Responses in a Plant-Insect System to Fertilization by Cormorant Feces

Theoretical arguments suggest that increased plant productivity may not only increase consumer densities but also their fluctuations. While increased consumer densities are commonly observed in fertilization experiments, experiments are seldom performed at a spatial and temporal scale where effects on population fluctuations may be observed. In this study we used a natural gradient in soil fertility caused by cormorant nesting. Cormorants feed on fish but defecate on their nesting islands. On these islands we studied soil nutrient availability, plant nutrient content and the density of Galerucella beetles, main herbivores feeding on Lythrum salicaria. In a common garden experiment, we followed larval development on fertilized plants and estimated larval stoichiometry. Soil nutrient availability varied among islands, and several cormorant islands had very high N and P soil content. Plant nutrient content, however, did not vary among islands, and there was no correlation between soil and plant nutrient contents. Beetle densities increased with plant nutrient content in the field study. However, there was either no effect on temporal fluctuations in beetle density or that temporal fluctuations decreased (at high P). In the common garden experiment, we found limited responses in either larval survival or pupal weights to fertilization. A possible mechanism for the limited effect of fertilization on density fluctuations may be that the distribution of L. salicaria on nesting islands was restricted to sites with a lower N and P content, presumably because high N loads are toxic.