Nitrogen limitation and trophic vs. abiotic influences on insect herbivores in a temperate grassland

Plant resources, predators, and abiotic conditions represent three major factors that potentially influence insect herbivore abundance in terrestrial ecosystems. In nitrogen (N)‐limited environments the potential for bottom‐up (plant resource) control is strong because plant quality may limit herbivore abundance. However, extremes in abiotic conditions, such as temperature and moisture, can mask such effects. I tested these hypotheses in an 8‐yr field experiment that measured responses of plants and grasshoppers (Orthoptera) to N addition and exclusion of bird predators in an N‐limited old‐field prairie in east‐central Minnesota. Plant biomass increased by 150–400%, and plant tissue N increased by 78% in response to N addition of 17 g·m⁻²·yr⁻¹. Total grasshopper density responded positively to N addition following warm years. However, multiple regressions suggested that grasshopper densities were related much more strongly to thermal conditions than to soil N. Bird exclusion yielded weak effects that varied over time and may have been influenced by compensatory responses of other grasshopper predators to bird exclosures. Grasshopper feeding guilds differed in the relationship between their in vivo dry‐matter digestibility (DMD) and plant tissue N, and this physiological difference explained their drastically different responses to N addition. Plant quality (in vivo DMD) increased with plant tissue N for mixed‐feeding grasshoppers, and accordingly, their density was positively correlated with soil N. Plant quality did not change with plant tissue N for grass feeders, and their densities were negatively correlated with soil N. Both guilds responded positively to warmer thermal conditions, but mixed‐feeder densities were negatively related to previous year's precipitation, and grass‐feeder densities were not affected by precipitation. These results support the hypothesis that bottom‐up influences of insect herbivores can be important in N‐limited systems but do not support the hypothesis that more productive environments necessarily support greater top‐down influences. Thermal conditions may interact with or eliminate bottom‐up effects. Furthermore, different guilds within the herbivore trophic level may be influenced differently by N addition, predators, and abiotic conditions. These results suggest that exploring the mechanisms of interaction between abiotic and trophic influences within components of food webs is likely to yield many new insights into the regulation of herbivore communities.