Ecologists have recognized that animal populations are sensitive to climate changes, but the field has focused on shifts in species’ ranges and phenologies. Less work has been conducted studying climate effects on population dynamics and species interactions in real communities. Moreover, most research has focused on climatic means and ignored projected variances in temperature and precipitation. However, precipitation is expected to become more variable with increased frequency and severity of droughts and floods. We analyzed 30 years of survey data and found that precipitation positively correlated with years of high abundance of caterpillars of the ranchman’s tiger moth, Platyprepia virginalis. Interestingly, the number of large rainfall events (> 5 cm) was a better predictor than total annual accumulation. We considered three ecological mechanisms that could drive this relationship and conducted observations and manipulative experiments to evaluate these mechanisms: Large rainfall events 1) increased host plant abundance 2) increased leaf litter 3) decreased the abundance of predatory, ground nesting ants.
Results/Conclusions
We did not detect a relationship between precipitation or the number or large rainfall events on host plant abundance. A watering experiment showed that plants that received supplemental water produced more leaf litter than plants only receiving natural precipitation. Using multiple laboratory and field experiments, we found that litter is not used as food by caterpillars but provided a refuge from ant predation. Simulated flooding events did not decrease ant abundance in a mesocosm experiment. Our results demonstrated that large rainfall events affect the population dynamics of caterpillars by at least one mechanism: more rainfall increased the quality and depth of litter, decreasing the foraging efficiency of predatory ants. Precipitation along the California coast has been historically variable and almost all models predict that it will become more so. Variability in rainfall, particularly variability in large rainfall events, is likely to affect populations of insects, as exemplified in this study by P. virginalis. We found that both abiotic conditions (large precipitation events) and biotic players (predatory ants) were important determinants of the abundance of a common herbivore. Indeed, the interaction of abiotic conditions and biotic players exert strong effects that are likely to become even stronger in response to global change. A detailed understanding of the mechanisms that control insect populations will be required to better predict temporal and spatial dynamics.