Experimental gradients of soil fertility reveal shifting drivers of invertebrate herbivory in a grassland food web

Rebecca M. Prather, Department of Biological Science, Florida State University, Tallahassee, FL and Michael Kaspari, Department of Biology, University of Oklahoma, Norman, OK

Background/Question/Methods
Herbivory is an important population interaction regulated by many factors. Anthropogenic nutrient deposition is rapidly increasing and produces numerous interacting effects on grassland food webs, simultaneously changing soil and plant fertility, plant and arthropod communities, and ultimately, herbivory. Understanding how increased nutrient deposition will change these interactions is important for maintaining grassland communities in the future. We used a gradient of fertilization to track nutrient addition through a grassland food web, measuring changes to soil and plant nutrients, plant biomass and richness, and arthropod abundance. Our end goal was to quantify the amount and type of herbivory plants experience. We compared the response of the aforementioned variables to a nutrient pulse vs. press in a single year on two sets of plots. Using a multi-year experiment, we worked to uncover the mechanisms driving herbivory within and across fertilizer quantities and durations.
Results/Conclusions
Nutrient addition increased soil fertility 100-fold and generated 1.3-fold increases in herbivory. However, this herbivory increase weakened over time—from a 1-year pulse experiment to a 2-year press experiment—and shifted herbivory damage from sucking herbivores to chewing herbivores. Additionally, we saw decreases over time in the role that herbivore abundance plays in predicting herbivory levels (SEM, P = 0.36, Fisher C = 17.48). Overall, we found the rather paradoxical result that fertilization increased herbivore abundance but decreased herbivory (GLM, z = 2.88, P = 0.004). We demonstrate the rippling effects of changing soil fertility on the abundance and function of a prairie food web, predicting herbivore abundance and herbivory. This research fills knowledge gaps about the shifting response of food webs in the years immediately following nutrient addition. The mechanisms we uncover could help predict responses to long-term fertilization.