UC Riverside

Trophic cascades occur when changes in an ecosystem’s top trophic level indirectly drive changes in biomass at lower levels. Cascades of varying magnitudes have now been documented in every ecosystem, but explaining this variation remains a fundamental challenge in ecology; predictions from mathematical models have received equivocal experimental support. This is in part due to the mismatch between the closed-system food chain models that underlie most cascade theory and the complex food webs they aim to abstract. In the following chapters I report the results of field and laboratory experiments, and computational and analytical models that extend cascade theory toward scenarios involving complex, adaptive and open ecological systems. In chapter two I combine models and experiments to demonstrate that feedbacks between animal movement and the foraging behavior of adaptive omnivores in food webs controls the strength and sign of emergent cascades in freshwater habitats. This study reveals that a deeper understanding of how spatial processes modify behavior, and how this scales up to communities, will be required to predict cascades. With my third chapter, I report results of simulations of complex in silico food web networks. By using allometric scaling rules derived from a global species dataset to parameterize models of energy flow, I demonstrate that basic features of network architecture influence the probability that cascades will emerge and the magnitudes of those cascades. In chapter four, I use analytical models to show that bottom-up and top-down cascades can interact to yield non-intuitive patterns in emergent community structure. Namely I demonstrate that under certain conditions, consumption of basal species by top predators (i.e., omnivory) can lead to a greater facilitation of basal species’ biomasses than if predators did not consume producers at all. This result contends with current theory, which expects cascades to be weaker in omnivory systems because of direct consumption of producers by predators. Overall, I demonstrate that cascades are strongly influenced by processes that operate on very different scales: from animal movement at the landscape scale, to community architecture at the network scale, gross diet at the population scale and behavior at the scale of the individual.