| Understanding whether a prey's risk of predation from individual predator species can be summed together to predict the predation risk when multiple predator species are together is crucial for predicting community dynamics. Observed predation risks can either match the predictions of a null model of additivity (additive risk) or vary from them (non-additive risk) in either a positive (risk enhancement) or negative direction (risk reduction). Additivity, risk enhancement, and risk reduction have all been documented in nature, but it is unclear how commonly each of these outcomes occurs within a large assemblage of predators because few studies have tested the assumption of additivity beyond one pair of predator species. To better understand how frequently the effects of multiple predator species are additive, I measured the risk of predation that mayfly larvae face from six different pairings of common stream predator species (represented by two insect and two fish species) and compared the observed risk to that predicted by a null model. Using controlled, replicated experiments in both artificial stream tanks and enclosures in a natural stream, I also evaluated whether the degree to which the observed risk differed from the expected risk was dependent on (1) the number of predator individuals or the number of predator species, (2) the identity of the prey species present, (3) the density of prey present, and (4) experimental venue. I found that observed risk levels were frequently reduced 15% or less compared to the model's predictions in the lab, despite the identity of the predator combination or prey present. The magnitude of the deviations from the model were similar in the presence of single and multiple predator species at the same predator density. In some cases, prey density influenced the level of risk reduction observed and the magnitude of risk reduction was greater in the field experiment than in lab experiments. By altering characteristics of the predators, prey, and experimental venue and evaluating their impacts on the outcome of multiple predator species interactions, my dissertation research provides a more complete picture of the realistic risk of predation prey face in the variable natural world. |
