Habitat structure changes the relationships between predator behavior, prey behavior, and prey survival rates.

The individual behavioral traits of predators and prey sometimes determine the outcome of their interactions. Here, we examine whether changes to habitat complexity alter the effects of predator and prey behavior on their survival rates. Specifically, we test whether behavioral traits (activity level, boldness, and perch height) measured in predators and prey or multivariate behavioral volumes best predict the survival rates of both trophic levels in staged mesocosms with contrasting structural complexity. Behavioral volumes and hypervolumes are a composite group-level behavioral diversity metric built from the individual-level behavioral traits we measured in predators and prey. We stocked mesocosms with a host plant and groups of cannibalistic predators (n = 5 mantises/mesocosm) and their prey (n = 15 katydids/mesocosm), and mesocosms varied in the presence/absence of additional non-living climbing structures. We found that mantis survival rates were unrelated to any behavioral metric considered here, but were higher in structurally complex mesocosms. Unexpectedly, katydids were more likely to survive when mantis groups occupied larger behavioral volumes, indicating that more behaviorally diverse predator groups are less lethal. Katydid mortality was also increased when both predators and prey exhibited higher average perch heights, but this effect was increased by the addition of supplemental structure. This is consistent with the expectation that structural complexity increases the effect of intraspecific behavioral variation on prey survival rates. Collectively, these results convey that the effects of predator and prey behavior on prey survival could depend highly on the environment in which they are evaluated.