Persistence in spatial multi-species food webs: The conflicting influences of isolated food web feasibility and spatial asynchrony

Abstract

A central goal of ecology is identifying the mechanisms that allow large, complex food webs to persist. Spatial mechanisms resulting from dispersal connections among local food webs are one factor shown to play a significant role in enabling species persistence, particularly by driving asynchrony in the dynamics among local food webs. However, it is still unknown how these spatial persistence mechanisms operate across food webs. Using simulations of full non-linear food web models, we investigate how spatial persistence mechanisms emerge in multi-species food webs that possess different structural metrics. Specifically, we ask whether 1) spatial persistence mechanisms work similarly across food webs, and 2) if differences can be explained by food web features influencing stability in the absence of dispersal, particularly trophic structure. Food web structures are generated using the allometric niche model that is capable of reproducing realistic feeding patterns and interaction strengths. Our analyses quantify the tendency of modeled food webs to achieve asynchrony in the presence of dispersal and show that this positively affects the ability of species in the food web to persist. We observe an inverse relationship between the ability of food webs to persist when isolated and their tendency to be asynchronous when spatial, indicating a limited ability of food webs that persist when isolated to benefit from spatial persistence mechanisms. Our results demonstrate a relatively unexplored layer of food web properties which determine the ability of a food web to capitalize on the stabilizing opportunities created by dispersal, specifically those that influence the tendency for dispersal-linked food webs to be asynchronous. Future studies should expand on our results by examining how properties of spatial connections and food webs influence the ability of food webs to achieve asynchrony.