The effects of environmental stochasticity on population interaction webs

Interaction webs are networks of populations (or species) interlinked by trophic as well as non-trophic relationships. Very little is known about the effects of the abiotic environmental context on assembly and persistence of such webs. This study focuses on two environmental features that show distinct geographical (such as latitudinal) variation: energy availability (and potential primary productivity) and environmental stochasticity. While the former increases with decreasing latitude (towards the tropics), the latter decreases along the same gradient. To understand the joint effects of these two factors, we simulate web assembly under constant immigration-extinction dynamics with mathematical incorporating allometric life history scaling. The results provide interesting new insights and predictions about the effects of environmental conditions on dynamical and structural properties of interaction webs. We show that environmental fluctuations can select for component populations with certain growth rate and body size distributions, resulting in the assembly of persistent webs with characteristic dynamical and structural properties. We conclude that certain structural properties of interaction webs can be expected to be associated consistently with both potential productivity, and pattern of stochasticity in the external environment. These results are robust to a wide range of variation in model parameters and types of stochasticity, and also hold if related mathematical models are used. We test some of these predictions with empirical data from published studies across geographical gradients.