Populations and Evolution (q-bio.PE)

Abstract: The dynamics of species communities are typically modelled considering fixed parameters for species interactions. The problem of over-parameterization that ensues when considering large communities has been overcome by sampling species interactions from a probability distribution. However, species interactions are not fixed in time, but they can change on a time scale comparable to population dynamics. Here we investigate the impact of time-dependent species interactions using the generalized Lotka-Volterra model, which serves as a paradigmatic theoretical framework in several scientific fields. In this work we model species interactions as stochastic colored noise. Assuming a large number of species and a steady state, we obtain analytical predictions for the species abundance distribution, which matches well empirical observations. In particular, our results suggest the absence of extinctions, unlike scenarios with fixed species interactions.