Demographic stochasticity and resource autocorrelation control biological invasions in heterogeneous landscapes

Abstract

Mounting theoretical evidence suggests that demographic stochasticity, environmental heterogeneity and biased movement of organisms individually a ect the dynamics of biological invasions and range expansions. Studies of species spread in heterogeneous landscapes have traditionally characterized invasion velocities as functions of the mean resource density throughout the landscape, thus neglecting higher-order moments of the spatial resource distribution. Here, we show theoretically that di erent spatial arrangements of resources lead to di erent spread velocities even if the mean resource density throughout the landscape is kept constant. Speci cally, we nd that increasing the resource autocorrelation length causes a reduction in the speed of species spread. e model shows that demographic stochasticity plays a key role in the slowdown, which is strengthened when individuals can actively move towards resources. We then experimentally corroborated the theoretically predicted reduction in propagation speed in microcosm experiments with the protist Euglena gracilis by comparing spread in landscapes with di erent resource autocorrelation lengths. Our work identi es the resource autocorrelation length as a key modulator and a simple measure of landscape susceptibility to biological invasions, which needs to be considered for predicting invasion dynamics within naturally heterogeneous environmental corridors.