The effect of "very fast" dispersal on two species competition with drift
Erin Ellefsen, Rana Parshad, Vaibhava Srivastava
TL;DR
The paper addresses how a faster dispersal mechanism, modeled by the $p$-Laplacian, competes with a slower species under downstream drift. It develops a degenerate two-species reaction-diffusion model combining regular diffusion and $p$-Laplacian diffusion, proves global existence of weak solutions for $rac{3}{2} < p < 2$ via regularization and compactness arguments, and identifies regimes where the faster diffuser may lose to the slower one, contrasting with the classical $p=2$ results. Numerical simulations corroborate the theory, revealing rich dynamics, including initial-data-dependent outcomes and partial-diffusion scenarios. The findings have ecological and applied relevance for biodiversity, refuge design, and biological control in fragmented or changing habitats.
Abstract
Classical theory predicts that for two competing populations subject to a constant downstream drift, the faster disperser will competitively exclude the slower disperser. In the current work, we consider a novel model of a "much faster" dispersing species, modeled via a $p$-Laplacian operator, competing with a slower disperser. We prove global existence of weak solutions to this model for any positive initial condition, in the regime $\frac{3}{2} < p <2$. Counterintuitively, we show that while the faster disperser always wins - the "much faster" disperser could actually lose, for certain initial data. Several numerical simulations are conducted to confirm our analytical findings. Our results have implications for biodiversity, refuge design, and improved biological control, driven by habitat fragmentation and climate change.
