Extinction threshold in spatial stochastic logistic model: Space homogeneous case

Abstract

We consider the extinction regime in the spatial stochastic logistic model in ${\mathbb{R}}^d$ (a.k.a. Bolker--Pacala--Dieckmann--Law model of spatial populations) using the first-order perturbation beyond the mean-field equation. In space homogeneous case (i.e. when the density is non-spatial and the covariance is translation invariant), we show that the perturbation converges as time tends to infinity; that yields the first-order approximation for the stationary density. Next, we study the critical mortality---the smallest constant death rate which ensures the extinction of the population---as a function of the mean-field scaling parameter $\epsilon >0$. We find the leading term of the asymptotic expansion (as $\epsilon \to 0$) of the critical mortality which is apparently different for the cases $d\ge 3$, $d=2$, and $d=1$.

v2 - Proof of the former Theorem 3.1 has been essentially simplified by introducing an additional lemma