Front propagation of a sexual population with evolution of dispersion: a formal analysis

TL;DR

This paper provides a detailed mathematical analysis of how sexual populations with evolving dispersal traits spread over space, revealing precise long-term behavior and trait distribution at the invasion front.

Contribution

It offers explicit asymptotic descriptions of the invasion front and trait distribution for sexual populations, advancing understanding beyond previous models.

Findings

The invasion front accelerates with a rate proportional to t^{5/4}.

Trait distribution at the front develops specific spatial structures.

Numerical simulations confirm the theoretical asymptotic approximations.

Abstract

The adaptation of biological species to their environment depends on their traits. When various biological processes occur (survival, reproduction, migration, etc.), the trait distribution may change with respect to time and space. In the context of invasions, when considering the evolution of a heritable trait that encodes the dispersive ability of individuals, the trait distribution develops a particular spatial structure that leads to the acceleration of the front propagation. That phenomenon is known as spatial sorting. Many biological examples can be cited like the bush cricket in Britain, the cane toad invasion in Australia or the common myna one in South Africa. Adopting this framework, recent mathematical studies have led to highlight the influence of the reproductive mode on the front propagation. Asexual populations have been shown to spread with an asymptotic rate of t 3/2 in a minimal reactiondiffusion model, whereas the analogous rate for sexual populations is of t 5/4 (where t denotes the time). However, the precise description of the behaviour of the front propagation in the sexual case is still an open question. The aim of this paper is to give precise approximations for large times of its position, as well as some features of the local trait distribution at the front. To do so, we solve explicitly the asymptotic problem derived formally. Numerical simulations are shown to confirm these calculations.