Lyapunov exponents and the mathematics of invasion in oscillatory or chaotic populations (Q1903088)

Invasion in non-growing populations is a common denominator to many issues arising in ecology and genetics. The theoretical study of epidemics or community diversity, for instance, is usually faced with the problem of the invasion (defined as an initial increase in frequency), spread and persistence of a new, initially rare, disease or species. Also, similar questions, dealing with genotypes or strategies, have to be answered in the field of population genetics and life-history theory. In any case, the outline of the mathematical approach to the problem of invasion is the same: Starting from a dynamical model which describes the behavior of a non-growing resident population of a single type (with type we refer to the characteristics of the disease, species, genotype\dots, under consideration), one should seek to assess the long-run growth rate of an initially small population of an alternative type (e.g., a newly arrived disease or species, or a mutant genotype), taking account of a possible interaction between the two types. If there is no interaction, the appropriate growth rate can be computed disregarding the resident type. If the two types interact, then the growth rate of the possibly invading type depends on the resident type itself. When the resident type is at a stable equilibrium, the criterion of invasion is simple, as it is related to the eigenvalues of the Jacobian matrix that describes the linearized dynamics of the two interacting types around the equilibrium: invasion occurs if the absolute value of the dominant eigenvalue is greater than unity. There are many cases, however, for which the stable regime of the resident type is not an equilibrium. Well known are the persistent temporal fluctuations of lemmings, voles, and in general small rodent populations as well as of many insects. The analysis of these fluctuations reveals the existence of complex quasiperiodic or chaotic dynamics. Therefore, there is increasing evidence that the stable regime of many populations is better described by the more general concept of attractor, be it an equilibrium, a cycle, a quasiperiodic trajectory, or chaos. Thus, the problem of invasion must necessarily be stated in this broader context: What are the conditions of invasion of a resident type characterized by a generic attractor? The purpose of this paper is to give insights into how the eigenvalue criterion can be extended to the more general case of the invasion of a resident attractor and to illustrate the relevant results. Hereafter we dwell on the most general case where resident and alternative types interact. We will assume that the interaction involved is competition for common environmental resources and, as such, is best modelled through a deterministic density-dependent time-discrete mechanism. We will also discuss how the results can be extended to account for stochastic fluctuations in the environment.