Lambda-Fleming-Viot processes arising in logistic Bienaymé-Galton-Watson processes with a large carrying capacity

TL;DR

This work analyzes a family of logistic Bienaymé-Galton-Watson populations with neutral markers and large carrying capacity, uncovering three distinct scaling regimes dictated by the offspring-tail behavior. On the finite-variance tail, the population stabilizes near a carrying-capacity level and the neutral marker distribution converges to a Fleming-Viot diffusion with speed $1/N_e$, after time-rescaling by $[0,KT]$. In the α-stable regime, the population remains near $n_*K$ on $[0,K^{\alpha-1}T]$ and the neutral types converge to a $\Lambda$-Fleming-Viot process with a Beta-type coalescent, derived via stochastic averaging. At the critical Neveu case $\alpha=1$, the time scale shifts to $[0,T]$ with a $K\log K$ population scaling, and the neutral-marker dynamics converge to the dual of the Bolthausen-Sznitman coalescent, consistent with known results for Bolthausen-Sznitman-type genealogies. Across all regimes, the authors connect the forward-in-time population dynamics to well-known coalescent structures through a precise martingale-problem formulation and averaging principle, highlighting how carrying-capacity-driven fluctuations shape genealogies in interacting populations.

Abstract

We consider a continuous-time Bienaymé-Galton-Watson process with logistic competition in a regime of weak competition, or equivalently of a large carrying capacity. Individuals reproduce at random times independently of each other but die at a rate which increases with the population size. When individuals reproduce, they produce a random number of offspring, drawn according to some probability distribution on the natural integers. We keep track of the number of descendants of the initial individuals by adding neutral markers to the individuals, which are inherited by one's offspring. We then consider several scaling limits of the measure-valued process describing the distribution of neutral markers in the population, as well as the population size, when the competition parameter tends to zero. Three regimes emerge, depending on the tail of the offspring distribution. When the offspring distribution admits a second moment (actually a $ 2+δ$ moment for some positive $ δ$), the fluctuations of the population size around its carrying capacity are small and the neutral types asymptotically follow a Fleming-Viot process. When the offspring distribution has a power-law decay with exponent $ α\in (1,2) $, the population size remains most of the time close to its carrying capacity with some (short-lived) fluctuations, and the neutral types evolve in the limit according to a generalised $ Λ$-Fleming-Viot process. When the exponent $ α$ is equal to 1, the time scale of the fluctuations changes drastically, as well as the order of magnitude of the population size. In that case the limiting dynamics of the neutral markers is given by the dual of the Bolthausen-Sznitman coalescent.

Figures (1)

• Figure 1: Illustration of a realisation of a Cannings model with $N = 8$ individuals, run for 4 generations. The offspring numbers in the last step are indicated on the figure.

Theorems & Definitions (44)

• theorem 1: Theorem 4 in schweinsberg_coalescent_2003
• theorem 2
• remark 1
• theorem 3
• theorem 4
• lemma 1
• lemma 2
• proof : Proof of Lemma \ref{['lemma:rescaled_generator_rho']}
• lemma 3
• proof