Splitting trees stopped when the first clock rings and Vervaat's transformation

TL;DR

This paper studies the genealogical tree of a branching population stopped at the first clock ringing, revealing a deep connection with Lévy processes and Vervaat's transformation, with applications in epidemiology and population genetics.

Contribution

It introduces a novel probabilistic identity linking the stopped genealogical tree to Lévy processes and Vervaat's transform, providing new insights and explicit distributions.

Findings

Conditional on the stopping event, the population size distribution is geometric.

A new representation of the genealogical tree using Lévy processes is established.

Explicit formulas for ages and residual lifetimes of individuals at stopping time.

Abstract

We consider a branching population where individuals have i.i.d.\ life lengths (not necessarily exponential) and constant birth rate. We let $N_t$ denote the population size at time $t$. %(called homogeneous, binary Crump--Mode--Jagers process). We further assume that all individuals, at birth time, are equipped with independent exponential clocks with parameter $\delta$. We are interested in the genealogical tree stopped at the first time $T$ when one of those clocks rings. This question has applications in epidemiology, in population genetics, in ecology and in queuing theory. We show that conditional on $\{T<\infty\}$, the joint law of $(N_T, T, X^{(T)})$, where $X^{(T)}$ is the jumping contour process of the tree truncated at time $T$, is equal to that of $(M, -I_M, Y_M')$ conditional on $\{M\not=0\}$, where : $M+1$ is the number of visits of 0, before some single independent exponential clock $\mathbf{e}$ with parameter $\delta$ rings, by some specified L{\'e}vy process $Y$ without negative jumps reflected below its supremum; $I_M$ is the infimum of the path $Y_M$ defined as $Y$ killed at its last 0 before $\mathbf{e}$; $Y_M'$ is the Vervaat transform of $Y_M$. This identity yields an explanation for the geometric distribution of $N_T$ \cite{K,T} and has numerous other applications. In particular, conditional on $\{N_T=n\}$, and also on $\{N_T=n, T<a\}$, the ages and residual lifetimes of the $n$ alive individuals at time $T$ are i.i.d.\ and independent of $n$. We provide explicit formulae for this distribution and give a more general application to outbreaks of antibiotic-resistant bacteria in the hospital.