Séminaire de Probabilités et Statistique :

Le 18 mars 2019 à 13:45 - UM - Bât 09 - Salle de conférence (1er étage)

Présentée par Lambert Amaury - Sorbonne Université

Ranked tree shapes, nonrandom extinctions and the loss of phylogenetic diversity

Phylogenetic diversity (PD) is a measure of the evolutionary legacy of a group of species, defined as the total length of the phylogenetic tree subtending these species. When the phylogeny is given by a Kingman coalescent, an important loss of species diversity leads to a much less important loss of PD (Nee & May Science 1997). However, PD loss is expected to strongly depend on the shape of the species tree, on the clade relative ages and on the order in which species are lost. Here, we propose a unifying framework to study these three different effects. Inspired by Aldous' beta-splitting model as well as self-similar fragmentations, we propose a new sampling-consistent, three-parameter model generating random trees with covarying shape, clade relative ages and clade relative abundances, the latter serving as proxy for robustness in the face of extinctions. This approach allows us to better understand the conditions in which species extinctions lead to a higher loss of PD. We propose a procedure to compute the ML estimates of the three parameters on empirical ranked phylogenies and apply it to phylogenies of bird families. The values of parameters inferred lie close to a 'danger zone' of parameter space where the loss of PD can go faster than the loss of species, i.e., where phylogenetically distinct species disappear first.