Rencontre de la chaire
19 novembre 2026 matin
(Université Paris Cité, Paris).
Programme:
9h45 - 10h00 : Accueil café
10h00 - 10h40 : Patrick Hoscheit (INRAE) - Phylodynamique
de virus dans des paysages agricoles dynamiques
Les méthodes phylodynamiques
permettent de reconstruire la dynamique spatiale et temporelle
de virus à partir d’échantillons de séquences datées et
géolocalisées. Dans le cas particulier de pathogènes de
plantes, les paysages agricoles dans lesquels ils se propagent
sont à la fois inhomogènes et variables dans le temps. Je
m’appuierai sur l’exemple de la jaunisse nanisante de l’orge
(BYDV), un virus transmis par pucerons aux céréales et aux
graminées, pour présenter les différentes approches possibles
pour ce problème, notamment la phylogéographie continue
reposant sur des marches aléatoires, les processus de
Λ-Fleming-Viot spatial et les coalescents structurés. Je
parlerai des choix de modélisation inhérents à chaque cadre,
et leurs limites actuelles.
10h45 - 11h25 : Pierre Véron (ENS-PSL) - Modelling
speciation to identify microevolutionary processes that
explain differences in diversification between lineages
Speciation is the process by
which one species gives birth to two distinct species.
Phylogenetic estimates of speciation rates have revealed a
high heterogeneity across lineages. In this talk, I show how
the use of two mathematical models and empirical data helps
understanding the link between between microevolutionary
processes and macroevolutionary speciation rate. First, I use
a speciation model based on the prediction of genetic
divergence and polymorphism between populations under
different scenarios. This model predicts the duration of
speciation and and the shape of the grey zone, i.e., when at
which speed between-populations compatibility drops. By
varying parameters that describe species traits, evolutionary
processes, and speciation scenarios, I assess how these
factors shape speciation dynamics. I show that the link
between population size and speciation duration depends on the
speciation scenario, in particular that faster speciation
between smaller populations is typical of non-ecological
speciation. Comparing this expectation with empirical data on
several plants, I show that speciation occurs primarily
through a non-ecological scenario. Second I focus on the
influence of the different steps of speciation on the
macroevolutionary process. I calculate “equivalent” speciation
and extinction rates arising from the protracted birth-death
model. I show that the speed at which two distinct lineages
become new species has in general relatively little influence
on the equivalent speciation rate. For parameters estimated on
a family of Australian lizards, my analyses suggest that the
most limiting steps in speciation are the formation of
populations and their survival, rather than the speed of
speciation completion.
11h25 - 11h55 : pause
11h55 - 12h35 : Josquin Guerber
(MNHN) - Conjointly studying long-term
trends and temporal variability for a better
understanding of ecological dynamics
Monitoring
ecological communities over time reveals
long-term directional trends of population
abundances as well as variable population and
community abundances. However, variability is
better understood in stationary regimes, i.e.
without trends, while estimating trends from
biodiversity monitoring data often overlooks
variability around these trends. This
presentation deals with two cases where
acknowledging both trends and variability
builds a better understanding of ecological
dynamics. In one part, moment closure
approximations to stochastic population
dynamics models help quantify how the presence
of long-term trends causes variability to
deviate from stationary expectations. In a
second part, a Poisson-log-Normal model for
population abundance is used to estimate
population variability in large-scale
biodiversity monitoring data, allowing to
measure changes in population and community
variability of bird species monitored by the
French Breeding Bird Survey.
