Modelling the effect of natural selection on the distribution of genotypes in environmental gradients (MODEGRAD)

Context and challenges

Environmental conditions fluctuate in time and space, which induces the maintenance of genetic diversity within populations through a double mechanism: on the one hand, spatial heterogeneity of selection pressures leads to different adaptive optima, and on the other hand, gene flow between subsets of the global population subjected to different selection pressures means that local optima are never reached. Spatial variations impose a variable adaptation regime within the stand. This spatial heterogeneity may favour the maintenance of genetic diversity in the stand as a whole, which can then be mobilised in adaptation to climate change (CC).

In forest trees, a correlation between values of environmental variables and allelic frequencies is regularly observed. This suggests that selection and adaptation processes have shaped these patterns, producing a non-homogeneous distribution of genotypes in space according to environmental conditions. These empirical results are consistent with theoretical predictions of the establishment of divergence between subpopulations by divergent selection even in the presence of gene flow. However, while these empirical studies have the merit of identifying processes of divergence between groups, they have not addressed the distribution of individual genotypes within stands. This task is complicated by the need to observe processes of potentially low intensity, which requires high analytical potential.

The accurate characterisation of this adaptive diversity in a population of interest requires the identification of the traits responsible for the adaptations and their genetic architectures. In natural populations of non-model organisms, this information is costly and slow to obtain. However, the study of molecular genetic diversity makes it possible to highlight the action of local selection, even at the fine scale of the landscape. Is it possible to predict the adaptive potential of the population from this information? The MODEGRAD project will address this question and the challenge of using this approach to predict the overall evolutionary trajectory of the population under climate change scenarios, as well as the impact of management on adaptation in the short term (one to a few generations) and on the maintenance of adaptive potential in the medium term.

Goals

MODEGRAD will establish a framework for modelling adaptive processes at the stand level and apply it to empirical genetic data from forest tree stands. Currently, population genetics is mainly a discipline focused on the study of the past: it is able to make inferences about the demographic history of populations and identify regions of the genome affected by selection events in the past. The prediction of future scenarios is what makes the MODEGRAD project highly original.

MODEGRAD will answer two fundamental questions:

• How does spatial and temporal variation in selection pressure, in interaction with dispersal abilities, shape the adaptation of a stand?
• How will adaptation be impacted by CC?

It will also answer one applied question:

• How can silvicultural management guide/facilitate adaptation processes in order to facilitate stand resilience?

MODEGRAD brings together researchers on the adaptation of wild stands and on the study of spatial-temporal dynamic processes. The complementarity of the two disciplines allows for original theoretical and methodological development.

INRAE units involved

• URFM, Avignon
• LESSEM, Grenoble
• CBGP, Montpellier
• ECOBIOP, Saint-Pée sur Nivelle
• BioSP, Avignon