SPATIO-TEMPORAL COLONIZATION PATTERNS IN EXPANDING TREE POPULATIONS: an integrated genetic and genomic approach – ExpandTree

Three complementary approaches are associated in EXPANDTREE: (i) reconstruction of chronosequences from the analysis of aerial pictures and field measurements, (ii) statistical analysis in landscape genetics and landscape genomics to exploit the spatio-temporal data concerning genetical markers and (iii) computer modeling of demographic and genetic interacting processes.
The methodological developments obtained in the project will enable the disentangling of processes effects from the observed patterns. They will also enable the predictions of future paths for the populations

First results in the EXPANDTREE project show that the studied tree populations were able to colonize rapidly (a few dozen years) free spaces at the scale of several kilometers. Long-distance dispersal events, by animals for Juniperus turbinata and by wind for Cedrus atlantica partly explain this rate of advance.
We developed new micro satellites for our two species (26 loci for Juniperus turbinata and 16 loci for Cedrus atlantica). We could then genotype several hundreds of individuals in each of our two experimental sites. The analysis of the genetic diversity and the importance of long-distance dispersal on the maintenance of diversity in the colonisation front is ongoing following an Approximate Bayesian Computation approach specifically developed for this type of data.
A set of 96 SNP markers have been selected for Cedrus atlantic and a comparable set of SNPs will be developed soon for Junipers turbinata.

The fine understanding of demographic and genetic processes ongoing in the spatio-temporal dynamics that span over a hundred years and a few kilometers is necessary to provide advices for a pragmatic management of tree stands that account for the challenges raised by global warming. The Mediterranean areas are particularly concerned since aridity will increase simultaneously with important land-use changes (recolonisation of free spaces following agriculture withdrawal, fragmentation of natural areas due to urbanization). Using demo-genetic models based on ecological processes to build with managers some pertinent guidelines for attenuating the effects of global changes is the option that EXANDTREE should make possible.

The first steps of EXPANDTREE focus on gathering ressources, data and new statistical methods. Besides the specific uses that will be made within EXPANDTREE, these ressources and methods will be useful in various contexts.In particular, this project provided ressources for population genetics studies on the two species of interest here: 26 microsatellites for Juniperus turbinata and 16 microsatellites pour Cedrus atlantica, a species for which marker development is notoriously difficult.
On the other hand, the computer programs that apply the statistical methods developed in EXPANDTREE are available for future uses on comparables scenarios: ABCCedrus enables the analysis of spatial genetic data resulting form a colonisation over o few number of populations; MEMM_Seedlings enables the analysis of genetic data from seedlings in a plot where all adults are georeferenced and genotyped.

Population responses to on-going climate changes can consist on: (i) allocating all ressources in surviving to new stressful conditions with a gradual loss of fecundity that eventually might lead to the population extinction ; (ii) colonize new suitable areas, thus expanding their current distribution ; or (iii) undergoing some evolutionary change allowing local populations to increase their overall fitness, (i.e., local adaptation). In fact, forest populations need to combine the three types of responses to overcome external changes, and we need to consider the interplay between demo-genetic-processes (involving demographic changes with genetic consequences such as increased fecundity variance) and evolutionary responses (involving adaptive changes across generations).
A sizeable proportion of Mediterranean forests are currently undergoing a demographic expansion or a displacement of their distribution area due to/in addition to climate change. In these populations, we hypothesize that the wave front of the expansion is a location where demographic and selective processes have strong and rapid effects on the evolution and structuration of genetic diversity. Our objective is to detect in genotypic data patterns indicating : (i) the role of long distance dispersal in favouring the mix of genotypes in the front wave ; (ii) demo-genetic processes in the front wave that would enhance local recruitment ; and (iii) adaptation of traits favouring survival in increasingly arid conditions.
We investigate these questions by studying two populations that have successfully regenerated and expanded in the latest decades from forest remnants (J. p. turbinata) or from a single founding population (C. atlantica) and where all trees are still present. This providing the unique opportunity to study in the same environment several cohorts of trees that were born in the last hundred year, the management time-scale at which we aim to understand micro-evolution.
Our project will combine three contrasted approaches (i) reconstruction of chronosequences of the expansion based on aerial pictures and field measurements that will provide information of how the expansion took place in space, time and ecological conditions (ii) landscape genetics and genomics where statistical analysis of the spatio-temporal patterns of genetic diversity (of both microsatellites and SNPs associated to resistance to drought) will inform us about the parameters of on-going demo-genetic processes and (iii) modelling the interplay between demographic and genetic factors in order to predict the fate of forest populations under different climate scenarios and for a variety of sets of demo-genetic parameters.
The ExpandTree project relies on the tight collaboration between the teams of C Garcia (CIBIO, Porto) and E Klein/S Oddou-Muratorio/F Lefèvre (INRA, Avignon) and takes benefit of the comparable situations of the two study sites identified for juniperus woodland (Doñana, Spain) and Cedrus (Mont-Ventoux, France). The participants will acquire comparable data and will conduct similar analyses. This will enable the comparison of the ecology of two species that inhabit two of the most threatened Mediterranean ecosystems : semi-arid coastal dunes (J. p. turbinata) and Mediterranean mountain forest (C. atlantica).