Inter-individual variation in fecundities has major consequences on population evolutionary potential, through genetic drift and selection. Using two spatially explicit mating models that analyze the genotypes of seeds and seedlings, we investigated the variation of male and female fecundities within and among three European beech (Fagus sylvatica) stands situated along an elevational gradient. Female and male individual fecundity distributions were both skewed in this monoecious species, and we found a higher variance in female as compared to male fecundities. Both female and male fecundities increased with tree size and decreased with density and competition in the neighborhood, the details of these effects suggesting sex-specific strategies to deal with the impact of limited resource on fecundity. The studied populations were functionally male-biased. Among-individual variations in functional gender were not driven by tree size but by density and competition in the neighborhood, consistently with the expectation of a decreasing femaleness under limited resource availability due to higher cost of female reproduction. Considering the variation of gene flow and genetic drift across elevation, our results suggest that the adaptive potential could be enhanced by low genetic drift at low elevation, and by high pollen-mediated gene flow at high elevation. Finally, this study predicts a more efficient response to selection for traits related to male versus female fitness, for a given selection intensity.

Global change is generating widespread local-scale expansions of tree populations. During demographic expansions, even small differences in individual reproductive success can generate large differences in the genetic composition of the resulting population. Colonizing tree populations almost invariably show highly skewed distributions of seed production, but the evolution of this reproductive skew during plant recruitment remains understudied. We examine how recruit mortality modifies initial patterns of acorn dispersal and maternal reproductive success in a recently established, expanding Pedunculate oak (Quercus robur) forest stand and test different predictors of successful recruit establishment. We mapped and genotyped a cohort of seedlings (n = 809) emerging within and around the stand (n = 254 adult trees), identified their habitats and monitored their fate until an age of 3 years. Their mother trees were inferred by Bayesian parentage analysis, and the number of assigned descendants was used as a proxy for tree maternal fertility. We examined temporal trends in tree fertility and patterns of dispersal, and tested effects of seed number-quality trade-offs, dispersal failure, dispersal distance and habitat of propagule arrival on the probability of seedling establishment. We observed marked inequality in maternal reproductive success at the time of seedling emergence. A positive relationship between tree fertility and the proportion of non-dispersed seedlings pointed to disperser satiation. An overall seedling mortality of 69% generated considerable reshuffling in the fertility ranking of individual trees. Non-dispersal more than doubled seedling mortality. The establishment success of actually dispersed seedlings was independent of mother fertility and dispersal distance but strongly related to the habitat of arrival. Recruit survival was markedly higher in pine plantations or non-forested areas than in broadleaved forest. Synthesis. A few highly fertile trees dominated reproduction in the stand. Differential recruitment success tended to reduce their initial advantage, however, suggesting that reproductive inequality might rapidly decline as more trees start to reproduce along the expansion process. Our results also suggest that field estimates of size or seed production are unlikely to provide reliable estimates of lifetime reproductive success in trees.

A central issue in plant evolutionary ecology is to understand how several coordinated suites of traits (i.e. traits syndrome) may be jointly selected within a single species. This study aims to describe patterns of variation and co-variation of functional traits in a water-stressed tree population and test their relationships with performance traits. Within a Mediterranean population of Fagus sylvatica experiencing recurrent summer droughts, we investigated the phenotypic variation of leaf unfolding phenology, Leaf Area (LA), Leaf Mass per Area (LMA), Leaf Water Content (LWC), water use efficiency (WUE) estimated by carbon isotopic discrimination (d13C), twig Huber-value (HV: the stem cross-section divided by the leaf area distal to the stem), wood density (WDens), and leaf nitrogen content (Nmass). First, a Principal Component Analysis revealed that two main axes structured the phenotypic variability: the first axis opposed leaf unfolding earliness and LWC to LMA and WUE; the second axis opposed LA to HV. These two axes can be interpreted as the opposition of two strategies (water economy versus water uptake) at two distinct scales (leaf for the first axis and branches for the second axis). Second, we found that LMA, LA, leaf unfolding and LWC responded differently to competition intensity, while WUE, WDens and HV did not correlate with competition. Third, we found that all studied functional traits were related to growth and/or reproductive performance traits and that these relationships were frequently non-linear, showing strong interactions between traits. By highlighting phenotypic clustering of functional traits involved in response to water stress and by evidencing antagonistic selection favouring intermediate trait values as well as trait combinations, our study brought new insights on how natural selection operates on plant functional traits in a stressful environment.

Understanding local adaptation in forest trees is currently a key research and societal priority. Geographically and ecologically marginal populations provide ideal case studies, because environmental stress along with reduced gene flow can facilitate the establishment of locally adapted populations. We sampled European silver fir (Abies alba Mill.) trees in the French Mediterranean Alps, along the margin of its distribution range, from pairs of high and low elevation plots on four different mountains situated along a 170 km east-west transect. The analysis of 267 SNP loci from 175 candidate genes suggested a neutral pattern of east-west isolation-by-distance among mountain sites. FST outlier tests revealed 16 SNPs that showed patterns of divergent selection. Plot climate was characterized using both in-situ measurements and gridded data that revealed marked differences between and within mountains with different trends depending on the season. Association between allelic frequencies and bio-climatic variables revealed eight genes that contained candidate SNPs of which two were also detected using FST outlier methods. All SNPs were associated with winter drought and one of them showed strong evidence of selection with respect to elevation. QST - FST tests for fitness related traits measured in a common garden, suggested adaptive divergence for the date of bud flush and for growth rate. Overall, our results suggest a complex adaptive picture for A. alba in the southern French Alps where, during the east to west Holocene recolonization, locally advantageous genetic variants established both at the landscape and local scales.

One challenge of evolutionary ecology is to predict the rate and mechanisms of population adaptation to environmental variations. The variations in most life history traits are shaped both by individual genotypic and by environmental variation. Forest trees exhibit high levels of genetic diversity, large population sizes, and gene flow, and they also show a high level of plasticity for life history traits. We developed a new Physio-Demo-Genetics model (denoted PDG) coupling (i) a physiological module simulating individual tree responses to the environment; (ii) a demographic module simulating tree survival, reproduction, and pollen and seed dispersal; and (iii) a quantitative genetics module controlling the heritability of key life history traits. We used this model to investigate the plastic and genetic components of the variations in the timing of budburst (TBB) along an elevational gradient of Fagus sylvatica (the European beech). We used a repeated 5 years climatic sequence to show that five generations of natural selection were sufficient to develop nonmonotonic genetic differentiation in the TBB along the local climatic gradient but also that plastic variation among different elevations and years was higher than genetic variation. PDG complements theoretical models and provides testable predictions to understand the adaptive potential of tree populations.

Studies addressing the variation of mating system between plant populations rarely account for the variability of these parameters between individuals within populations, although this variability is often non-negligible. Here, we propose a new direct method based on paternity analyses (Mixed Effect Mating Model) to estimate individual migration (mi) and selfing rates (si) together with the pollen dispersal kernel. Using this method and the KINDIST approach, we investigated the variation of mating system parameters within and between three populations of Fagus sylvatica along an elevational gradient. Among the mother trees, si varied from 0% to 48%, mi varied from 12% to 86% and the effective number of pollen donors (Nepi) varied from 2 to 364. The mating patterns differed along the gradient, the top population showing higher m and lower s, and a trend to higher Nep than the bottom populations. The phenological lag shaped long-distance pollen flow both within population (by increasing mi at mother-tree level) and between populations (by increasing m at high elevation). Rather than the mate density, the canopy density was detected as a major mating system determinant within population; it acted as a barrier to pollen flow, decreasing the proportion of long-distance pollen flow and increasing si. Overall, the effects of ecological factors on mating system were not the same within vs. between populations across the gradient, and these factors also differed from those traditionally found to shape variation at range-wide scale, highlighting the interest of multiscale approaches.

Negative frequency dependent selection (NFDS) is supposed to be the main force controlling allele evolution at the gametophytic self-incompatibility locus (S-locus) in strictly outcrossing species. Genetic drift also influences S-allele evolution. In perennial sessile organisms, evolution of allelic frequencies over two generations is mainly shaped by individual fecundities and spatial processes. Using wild cherry populations between two successive generations, we tested whether S-alleles evolved following NFDS qualitative and quantitative predictions. We showed that allelic variation was negatively correlated with parental allelic frequency as expected under NFDS. However, NFDS predictions in finite population failed to predict more than half all S-allele quantitative evolution. We developed a spatially-explicit mating model which included the S-locus. We studied the effects of self-incompatibility and local drift within populations due to pollen dispersal in spatially distributed individuals, and variation in male fecundity on male mating success and allelic frequency evolution. Male mating success was negatively related to male allelic frequency as expected under NFDS. Spatial genetic structure combined with self-incompatibility resulted in higher effective pollen dispersal. Limited pollen dispersal in structured distributions of individuals and genotypes, non-random distribution of individuals and unequal pollen production significantly contributed to S-allele frequency evolution by creating local drift effects strong enough to counteract the NFDS effect on some alleles.