Rapid phenotypic evolution of quantitative traits can occur within years, but its underlying genetic architecture remains uncharacterized. Here we test the theoretical prediction that genes with intermediate pleiotropy drive adaptive evolution in nature. Through a resurrection experiment, we grew Arabidopsis thaliana accessions collected across an 8-year period in six micro-habitats representative of that local population. We then used genome-wide association mapping to identify the single-nucleotide polymorphisms (SNPs) associated with evolved and unevolved traits in each micro-habitat. Finally, we performed a selection scan by testing for temporal differentiation in these SNPs. Phenotypic evolution was consistent across micro-habitats, but its associated genetic bases were largely distinct. Adaptive evolutionary change was most strongly driven by a small number of quantitative trait loci (QTLs) with intermediate degrees of pleiotropy; this pleiotropy was synergistic with the per-trait effect size of the SNPs, increasing with the degree of pleiotropy. In addition, weak selection was detected for frequent micro-habitat-specific QTLs that shape single traits. In this population, A. thaliana probably responded to local warming and increased competition, in part mediated by central regulators of flowering time. This genetic architecture, which includes both synergistic pleiotropic QTLs and distinct QTLs within particular micro-habitats, enables rapid phenotypic evolution while still maintaining genetic variation in wild populations.

1. While competition is recognized as a major factor responsible for plant community dynamics, the genetics of intra-and interspecific competitive ability of a target species (i.e. level of intra-population genetic variation, identity of phenotypic traits under selection and genetic bases) still deserves a deeper investigation at the local spatial scale by considering both numerous genotypes and several interacting species. 2.In this study, we tested whether the genetics of competitive response and effect in Arabidopsis thaliana was dependent on the competitive environment at both the intraspecific and interspecific levels. We used a mapping population of 48 accessions (i) that maximize the genetic diversity of a local population of A. thaliana and (ii) that have been genotyped for 168 503 single nucleotide polymorphisms. 3. In a common garden experiment, those 48 accessions were grown in six competitive environments: the absence of competition, intraspecific competition and interspecific competition with four species frequently associated with A. thaliana in natural plant communities (i.e. Poa annua, Stellaria media, Trifolium repens and Veronica arvensis). A suite of nine phenotypic traits, including a proxy of fitness, were scored on each target A. thaliana plant and the aboveground dry biomass of its corresponding competitor was estimated. 4. We first showed that crossing reaction norms of competitive response (A. thaliana performance) and effect (competitor biomass) might promote maintenance of genetic variation in a local population of A. thaliana and species coexistence at a fine spatial scale. By estimating genotypic gradients of selection, we then demonstrated that the optimal phenotypic strategies in response to competition depend on the identity of the competitor species. Finally, a genomewide association mapping approach highlighted that genomic regions associated with direct genetic effects were (i) dependent on the competitor species and (ii) different from genomic regions associated with interspecific indirect genetic effects. 5. While a first step, this study highlighted the power of adding ecology to genomics in A. thaliana to identify genetic bases underlying micro-geographic adaptation to competition. Nextgeneration sequencing technologies will undoubtedly facilitate the discovery of molecular and genetic mechanisms underlying competitive ability in other plant species, and thereby the prediction of evolutionary trajectories of plant communities.

Life history strategies of most organisms are constrained by resource allocation patterns that follow a ‘slow-fast continuum’. It opposes slow growing and long-lived organisms with late investment in reproduction to those that grow faster, have earlier and larger reproductive effort and a short longevity. In plants, the Leaf Economics Spectrum (LES) depicts a leaf-level trade-off between the rate of carbon assimilation and leaf lifespan, as stressed in functional ecology from interspecific comparative studies. However, it is still unclear how the LES is connected to the slow-fast syndrome. Interspecific comparisons also impede a deep exploration of the linkage between LES variation and adaptation to climate. Here, we measured growth, morpho-physiological and life-history traits, at both the leaf and whole-plant levels, in 378 natural accessions of Arabidopsis thaliana . We found that the LES is tightly linked to variation in whole-plant functioning, and aligns with the slow-fast continuum. A genetic analysis further suggested that phenotypic differentiation results from the selection of different slow-fast strategies in contrasted climates. Slow growing and long-lived plants were preferentially found in cold and arid habitats while fast growing and short-lived ones in more favorable habitats. Our findings shed light on the role of the slow-fast continuum for plant adaptation to climate. More broadly, they encourage future studies to bridge functional ecology, genetics and evolutionary biology to improve our understanding of plant adaptation to environmental changes.