Experimental validation of genome-environment associations in Arabidopsis

Identifying the genetic basis of local adaptation is a key goal in evolutionary biology. Allele frequency clines along environmental gradients, known as genotype-environment associations (GEA), are often used to detect potential loci causing local adaptation, but GEA are rarely followed by experimental validation. Here, we tested loci identified in three different moisture-related GEA studies on Arabidopsis. We studied 44 GEA-identified genes using t-DNA knockout lines under drought and tested for effects on flowering time, an adaptive trait, and genotype-by-environment (GxE) interactions on performance and fitness. We found that wrky38 mutants had significant GxE effects for fitness; lsd1 plants had a significant GxE effect for flowering time, while 11 genes showed flowering time effects with no drought interaction. However, most GEA candidates did not exhibit GxE. In the follow-up experiments we found wrky38 caused decreased stomatal conductance and specific leaf area under drought, indicating potentially adaptive drought avoidance. Additionally, we found that GEA identified natural putative LoF variants of WRKY38 associated with dry environments, as well as alleles associated with variation in LSD1 expression. While only a few GEA putative drought-adapted genes were validated for GxE interactions for fitness under drought, we likely overlooked some genes because experiments might not well represent natural environments and t-DNA insertions might not well represent natural alleles. Nevertheless, GEAs apparently identified some genes contributing to local adaptation. GEA and follow-up experiments are straightforward to implement in model systems and thus demonstrate prospects for GEA discovery of new local adaptations.