Rapid adaptation follows experimental assisted gene flow in subset of annual monkeyflower populations

Assisted gene flow, the human-facilitated movement of species within its range, has the potential to increase genetic diversity and facilitate adaptation for climate-threatened populations. However, concerns regarding outbreeding depression and gene swamping have limited the application of assisted gene flow, and experimental tests have been largely confined to simulations and laboratories. We conducted a landscape-scale manipulation, planting seeds and seedlings from historically hotter and drier source populations into climate-threatened populations of the common yellow monkeyflower (Mimulus guttatus) and tracked change in genomes, phenotypes, and fitness. Within the first three generations, source alleles introgressed into half of the target populations. Assisted gene flow increased fitness in one experimental block of populations and slightly decreased fitness in a second block. Fitness increases were associated with flowering earlier and producing fewer trichomes. Greater fitness changes occurred in populations where seeds rather than seedlings were introduced. Both the amount of introgression and fitness changes were heterogeneous between years, suggesting that temporal fluctuations in climate likely impact initial introgression and success of assisted gene flow. Initial increases in ancestry from source populations was low across the genome, limiting concerns of gene swamping. Our results are generally consistent with theoretical models, and provide cautious optimism for an often-maligned conservation strategy. Significance StatementScientists must validate potential management strategies for climate-threatened species. Assisted gene flow - human-assisted movement of individuals within a species range - is a controversial strategy that theoretically results in increased fitness in threatened populations. In a landscape-level test of assisted gene flow in an annual monkeyflower, we find evidence of both increased and decreased fitness due to introgression from source populations following assisted gene flow. The amount of genome-wide introgression depended on the type of germplasm introduced and declined over time. The direction of fitness changes was likely based on the differences in ecologically-important phenotypes in source vs. target populations. Our results suggest that assisted gene flow performs largely as predicted from theoretical models and is a viable conservation strategy.