Gene flow from central to edge populations along an ecological gradient has been hypothesized to either cause range limits, by preventing adaptation through maladaptive gene swamping, or enhance the adaptive process at range limits, by introducing the genetic variation necessary to adapt. The role of gene flow was investigated using experimental breeding at the species low-elevation range limit in Mimulus laciniatus, an annual plant endemic to Sierra Nevada seeps between 900 – 3,270 m. Seeds were collected from three population sets (1 transect in Yosemite National Park and 2 transects in the Sierra National Forest), each having 7-12 populations spaced at roughly 300 m along the elevation gradient (24 populations total). Common gardens were established at the upper and lower elevation range limits and at a central elevation to test for elevation-based adaptive differentiation. Gene flow to two replicate low-edge populations was simulated by crossing sires from varying distances along and across the elevation gradient to dams from the target populations.  Within each target edge population dams received pollen from sires originating from 5 populations representing random, local breeding, counter-elevation-gradient breeding (far and near), and breeding along the same range limit (far and near). 
Results/Conclusions

Elevation-based, ecotypic variation was apparent throughout the species range, with low-elevation populations exhibiting increased fitness in warm, low-elevation environments. High-elevation populations had longer germination and development times. Gene flow enhanced the adaptive process at the species range limit.  Fitness (seed production) at the low-elevation range limit garden increased through random mating in general, and was further increased by breeding with outside populations. Breeding with the farthest populations along the same range limit resulted in highest fitness, suggesting that the most novel combinations of genetic variation, but from similar environments, are most beneficial.  Migrational meltdown (gene swamping) as a cause for range limits was rejected in this system, at least at the low-elevation range limit.  Similar experiments at the high-elevation limit are underway. Under rapidly warming environments, higher elevation populations may benefit from gene flow from low-elevation genotypes, while low-elevation inhabitants might be bolstered by gene flow from populations inhabiting similar climates. Trailing-edge populations of species ranges should be assessed and prioritized for their adaptive value under predicted climate change scenarios.