Understanding mechanisms that underlie species range limits is at the core of evolutionary ecology. Asymmetric gene flow between larger core populations and smaller edge populations can swamp local adaptation at the range edge and inhibit further range expansion. However, empirical tests of this theory are exceedingly rare. We tested the hypothesis that asymmetric gene flow can constrain local adaptation and thereby species’ range limits in an endemic US salamander (Ambystoma barbouri) by determining if gene flow is asymmetric between the core and peripheries of the species’ geographic distribution and testing whether local adaptation is swamped at range edges with a reciprocal transplant experiment. Using putatively neutral loci from populations across three core-to-edge transects that covered nearly the entire species’ geographic range, we found evidence for asymmetric, core-to-edge gene flow along western and northern transects, but not along a southern transect. Subsequently, the reciprocal transplant experiment suggested that northern and western edge populations are locally adapted despite experiencing asymmetric gene flow yet have lower fitness in their respective home regions than center populations do. Conversely, southern populations exhibit low deme quality, experiencing high mortality regardless of where they were reared, likely due to harsher edge habitat conditions. Consequently, we provide rare species-wide evidence that local adaptation can occur despite asymmetric gene flow, though migration from the core may prohibit range expansion by reducing fitness in edge populations. Further, our multi-transect study shows that multiple, non-mutually exclusive mechanisms can lead to range limits within a single species.

Data originate from field notes. Contained within are detailed measurements and observations from four transplantation cites in Kentucky and Ohio, US.

dTable1: Diversity and genetic measurements for each collection stream using a subset of invidividuals. N = number of individuals, AR = allelic richness, Ho = observed heterozygosity, He = expected heterozygosity, ID = geo identification, Fis = inbreeded coefficient (+ highest and lowest value).

dTable2: Details of each metamorph collected from the experiment. Region = transplantation regions (Center, South, North, West), Position = block position, Cont. ID = container ID, Mass (g) = dry metamorph mass, Date = date of metamorphosis, TL = total length of metamorph, SVL = snout-vent length of metamorph, Gape = gape of metamorph.

dTable3: Water and air logger data from each site. Each transplantation site has the water temperature, light intensity, and air temperature for the entirety of the experiment. 

dTable4: Randomized block experiment design. Mapping of blocks and experimental setup for each transplantation site (tabs).

dTable5: Raw field notes for each visit to transplantation sites. Date of visit, transplantation site, block, remaining salamanders in container, total dissolved solutes, water pH, number of Dead salamanders observed, picture of salamanders, Notes, mass of salamanders.

dTable6: Input data for survival analysis. Each row indicates a dead salamader, obs is the number of days until death, treatment = transplantation treatment (center in center, south in center, west in center, etc.).