Environmental change is multidimensional, with local anthropogenic stressors and global climate change interacting to differentially impact populations throughout a species’ geographic range. Within species, the spatial distribution of phenotypic variation and its causes (i.e., local adaptation or plasticity) will determine species’ adaptive capacity to respond to a changing environment. However, comparatively less is known about the spatial scale of adaptive differentiation among populations and how patterns of local adaptation might drive vulnerability to global change stressors. To test whether fine‐scale (2–12 km) mosaics of environmental stress can cause adaptive differentiation in a marine foundation species, eelgrass (Zostera marina), we conducted a three‐way reciprocal transplant experiment spanning the length of Tomales Bay, CA. Our results revealed strong home‐site advantage in growth and survival for all three populations. In subsequent common garden experiments and feeding assays, we showed that countergradients in temperature, light availability, and grazing pressure from an introduced herbivore contribute to differential performance among populations consistent with local adaptation. Our findings highlight how local‐scale mosaics in environmental stressors can increase phenotypic variation among neighboring populations, potentially increasing species resilience to future global change. More specifically, we identified a range‐center eelgrass population that is pre‐adapted to extremely warm temperatures similar to those experienced by low‐latitude range‐edge populations of eelgrass, demonstrating how reservoirs of heat‐tolerant phenotypes may already exist throughout a species range. Future work on predicting species resilience to global change should incorporate potential buffering effects of local‐scale population differentiation and promote a phenotypic management approach to species conservation.