AbstractLocal adaptation to broad-scale environmental heterogeneity can increase species’ distributions and diversification, but which environmental components commonly drive local adaptation— particularly the importance of biotic interactions—is unclear. Biotic interactions should drive local adaptation when they impose consistent divergent selection; if this is common we expect experiments to detect more frequent and stronger local adaptation when biotic interactions are left intact. We tested this hypothesis using a meta-analysis of common-garden experiments from 138 studies (149 taxa). Across studies, local adaptation was common and biotic interactions affected fitness. Nevertheless, local adaptation was neither more common nor stronger when biotic interactions were left intact, either between experimental treatments within studies (control vs. biotic interactions experimentally manipulated) or between studies that used natural vs. biotically-altered transplant environments. However, tropical studies, which comprised only 7% of our data, found strong local adaptation in intact environments but not when negative biotic interactions were ameliorated, suggesting that interactions frequently drive local adaptation in the tropics. Our results suggest that biotic interactions often fail to drive local adaptation even though they affect fitness, perhaps because the temperate-zone biotic environment is less predictable at the spatiotemporal scales required for local adaptation.
Common garden experiments to study local adaptation need to account for population structure
