Spatial isolation is a key driver of population-level variability in traits and genotypes worldwide. Geographical distance between populations typically increases isolation, but organisms face additional environmental barriers when dispersing between suitable habitat patches. Despite the predicted universal nature of the causes of isolation, global comparisons of isolation effects across taxa and geographic systems are few. We assessed the strength of isolation due to geographic and macroclimatic distance for paired marine island and paired mainland populations within the same species. Our meta-analysis included published measurements of phenotypic traits and neutral genetic diversity from 1832 populations of 112 plant and animal species at a global scale. As expected, phenotypic differentiation was higher between marine islands than between populations on the mainland, but spatial patterns of neutral genetic diversity did not vary between the two systems. Geographic distance had comparatively weak effects on the spatial patterns of phenotypes and neutral genetic diversity, but only phenotypic trait variability showed signal of system-dependence. These results suggest that spatial patterns of phenotypic variation are determined by system-dependent eco-evolutionary pressures, while the spatial variability of neutral genetic diversity might be universal. Our approach demonstrates that global biodiversity models that include island biology studies may progress our understanding of the interacting effects of spatial habitat structure, geographic- and environmental distances on biological processes underlying spatial population variability. We formulate future research directions for empirical tests and global syntheses in the field.