A novel framework for assessing the power of genomic animal models in wild non-model organisms

ABSTRACT In highly dispersive taxa, gene flow can homogenise genetic variation over broad spatial and temporal scales. However, phenotypic differentiation can still arise when heritable genetic variation is available to selection and (or) when organisms exhibit plasticity. Disentangling heritable versus non-heritable components of variation is possible using genomic animal models, but this approach has been underused in wild organisms. We investigated head shape variation in the marine cryptobenthic fish Bathygobius cocosensis , a highly dispersive species occupying heterogeneous environments. Head shape, a diet-related trait, is spatially structured in Australian populations. We curated a unique dataset comprising genotypic, phenotypic, and environmental variation to: (1) contrast the scales over which genetic and phenotypic variation are structured; and (2) quantify heritable versus non-heritable contributions to head shape variation using genomic animal models and comparisons between F ST , P ST , and Q ST . At macrogeographic scales (hundreds of kilometres) and over time (2 years), phenotypic differentiation (0.0001 ≤ P ST ≤ 0.47) exceeded both genomic and additive genetic differentiation (0.001 ≤ F ST ≤ 0.004 and 0.00001 ≤ Q ST ≤ 0.004, respectively). This suggests head shape variation may primarily reflect (non-heritable) plastic responses, rather than (heritable) adaptive genetic changes in mean head shape. At microgeographic scales, associations between head shape and tide pool variables were temporally variable, indicating that local environmental conditions may structure phenotypes within populations. Overall, this study highlights the value of genomic animal models for disentangling heritable and non-heritable components of phenotypic variation in dispersive marine taxa. It provides an integrative analysis linking genetic, phenotypic, and environmental variation in a cryptobenthic fish, highlighting how phenotypic divergence can emerge without strong genetic differentiation, and challenging the common assumption that phenotypic divergence necessarily implies adaptation.