Evidence of parallel multi-dimensional phenotypic reorganization in Brassica incana on two Mediterranean islands

Abstract Background and Aims Island populations represent natural evolutionary experiments, often showing predictable divergence in life-history strategies (i.e., island syndrome). The multi-dimensional expression of this island syndrome in plants, spanning vegetative morphology, floral architecture, and seed provisioning, and its downstream consequences for offspring performance remains poorly characterized. Here, we aim to investigate whether island colonization reshapes phenotypic traits and reproductive investment strategies in insular Brassica incana populations occurring on two proximate Mediterranean islands. Methods Focusing on two mainland-insular population pairs across contrasting substrates (volcanic and limestone), we integrated genomic, phenotypic, and experimental approaches. We measured vegetative, floral, and reproductive traits across two flowering seasons, estimated standardized linear selection gradients to identify which phenotypic traits are under selection, and estimated the strength of phenotypic integration, i.e., the correlation among phenotypic traits, within each population. Finally, we tested the effect of seed mass on germination probability and early seedling performance in a controlled experiment. Key Results Insular populations showed moderate genomic differentiation from mainland populations and comparable levels of genetic diversity. Across both mainland-insular comparisons and both flowering seasons, insular populations showed larger leaf areas, larger corollas, a higher ratio of reproductive to aboveground vegetative biomass proxy, more seeds per fruit, and heavier seeds than mainland populations, while producing fewer leaves and fewer fruits, revealing different allocation trade-offs. Corolla size was under significant positive selection in insular populations. Seed mass did not predict germination probability, but was a strong positive predictor of seedling biomass under home-soil conditions. Phenotypic integration was overall low across all populations. Conclusions Island colonization in B. incana shows a consistent multi-dimensional syndrome across two independent colonization events and across years. This replicated demonstration expands our understanding of the island syndrome in plants, while revealing small and proximate Mediterranean islands as valuable, underexplored contexts for studying insular evolution.