Colonization history of the Canary Islands endemic Lavatera acerifolia, (Malvaceae) unveiled with genotyping-by-sequencing data and niche modelling

Aim: Differentiation of island lineages from mainland relatives and radiation after initial colonization are two important processes generating island diversity. Both of these processes are influenced by colonization dynamics and contemporary connections between island species and their source populations. The classic island progression rule model, that is dispersal from mainland to older islands and subsequently to younger islands, can be considered a null hypothesis, particularly for islands that are geographically aligned perpendicular to the mainland coast with ages inversely proportional to distance from the mainland. Alterations to this pattern have been reported, particularly in archipelagos that are geographically closely adjacent to mainland source populations. Here we aim to integrate genomic and environmental niche data to infer the colonization history of a Canary endemic species and to understand its current diversity patterns. Location: Canary Islands. Taxon: Lavatera acerifolia (Malvaceae). Methods: We used high-throughput genotyping-by-sequencing (GBS) combined with species distribution modelling (SDM) projected onto past conditions. Genetic structure (clustering methods), relatedness (coalescent and ML trees), nucleotide diversity and differentiation (population genetics) were assessed based on SNPs obtained from three alternative bioinformatics pipelines. The influence of environmental variables over time was assessed with a generalized linear model in which the response variable was amount of heterozygous sites per individual. Results: Four genetic groups were identified arranged along a longitudinal gradient, and the earliest diverging coincides with the older, and easternmost, islands (Lanzarote and Fuerteventura). Genetic diversity is reduced in the westernmost islands, which are more distant from the mainland, host few populations and yet apparently offer more suitable habitats. Main conclusions: The inferred colonization scenario generally fits the progression rule model, but suggests a more complex pattern for the central islands. For the westernmost islands, the contrast between high availability of suitable habitats and reduced genetic diversity and number of populations suggests a colonization front moving at a slow pace, rather than local extinctions, as an explanation for the scarcity of populations in those islands. Historical projections of SDM lend support to this interpretation