Beyond isolation by distance: What best explains functional connectivity among populations of three sympatric plant species in an ancient terrestrial island system?

Abstract

Aim: Understanding how landscape features affect gene flow is critical to connectivity conservation and restoration management. Here, we examined the relationship between functional connectivity (gene flow) and structural connectivity (area and spatial configuration of habitats) in three co-occurring short-range plant taxa in an ancient terrestrial island system. Location: Low-altitude mountain range south-western Australia. Methods: We analysed spatial patterns of genetic differentiation at nuclear microsatellite loci using Bayesian clustering. Circuit theory modelling was used to generate all possible pathways that connect populations as resistance distance matrices based on two surfaces for each taxon. The first surface assumes a flat terrain and tests whether genetic similarity declines only with distance—isolation by distance (IBD). The second surface is habitat suitability based on species distribution modelling (SDM), which tests whether genetic similarity is a function of connected and suitable habitat. Multiple matrix regression with randomization was used to test the significance of the resistance distance matrices at predicting two metrics of genetic differentiation (FST and DEST). Variance explained was partitioned using redundancy analysis. Results: Genetic structure for the insect-pollinated taxa—Acacia adinophylla—and Tetratheca aphylla subsp. aphylla was at similar spatial scales. Unexpectedly, a higher level of genetic structure was found in the wind-pollinated Lepidosperma bungalbin. IBD best explained the gene flow of A. adinophylla (R2 = 0.41–0.43) with partial support provided by habitat suitability (R2 = 0.04–0.07). In contrast, connectivity by habitat suitability was highest for T. aphylla subsp. aphylla (R2 = 0.56–0.59). Drivers of L. bungalbin connectivity were inconsistent between the two measures of genetic differentiation. Main conclusions: Gene flow is facilitated by different factors for the three taxa. Habitat fragmentation would most strongly impede gene flow for T. aphylla subsp. aphylla. Geographical distance cannot be assumed as the sole or best determinant of gene flow among populations, nor can findings be generalized to coexisting taxa.