Does population distribution matter? Influence of a patchy versus continuous distribution on genetic patterns in a wind‐pollinated shrub

Abstract

Aim Uniform spatial population distributions are predicted to result in lower among-population genetic differentiation and higher within-population genetic diversity than naturally patchy distributions, but there have been surprisingly few attempts to isolate this effect from confounding factors. We studied the widespread wind-pollinated shrub Allocasuarina humilis that is common in a geologically stable landscape characterized by long-term population persistence to test the influence of semi-continuous versus patchy population distributions on genetic patterns. We also investigated whether A. humilis shows the high population connectedness and genetic diversity typically associated with wind pollination, a relatively uncommon and little-studied syndrome in this landscape.Location Heath-shrublands ('heath') and forests of south-western Australia.Methods Populations were sampled from heath and forest regions, which respectively exhibited semi-continuous and patchy population distributions. Genetic structure and diversity were assessed for 27 populations using eight nuclear microsatellite markers and three chloroplast regions. Phylogeographical history was examined using Bayesian phylogeny reconstruction, parsimony analysis and tests of expansion.Results High haplotype diversity is consistent with long-term population persistence across most of the species' range. Nuclear markers showed low overall population differentiation and no geographical structure over c. 900 km, reflecting extensive pollen dispersal. For both marker types, patchily distributed forest populations were substantially more differentiated with significantly lower within-population diversity than semi-continuous heath populations. Phylogeographical analysis revealed evidence for earlier colonization of heath than forest and recent expansion into wetter forests, consistent with progressive long-term climatic drying.Main conclusions High population connectedness and genetic diversity probably resulted from wind pollination in combination with dioecy and long life span. Patchy population distributions appear to have influenced genetic structure and diversity through lower pollen and seed dispersal, lower effective population sizes and greater genetic drift. Our approach illustrates the value of minimizing confounding variables by testing the effect of a variable ecological trait within a single species. Population code; dominant vegetation type; N, number of adults sampled; A r , mean allelic richness per locus, estimated using rarefaction with a minimum sample size of 11 individuals; N p , total number of private alleles; H o , observed heterozygosity; H e , unbiased expected heterozygosity; F IS , inbreeding coefficient; *F IS > 0 at P ≤ 0.05; means across loci are given with standard errors in parentheses.

Human footprint differentially impacts genetic connectivity of four wide‐ranging mammals in a fragmented landscape

“…This has been demonstrated in a few recent studies on plants and birds. Two comparative studies on plants find that semi‐continuously or continuously distributed populations have lower genetic differentiation than patchily distributed populations (Levy et al, ; Llorens et al, ). Robin, Vishnudas, Gupta, and Ramakrishnan () studied an entire community of montane birds of the Western Ghats in India and found, with a few exceptions, highest genetic differentiation in montane‐restricted species, while the widespread species showed no genetic differentiation.…”

“…Density of linear features offered high resistance to jungle cat gene flow, though the model fit was very poor. Density of linear features (highway, rivers and canals) has been shown to influence genetic connectivity of roe deer ( Capreolus capreolus , Coulon et al, ) and has also been shown to affect abundance (Beazley, Snaith, Mackinnon, & Colville, ; Boulanger & Stenhouse, ) and occurrence (Wasserman, Cushman, Wallin, & Hayden, ) of mammals, which in‐turn can impact genetic differentiation in a landscape (Llorens et al, ; Weckworth et al, ).…”

“…Landscape genetic analysis revealed that human footprint has an impact on connectivity of the study species in the landscape. LULC Coulon et al, 2006) and has also been shown to affect abundance (Beazley, Snaith, Mackinnon, & Colville, 2004;Boulanger & Stenhouse, 2014) and occurrence (Wasserman, Cushman, Wallin, & Hayden, 2012) of mammals, which in-turn can impact genetic differentiation in a landscape (Llorens et al, 2017;Weckworth et al, 2013).…”

Environmental variation shapes genetic variation in Bouteloua gracilis: Implications for restoration management of natural populations and cultivated varieties in the southwestern United States

“…This is reflected, in part, in the genetic analysis, where genetic variation due to differences among individuals within sampling sites across the Colorado Plateau was fairly high (PhiPT = 0.15). This level of within-population genetic diversity has been documented in other wind-pollinated species, even within populations that occupy fragmented habitat (Gray et al, 2014;Llorens et al, 2016). This same model could be bolstered with genetic and cytotypic patterns across environmental gradients to help delineate the most appropriate seed transfer zones.…”

“…Fragmented populations typically have few prospects for long-distance seed dispersal and may exhibit restricted gene flow (Anderson, 2003), though gene flow could be expected to remain high even in patchy environments due to long-distance wind dispersal of pollen (Llorens et al, 2016). Bouteloua gracilis is also evidently an autopolyploid across the study region, with diploid, tetraploid, and mixed cytotype sites distributed across the Colorado Plateau (Butterfield & Wood, 2015).…”