Water availability drives signatures of local adaptation in whitebark pine (Pinus albicaulisEnglm.) across fine spatial scales of the Lake Tahoe Basin, USA

Abstract

lis ia 69 of 80 Lind et al: Local adaptation of P. albicaulis 2 Running Title: Local adaptation of P. albicaulis 19 20 ABSTRACT 38 For populations exhibiting high levels of gene flow, the genetic architecture of fitness-related 39 traits is expected to be polygenic and underlain by many small-effect loci that covary across a 40 network of linked genomic regions. For most coniferous taxa, studies describing this 41 architecture have been limited to single-locus approaches, possibly leaving the vast majority of 42 the underlying genetic architecture undescribed. Even so, molecular investigations rarely search 43 for patterns indicative of an underlying polygenic basis, despite prior expectations for this signal.44 Here, using a polygenic perspective, we employ single and multilocus analyses of genome-wide 45 data (n = 116,231 SNPs) to describe the genetic architecture of adaptation within whitebark pine 46 (Pinus albicaulis Engelm.) across the local extent of the environmentally heterogeneous Lake ୗ = 0.0069) there is 48 strong evidence for polygenic adaptation to the rain shadow experienced across the eastern 49 Sierra Nevada. Specifically, we find little evidence for large-effect loci and that the frequencies 50 of loci associated with 4/5 phenotypes (mean = 236 SNPs), 18 environmental variables (mean = 51 99 SNPs), and those detected through genetic differentiation (n = 110 SNPs) exhibit 52 significantly higher covariance than random SNPs. We also provide evidence that this 53 covariance tracks environmental measures related to soil water availability through subtle allele 54 frequency shifts across populations. Our results provide replicative support for theoretical 55 expectations and highlight advantages of a polygenic perspective, as unremarkable loci when 56 viewed from a single-locus perspective are noteworthy when viewed through a polygenic lens, 57 particularly when considering protective measures such as conservation guidelines and 58 restoration strategies. Lind et al: Local adaptation of P. albicaulis 5 empirical (Hall et al. 2007; Luquez et al. 2007) investigations exploring the relationship between 86 phenotypic differentiation (e.g., Q ST ) and that of the underlying loci (e.g, G ST or F ST ) have shown 87 that discordance between these two structural indices can occur under adaptive evolution. 88Moreover, as the number of underlying loci increases, the divergence between these indices 89 increases as well, and the contribution of F ST to any individual underlying locus decreases. In 90 cases that exhibit strong diversifying selection and high gene flow, this adaptive divergence 91 results from selection on segregating genetic variation (Hermisson and Pennings 2005; Barret 92 and Schluter 2008) and is attributable to the between-population component of linkage 93 disequilibrium (Ohta 1982, Latta 1998). In the short term, local adaptation will be realized 94 through subtle coordinated shifts of allele frequencies across populations causing covariance 95

Genetic evidence for species cohesion, substructure and hybrids in spruce

“…Our understanding of the evolutionary ecology of trees is also shaped by the evolutionary historical context in which comparisons are made. For example, many studies have evaluated local adaptation and the fit of plant phenotypes and genotypes to contemporary environments and climate variation, and have done so within an implicit or explicit framework for the relevant evolutionary history, including species boundaries, hybridization and gene flow (Alberto et al, ; Lind et al, ; Lindtke, Gompert, Lexer, & Buerkle, ; Yeaman et al, ). The evolutionary and historical context for studies of trait variation will determine what processes and dynamics are likely to account for the variation.…”

“…Our understanding of the evolutionary ecology of trees is also shaped by the evolutionary historical context in which comparisons are made. For example, many studies have evaluated local adaptation and the fit of plant phenotypes and genotypes to contemporary environments and climate variation, and have done so within an implicit or explicit framework for the relevant evolutionary history, including species boundaries, hybridization and gene flow (Alberto et al, 2013;Lind et al, 2017;Lindtke, Gompert, Lexer, & Buerkle, 2014;Yeaman et al, 2016). The evolutionary and historical context for studies of trait variation will determine what processes and dy- obtained through DNA sequencing means that some previously intractable relationships between groups can be reconstructed (e.g., Lamichhaney et al, 2015;Martin et al, 2015;McVay et al, 2017;Meier et al, 2017;Wagner et al, 2013) and that sampling of taxa and populations can more comprehensively assay genetic variation near the species level and tie it to evolutionary processes (e.g., Gompert et al, 2014;Mandeville, Parchman, McDonald, & Buerkle, 2015;Parchman, Buerkle, Soria-Carrasco, & Benkman, 2016;Zhou et al, 2017).…”

The role of hybridization during ecological divergence of southwestern white pine (Pinus strobiformis) and limber pine (P. flexilis)

“…The absence of elevated islands of divergence in this study, however, does not necessarily indicate an absence of adaptive divergence during speciation with gene flow. The lack of islands of divergence is expected in conifers, given the prevalence of polygenic architectures defining continuous trait variation across species boundaries and the expected prevalence of soft sweeps (Pritchard & Rienzo 2010; Alberto et al 2013; Rajora et al 2016; Lind et al 2017). Alternatively, given the large and complex genomes of conifers (reviewed by De La Torre et al 2014a) our ddRADseq markers underrepresented genic regions, which are often identified as islands of divergence (Nosil & Feder 2012; Zhou et al 2014; Moreno-Letelier & Barraclough 2015; Marques et al 2017).…”

Large‐effect loci affect survival in Tasmanian devils (Sarcophilus harrisii) infected with a transmissible cancer

“…We ran ≥5 million iterations following a 500,000‐iteration burn‐in, and we chose the linear BSLMM option except for the case–control phenotype, for which we chose the probit model. Similar to previous work (Lind et al., ), we used the posterior probability of a SNP having a large effect on a phenotype (after accounting for the effects of other SNPs in the genome) to identify candidate genes (see below) because we felt that this metric aptly captured a variant's contribution to that particular trait. SNP counts and sample sizes are given in Table .…”