Table 2.

Linear Models Testing Relationships between Two Measures of Directional (Positive) Selection and Genomic Diversity in Intergenic Regions (A) or Census Size, the Mating System, Ancestral Cluster, Distance to Cluster Core, and Admixture (B).

	FgenesposSel		α
Source	Estimate	Var. (%)	Estimate	Var. (%)
(A)
Intercept	0.0016∗∗∗		0.0217∗
π, intergenic regions	0.5226∗∗∗	28.04	70.8561∗∗∗	62.72
Depth of sequencing	1.9 ×10−5∗∗∗	17.08
(B)
Intercept	0.0018∗∗∗		0.0335
log10 (census size)	0.0000	0.81	0.0092	0.93
Mating system (selfing)	−0.0009∗	7.63	−0.1086∗∗∗	17.57
Ancestral cluster (west)	−0.0004	5.93	−0.0036	3.67
log10 (distance to core)	−0.0024∗∗	11.50	−0.3292∗∗∗	34.36
Admixture (yes)	0.0013∗∗	9.39	0.0837∗	2.36
Depth of sequencing	2.1 ×10−5∗∗∗	16.63

The response variables are the fraction of genes with a significant McDonald–Kreitman test (F_genesposSel) and the rate of adaptation estimated on non-synonymous versus synonymous sites (α). F_genesposSel was influenced by the depth of sequencing, which was therefore included as a covariate. The sample size was 52 populations. Variance explained by genomic diversity (R²) was 0.28 and 0.63, respectively, for the two variables; variance explained by the model with demographic factors was 0.52 (including depth of sequencing) and 0.59, respectively. Estimated intercept and coefficients, and variation explained are reported. Significance is indicated: ∗0.01 < P ≤ 0.05, ∗∗0.001 < P ≤ 0.01, ∗∗∗P ≤ 0.001.