Table 6.
Redundancy analyses (RDAs) to partition among‐population genetic variation (F) in Pinus strobus and P. monticola into three components: climate (IBE); geography (IBD); and north–south ancestry (IBC)
| Combined fractionsa | P. strobus | P. monticola | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| All (153) SNPs | Bayenv2 outlier (12) SNPsb | LFMM outlier (19) SNPsb | All (158) SNPs | Bayenv2 outlier (12) SNPsb | LFMM outlier (6) SNPsb | |||||||
| R 2 | p (>F)c | R 2 | p (>F)c | R 2 | p (>F)c | R 2 | p (>F)c | R 2 | p (>F)c | R 2 | p (>F)c | |
| F~clim. | .059 | .001*** | .295 | .001*** | .193 | .001*** | .089 | .001*** | .317 | .001*** | .109 | .002** |
| F~geog. | .064 | .001*** | .327 | .001*** | .198 | .001*** | .152 | .001*** | .400 | .001*** | .215 | .001*** |
| F~anc. | .045 | .001*** | .171 | .001*** | .139 | .001*** | .101 | .001*** | .386 | .001*** | .088 | .001*** |
| Individual fractions a | ||||||||||||
| F~clim. | (geog. + anc.) | .001 | .382 | .016 | .091● | .025 | .010** | −.006 | .722 | −.005 | .613 | −.048 | .962 |
| F~geog. | (clim. + anc.) | .007 | .023* | .034 | .002** | .026 | .001*** | .025 | .006** | .012 | .145 | .010 | .310 |
| F~anc. | (clim. + geog.) | .018 | .001*** | .005 | .136 | .021 | .002** | .021 | .001*** | .059 | .001*** | −.006 | .646 |
| F~clim.+geog. | anc. | .029 | – | .125 | – | .051 | – | .049 | – | .072 | – | .124 | – |
| F~geog.+anc. | clim. | −.001 | – | .012 | – | .001 | – | .034 | – | .077 | – | .060 | – |
| F~clim.+anc. | geog. | −.001 | – | −.002 | – | −.003 | – | .002 | – | .011 | – | .012 | – |
| F~clim. + anc. + geog. | .029 | – | .156 | – | .120 | – | .045 | – | .240 | – | .022 | – |
| Total explainedd | .084 | .348 | .244 | .176 | .471 | .228 | ||||||
| Total confoundedd | .058 | .293 | .172 | .130 | .400 | .218 | ||||||
| Total unexplained | .916 | .652 | .756 | .824 | .529 | .772 | ||||||
| Total | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | 1.000 | ||||||
aF = Independent matrix of population alleles frequencies; RDA tests are of the form: F~dependent matrices | covariate matrices. Clim. = climate (eight climatic variables); geog. = geography (P. monticola: x, y, xy, y 2; P. strobus: x, y, xy); anc. = north–south ancestry (Q‐values from STRUCTURE). Populations including five or more genotyped individual were used in this analysis.
bSubsets of SNPs detected by Bayenv2 (BF > 3) and by LFMM (q < 0.05). The number of SNPs for each subset is given in parentheses.
c●p < .10; *p < .05; **p < .01; ***p < .001. Significance of confounded fractions between climate, geography, and north–south ancestry was not tested.
dTotal explained = total adjusted R 2 of individual fractions. Total confounded = Total of individual fractions confounded between various combinations of climate, geography, and north–south ancestry. Negative R 2 values were considered null for this calculation.