Table 1.
Results of the analysis of molecular variance (AMOVA). AMOVA was used to assess genetic divergence within treatments (i.e., among biological replicates in a given treatment) and between treatments (i.e., between the different selection environments)
| σ 2 | % | p value | |
|---|---|---|---|
| 22 °C | |||
| Within population (σ2WP) | 39.76 | 74.99 | <0.05 |
| Among population (σ2AP) | 13.26 | 25.01 | <0.05 |
| Total (σ2T) | 53.02 | 100 | |
| 26 °C | |||
| Within population (σ2WP) | 31.07 | 52.29 | <0.01 |
| Among population (σ2AP) | 28.35 | 47.71 | <0.01 |
| Total (σ2T) | 59.42 | 100 | |
| 32 °C | |||
| Within population (σ2WP) | 43.66 | 86.51 | <0.05 |
| Among population (σ2AP) | 6.81 | 13.49 | 0.051 |
| Total (σ2T) | 50.47 | 100 | |
| FS | |||
| Within population (σ2WP) | 31.03 | 50.04 | <0.01 |
| Among population (σ2AP) | 30.98 | 49.96 | <0.01 |
| Total (σ2T) | 62.01 | 100 | |
| All treatments | |||
| Within population (σ2WP) | 28.97 | 49.69 | <0.05 |
| Among populations between treatments (σ2AP/BT) | 8.82 | 15.12 | <0.05 |
| Between treatments ( σ 2 BT ) | 20.522 | 35.19 | <0.05 |
| Total (σ2T) | 58.32 | 100 |
AMOVA revealed that a significant proportion of the variance in SNV frequencies at specific loci was attributable to differences between treatments. Within individual treatments, there was significant variation between biological replicates (“populations” in the table), with the largest amount of variation among populations, (“AP”), in the 26 °C and FS treatments. σ2 represents the variance of each hierarchical level (i.e., within population “WP”, among populations “AP” and between treatments “BT”), and the % gives the percentage of this variance calculated to a total of 100%. Significance testing was carried out through random permutation of the samples following the methods outlined in ref. 59