Fig 1.
Nucleotide diversity was lower in inbred lines of all inbreeding groups than in outbred lines.
Boxplot of nucleotide diversity (π) of the experimental lines of the three expected inbreeding levels (Low, Medium, and High) and the outbred lines expressed here as relative to the mean of the outbred lines. Expected inbreeding coefficients (F) of the Low, Medium, and High inbreeding levels are 0.125, 0.219, and 0.381, respectively. Letters denote significant differences in means as determined by Welch’s t-test, i.e. the mean π of outbred lines was significantly higher than the mean π of inbred lines in all three inbreeding groups. Numbers in parenthesis (n) show the number of lines within each inbreeding level for which π could be obtained (total n = 119). SE is: 1.621, 1.363, 1.672, and 2.383 for OB, Low, Medium, and High, respectively.
Fig 2.
Low genomic variation was associated with elevated extinction risk.
Boxplot of nucleotide diversity (π) of the experimental lines that went extinct in either the first generation, the second generation, or in the third to the ninth generation, as well as for lines that did not go extinct, with π expressed relative to the mean π of the outbred lines. Numbers in parenthesis (n) show the number of lines in each group for which π could be obtained (total n = 119). Letters denote significant differences in mean π as determined by Welch’s t-test, i.e. the mean π of lines that went extinct during the experiment (generations 1–9) was significantly lower than the mean π of lines that did not go extinct.
Fig 3.
Genomic variation predicts evolutionary responses in productivity and body mass better than expected inbreeding coefficients.
Correlations between nucleotide diversity (π) (A-B) or expected inbreeding coefficient (F) (C-D) and slope of evolutionary responses for productivity (Slope of productivity; A and C) and body mass (Slope of body mass; B and D). Black solid lines represent linear regressions to visualize the relationships, and R2 values are shown followed by asterisks denoting significant correlations; * P<0.05, ** P<0.01, *** P<0.001. We only considered slope for lines that did not go extinct, to ensure that unreliable slope estimates generated from information across e.g. just 2 generations were not biasing the results. Further, only lines with available nucleotide diversity measures were included. In total this yielded 87 lines for computing the correlations. Dashed grey lines are shown at slope = 0 for comparison.
Fig 4.
Genomic variation predicts evolutionary change in egg-to-adult viability in the stressful environment.
Correlations between nucleotide diversity (π) (A-B) or expected inbreeding coefficient (F) (C-D) and changes in egg-to-adult viability (Δ). Δ was calculated as the difference in egg-to-adult viability measured at generation 10 (Gen 10) and generation 0 (Gen 0) on the stressful medium on which the specific line had been reared on throughout the experiment (A and C), and on the benign medium (B and D). Thus, a positive change represents lines where the viability is higher after compared to before the experiment at the respective mediums. Black solid lines represent linear regressions to visualize the relationships, and R2 values are shown followed by asterisks denoting significant correlations; * P<0.05, ** P<0.01, *** P<0.001. We only considered difference in viability for lines that did not go extinct, and lines with available nucleotide diversity measures; in total this yielded 87 lines for computing the correlations. Dashed grey lines are shown at slope = 0 for comparison.