Figure 3.

Summaries of posterior distributions of (A) the difference in the variance in u _i versus the variance in a _i; (B) the variances in u _i (blue) and a _i (black); (C) the difference in the skew in u _i versus the skew in a _i; and (D) the skew in u _i (blue) and in a _i (black) across all individuals in each cohort during 1993–2018. Points show posterior means, with 95% credible intervals. Dashed black lines denote zero. In (B), the dotted black line denotes the direct base population estimate of V_A, adjusted for mean coefficient of inbreeding during 1993–2018 (V_A(1‐F), with F = 0.078). In (A) and (C), differences were calculated across pairwise posterior samples. Statistics shown above the plots are the posterior mean change across cohorts, with 95% credible interval (in parentheses) and posterior probability that the regression slope was positive (in square brackets). Statistics below the plots show the posterior probability that each cohort difference is negative.