Figure Legends
Figure 1. Map of observation locations for mitochondrial (A: π, B: Hd ) and nuclear (C: He ) genetic diversity. Populations were binned into 3° latitude x 3° longitude hexagons and the mean genetic diversity within each hexagon is plotted. Rug plots on the x- and y-axes illustrate the latitudinal and longitudinal sampling locations.
Figure 2. Relationship between absolute latitude and genetic diversity (A: mitochondrial π; B: mitochondrialHd ; C: nuclear microsatelliteHe ). Grey line represents the predicted relationship based on the mixed effects model with shaded 95% confidence intervals. Green violin plots show the distribution of genetic diversity binned every 10°, with the dark green points representing the medians in every 10° band.
Figure 3. Relationship between longitude and genetic diversity (A: mitochondrial π; B: mitochondrial Hd ; C: nuclear microsatellite He ). Gray line represents the predicted relationship based on the mixed effects model with shaded 95% confidence intervals. Blue circles represent median diversity binned every 10° with median average deviation (MAD) error bars.
Figure 4. A-C: Relationship between mean sea surface temperature (SST) (A, B, C) or mean chlorophyll (D, E, F) and genetic diversity (A, D: mitochondrial π; B, E: mitochondrialHd ; C, F: nuclear microsatelliteHe ). Black line represents the predicted relationship based on the mixed effects model with shaded 95% confidence intervals. Rug plots on the x-axis illustrate the SST or chlorophyll sampling extent. Mean chlorophyll is plotted on a common logarithm scale.
Table 1. Mitochondrial DNA (π and Hd ) model results for latitude and longitude. Model coefficients are reported, along with ΔAIC compared to the null model (model AIC - null AIC). For the null models, AIC is also reported in parentheses. The top model (with the lowest AIC) is bolded, while the null model is highlighted in grey. For longitude, the b-spline basis function coefficients are reported (1-3).