Figure 4.

Comparing CSD and nonlinearity indicators in simulated and empirical time series for exploited and unexploited populations. (a–c) Simulated data: in the demographic change scenario we assumed that populations were exploited if their growth rates were above 1.48. In the exploitation scenario, we categorized populations as exploited based on fishing rate: 0 < F ≤ 0.57 (unexploited), 0.57 < F < 1.18 (exploited). (d–f) Empirical data from the NES and CCE fish populations. Exploited populations are the ones targeted by commercial fisheries, while unexploited populations are the ones not fished or bycatch. CV, coefficient of variation; AR1, autocorrelation at-lag-1; Δρ, nonlinearity index. Boxplots reflect the 25, 75 percentiles. Whiskers show the minimum and maximum values for the empirical data and the 5 and 95 percentiles for the simulated data. P-values are derived by a non-parametric Mann–Whitney U-test.