Ecosystem-based fisheries management forestalls climate-driven collapse

Abstract

Climate change is impacting fisheries worldwide with uncertain outcomes for food and nutritional security. Using management strategy evaluations for key US fisheries in the eastern Bering Sea we find that Ecosystem Based Fisheries Management (EBFM) measures forestall future declines under climate change over non-EBFM approaches. Yet, benefits are species-specific and decrease markedly after 2050. Under high-baseline carbon emission scenarios (RCP 8.5), end-of-century (2075-2100) pollock and Pacific cod fisheries collapse in >70% and >35% of all simulations, respectively. Our analysis suggests that 2.1-2.3 °C (modeled summer bottom temperature) is a tipping point of rapid decline in gadid biomass and catch. Multiyear stanzas above 2.1 °C become commonplace in projections from ~2030 onward, with higher agreement under RCP 8.5 than simulations with moderate carbon mitigation (i.e., RCP 4.5). We find that EBFM ameliorates climate change impacts on fisheries in the near-term, but long-term EBFM benefits are limited by the magnitude of anticipated change.

Fig. 1. Model coupling framework.

a Regional downscaling where three global climate models driven by the IPCC AR5 CMIP5 emission scenarios determine boundary conditions of the coupled ROMSNPZ high resolution oceanographic model for the Bering Sea, AK. b Biological downscaling of annual indices from the ROMSNPZ were used to drive thermal parameters in the CEATTLE model (i.e., weight-at-age and predation) as well as climate-enhanced spawner-recruitment relationships. c Annual harvest recommendations (ABC) from the assessment model which were translated into annual catch using the ATTACH social-economic model of the effect of EBFM policies on harvest.

Fig. 2. Future Bering Sea bottom temperatures relative to the ~2.1 OC tipping point.

Bias-corrected projections of survey replicated annual mean summer bottom temperature (°C) for the Bering Sea under CMIP5 Representative Concentration Pathway (RCP) 4.5 (a) and RCP 8.5 (b). Annual temperatures (thin lines) and 20 year running mean (thick lines) temperatures are based on survey replicated samples from the downscaled ROMSNPZ Bering10K model. Persistence is based on 2006–2017 average conditions from the downscaled ROMSNPZ hindcast (gray line). Vertical dashed lines represent the start of the projection period (2018). Projections include three global climate models; the Geofluid Dynamics Lab Earth System Model (GFDL-ESM2M), the MIROC-ESM, and the National Center for Atmospheric Research Community Earth System Model (CESM1). See ref. for more detail.

Fig. 3. Future unfished spawning stock biomass.

Unfished spawning stock biomass for pollock (a, b), Pacific cod (c, d), and arrowtooth flounder (e, f) under future climate change scenarios: moderate mitigation (RCP 4.5; left column), high-baseline emissions (RCP 8.5; right column), and a persistence baseline climatology (solid gray lines in each panel). Vertical dashed line represents the start of the projection period (2018–2100). Solid lines represent the 50th quantile, while shading indicates the 10th and 90th quantiles from 100 random draws from estimated recruitment parameters.

Fig. 4. Change in catch (%) relative to the persistence scenario (constant climate).

Changes in catch relative to the persistence scenario for pollock (a, b), Pacific cod (c, d), and arrowtooth flounder (e, f) under future climate change scenarios: moderate mitigation (RCP 4.5; left column), high-baseline emissions (RCP 8.5; right column), and a persistence baseline climatology (solid gray lines in each panel). Lines represent management scenarios when catch is equal to the annual harvest limit (${\mathrm{ABC}}_y$) based on a sloping harvest control rule (i.e., no cap; thin lines) and when catch is equal to allocation using the same ${\mathrm{ABC}}_y$ in combination with a 2 MT ecosystem cap on groundfish harvest (i.e., 2 MT cap; thick lines). Vertical dashed lines represent the start of the projection period (2018–2100).

Fig. 5. Change in risk of decline in pollock catch under climate change (RCP 8.5) relative to the climate persistence scenario.

Risk includes risk of decline (>10% decline), severe decline (>50% decline) and collapse (>80% decline) in catch during four time periods for scenarios without the cap (gray points) and those with the 2 MT cap. The color scale represents relative risk (0–100) from low (teal) to high (red). Length of segments indicate the magnitude of change in risk between the no cap and 2 MT cap scenarios. Vertical gray segments indicate the mean of the “no cap” scenarios.

Fig. 6. Thermal tipping points for catch and biomass.

Analysis of proportional change in future catch relative to the persistence scenario (∆Catch) as a function of future bottom temperature (°C). Solid lines represent the mean smoothing function (s(x)); shading indicates the 2.5 and 97.5% quantiles from 1000 bootstrap replicates. Scenarios without the 2 MT cap (a, c, e); scenarios with the 2 MT cap (b, d, f). Rows correspond to each species. The thick white and orange lines indicate areas where the 95% CI of the first derivative (s(x)′) of the smoothing functions do not include zero; orange bar indicates indicate where the 95% CI of the second derivative (s(x)″) does not overlap zero; on each line, red circles indicate the best estimate of the tipping point (i.e., s(x)″ is most different from zero).