Trade-offs of managing Arctic predator harvest instability in fluctuating oceans

Sustainable human exploitation of marine living resources stems from a delicate balance between short-term yield stability and long-term population persistence to achieve socioeconomic and conservation goals. However, imperfect knowledge of how oscillations in ecosystem processes regulate fluctuations in exploited populations can obscure the risk of missing management targets. We illustrate how the harvest policy to suppress short-term yield fluctuation inadvertently disrupts population cycles and yield stability of exploited, long-lived predators under stochastically fluctuating environmental forces (food availability and regional climate) using Northeast Arctic (NEA) cod (Gadus morhua, an apex predatory fish) as a case study. We use a stochastic, empirically parameterized multispecies model to simulate NEA cod population dynamics through life-history processes; Barents Sea capelin (Mallotus villosus, a pelagic forage fish) modulates cod productivity through density-dependent cannibalism–predation dynamics, whereas sea temperature regulates cod consumption, growth, and recruitment. We first test how capelin and sea temperature fluctuations regulate patterns in cod yield fluctuation and then quantitatively assess how fishing pattern designed to limit yield between-year variance (within 50–5%) perturbs cod population–catch dynamics. Simulations suggest that capelin and temperature interactively contribute to shifting cyclic patterns in cod yield fluctuation primarily through cod cannibalism–predation dynamics. Wavelet analyses further show that muffling yield variance (30 % or less) reshapes the cyclicity (shorter period and greater amplitude) of cod population size and demography, thereby becoming progressively unsynchronized with fishing pressure. Our work reveals unintended consequences of managing transient dynamics of fished populations: the interworking of population cycle destabilized by inadvertently intensifying fishing pressure, amplifying yield fluctuation and, in turn, elevating overharvest risk when not accounting for compounded effects of stochasticity in ecologically connected processes. These policy implications underscore the need for an ecosystem approach to designing ecologically sound management measures to safely harvest shared living resources while achieving socioeconomic security in increasingly more dynamic oceans in the Arctic and elsewhere.


3 8
Capelin fluctuation may, thus, regulate the intensity of cod cannibalism and year-class 1 3 9 strength, likely contributing to depensatory effects of depleted food supply (Hjermann et al.  To simulate the Barents Sea cod-capelin dynamics, we used a stochastic, food-web model We used capelin biomass as proxy for cod carrying capacity, representing "good" and 1 8 6 "poor" years of habitat conditions that regulate cod growth and cannibalism. The model uses shrimp Pandalus borealis, and krill Euphausiidae) being aggregated as "other food'. We 1 9 0 statistically modeled between-year variability in capelin production (as an aggregate of 1-to 1 9 1 5-yr-olds) to capture observed patterns of declining biomass 1) when cod SSB exceeding 1 9 2 400,000 t or 2) when capelin biomass falling below three million t in the previous year (no 1 9 3 feedback for cod predation on capelin), while capelin biomass is capped at six million t when 1 9 4 cod SSB exceeds 800,000 t. Modeled capelin biomass stochastically varies between years and Sea temperature regulates cod population dynamics through recruitment, growth, and 1 9 7 consumption to capture variable metabolic rate in the model (Fig. 1f). In this study, we 1 9 8 primarily focused on short-term sea temperature fluctuations (rather than multidecadal shifts), AMT data were randomly selected from these three groups sequentially. The modeled 2 0 6 thermal periods (cold, moderate, and warm) last from one to five years; the length of each 2 0 7 period was set randomly (effectively generating roughly decadal oscillations, Fig. 1b fishing mortality rate of adults (a fixed age range, 5-to 10-yr-olds, is used for this 2 2 2 computation) in year 1; and F′ j is fishing mortality rate of j-yr-olds in year 1.

2 8
Population-fishery asynchrony in mean fishing mortality became progressively more amplified, peaking in year ~75 before 3 3 5 target exploitation rate was substantially reduced by the management model (Fig. 4a).

6
Stability-yield tradeoffs 3 3 7 Although average catch forecasts increased (by 10-22%) as the strength of stability control 3 3 8 increased (Table 1), overharvest risk became increasingly greater owning primarily to high 3 3 9 between-year variability (Fig. 5). With constraints of 30% or more, SSB fell below the lower constraints of 30% or less, SSB fell below B lim more frequently (up to 6.5%) because of 3 4 2 greater amplitudes resulting from unsynchronized cod population-harvest dynamics (Fig. 4).

4 3
These risk levels under variable capelin productivity and sea temperature were up to 600x We demonstrate that the harvest strategy to suppress short-term yield fluctuation can analyses reveal that the cyclicity of stochastic environmental forces (prey abundance and sea 3 5 2 temperature) can interactively shape the periodicity and amplitude of predator population and inadvertently elevate variability not only in catch but also in population behavior in processes that regulate cod productivity-a density-dependent interplay of cannibalism and productivity cycles when fishing pattern is not forced to suppress yield fluctuation.

8 8
Over the past half century, the Barents Sea has shown large fluctuations in productivity trophically mediated processes that exacerbated the cod population depletion, including 3 9 5 increased consumption of the young by marine mammals such as harp seals (Bogstad et al. forage species productivity may increase overharvest and depletion risks of fished predators. In the case of NEA cod, analyses indicate that an attempt to dampen the amplitude of yield groups of spawners (~6-to 8-year-olds for NEA cod, Appendix 1: Fig. 1 and Rouyer et al.  Although large short-term variation in catch is often disfavored by harvesters and seafood promoting population resilience (restoration of age and size structure, for example) rather 4 3 7 than imposing apparent stability (Anderson et al. 2008, Carpenter et al. 2015.

3 8
Our case study with the exploited Arctic piscivore-planktivore complex illustrates that the   projecting stock response to climate change: example from North East Arctic cod. Marine