Bottom-up and top-down controls of walleye pollock (Theragra chalcogramma) on the Eastern Bering Sea shelf

Control of walleye pollock (Theragra chalcogramma) recruitment in the Eastern Bering Sea involves complex interactions between bottom-up and top-down processes, although the mechanisms are poorly understood. We used statistical models to test the leading hypotheses linking recruitment variability to biotic and abiotic factors. Consistent with a “cold-pool hypothesis”, recruitment of pollock was significantly stronger if winters preceding the larval (age-0) and juvenile stages (age-1) were mild. However, our results did not support the proposed top-down mechanism (cannibalism) underlying this hypothesis. Several empirical relationships support an “oscillating control hypothesis”. As predicted by it, the effect of ice conditions on survival during the larval and early juvenile stages was modified by the abundance of adult pollock, implying stronger bottom-up control when adult abundance (hence cannibalism) was low. The proposed bottom-up mechanism predicts that the survival of pelagic-feeding walleye pollock (benthic-feeding yellowfin sole), should be higher during years with an early (late) ice retreat, which was confirmed by our analysis. Our results also provide additional evidence for a “larval transport hypothesis”, which states that cannibalism of larval and juvenile pollock is reduced in years when strong northward advection separates juveniles from cannibalistic adults.In addition to testing existing hypotheses, we identified new relationships between spawner-to-recruit survival rates of walleye pollock and several indicators of mixed layer dynamics during the spring and summer. Survival rates and recruitment were significantly reduced when larval or early juvenile stages experienced a delay in the (non-ice-associated) spring bloom as a result of stormy spring conditions, suggesting that the timing of the spring bloom is critical to both first-feeding larvae and age-1 juveniles. Furthermore, a dome-shaped relationship between pollock survival and summer wind mixing at the early juvenile stage is consistent with modeling and laboratory studies showing an increase in survival at low to moderate levels of wind mixing, but a decrease in feeding success at high levels of wind mixing.Top-down controls also regulate recruitment of walleye pollock. At least one-third of the variability in spawner-to-recruit survival could be accounted for by predation mortality at the early juvenile stage (age-1). Predation of juvenile pollock can be attributed largely to cannibalism, which varies with the abundance of adult pollock and with the availability of juveniles to adult predators. A simple index reflecting the spatial overlap between juvenile and adult pollock explained 30–50% of the overall variability in recruitment, similar to the variability explained by the best environmental predictors. Although environmental effects are difficult to separate from the effects of predation, we conclude that bottom-up and top-down processes are equally important in controlling the survival of pollock from spawning to recruitment at age 2. However, the magnitude of top-down control is itself modified by environmental factors that control the availability of juvenile pollock to adults (through impacts on spatial distribution) and the abundance of adult predators (through effects on productivity and carrying capacity).