Euphausiids in the eastern Bering Sea: A synthesis of recent studies of euphausiid production, consumption and population control

Over the period for which data were available (2004-2012), we observed a strong negative relationship between euphausiid biomass as determined in acoustic surveys and pollock biomass as estimated in the eastern Bering Sea pollock stock assessment (r2=0.82). During this time period, sea-surface temperature was the second strongest predictor of euphausiid biomass, (r2=0.63). However, for the period 2004-2010, bottom temperature (r2=0.94) was the strongest predictor, followed by pollock biomass from the pollock stock assessment (r2=0.82), and sea-surface temperature (r2=0.81). Mean pollock density in the acoustic surveys was not a powerful predictor of euphausiid biomass during either period. In spatially explicit multiple regression analyses for the periods 2004-2012 and 2004-2010 those formulations that included sea-surface and bottom temperatures as well as survey estimates of pollock had the greatest explanatory value. However, when either or both temperature terms were dropped, the explanatory value of the models dropped considerably. When pollock biomass was dropped from the models, there was little change in explanatory value compared to the full model. Euphausiid production and pollock consumption data coupled with a negative relationship between euphausiid biomass and stock assessment estimates of pollock biomass indicate a top-down predation effect. However, strong negative relationships between euphausiid biomass and water temperatures indicate the influence of a bottom-up mechanism. The apparent differences in these results may relate to the different spatial and temporal scales used to assess the pollock biomass used in the analyses. Alternatively, euphausiid biomass may be strongly controlled during a restricted portion of the year, such as spring, if critical food needs are not met in some years. We lack the data necessary to resolve these alternative hypotheses.