Estimating Overwinter Mortality of Age-0 Pacific Herring Based on Loss of Energy and Implications for Recruitment

An overwinter mortality model based on empirical mortality and energy density levels following forced fasting was applied to age-0 Pacific herring (Clupea pallasii) of Prince William Sound, Alaska. The frequency distribution of energy density measured during November 2007-2010 was used as model initial conditions. The model was validated by comparing observed energy distributions in March 2008-2011 to model predictions. Modeled mortality from November to March, April, and May resulted in survival rates of, respectively, 22, 5, and 1.2%.  Mortality from starvation from November through May thus explains the two orders of magnitude range observed for herring recruitment in Prince William Sound if there is no starvation mortality for cohorts leading to peak recruitment levels (~ 1 billion age-3 herring). Observations of November and March energy density for cohorts recruiting at 1 billion are needed to resolve whether low energy conditions at the beginning of winter or starvation during winter drives recruitment. Low energy levels in November may be due to lack of sufficient high-energy forage as well from energy losses such as those caused by external parasites. For example, recent in-situ observations of multiple sea lice, actually parasitic copepods, on age-0 herring during their first months following metamorphosis suggest this as a possible energy sink.