Stock structure and environmental effects on year class formation and population trends of Pacific herring, Clupea pallasi, in Prince William Sound, Alaska

Abstract

Fluctuating forage fish populations trigger large ecosystem responses in the North Pacific. A representative species, Pacific herring, Clupea pallasi, was chosen to model environmental effects on population fluctuations and recruitment with a case example in Prince William Sound (PWS), Alaska. A unique approach was used to 1) develop a spatially-explicit, life history-based conceptual stock model, 2) quantify population level effects of climatic trends, and 3) model key environmental factors affecting recruitment. Framed as a simulation model, the stock model was compartmentalized by life-history stages based on shared habitats and environmental forcing. Initial model conditions impacting year-class formation were adult size-at-age, spawn timing, location and spawner density, and conditions during egg incubation, all impacting a two-stage larval mortality rate. Larval survival probably dictates the extremes in year-class strength. Age-1 abundance should reflect recruitment levels 2-3 yrs later, unless a predator pit exists. A metapopulation structure was proposed with at least two local population groupings with spatial complexity required to maintain stock levels. Herring abundance correlated with long-term climate trends supporting hypotheses of bottom up environmental forcing. Adult growth was oscillatory over a 13 yr period in phase with zooplankton production and climatic trends. Spawn timing occurred progressively earlier over the last 30 yr period with a concurrent regional spawn allocation shift and decrease in recruits per spawner. Incorporating local stock structure and local environmental variables into nonlinear herring recruitment models improved explanatory power over traditional models. Best-fit variables were eastern PWS SST, salinity, SST variance, and salinity variance from spring to fall. Eight critical life stage periods were defined based on the season and lag of the best-fitting varibles. Examining other variables in these critical periods led to defining potential key processes affecting year class formation. Allocation of spawn and age-3 recruits to metapopulation regions also impacted recruitment to PWS as a whole and these results supported the metapopulation theory. The results led to formulation of a new theory, entitled 'opposing response', explaining the mechanism producing the observed pattern of alternating strong and week year class strengths in northern Pacific herring.