bsierp nprb org BESTBSIERP Bering Sea Project FEAST

bsierp. nprb. org BEST-BSIERP Bering Sea Project

FEAST Higher trophic level model Observational Data Nested models BEST-BSIERP Bering Sea Project Economic/ecological model 2 NPZ-B-D Lower trophic level ROMS Physical Oceanography Climate scenarios BEST bsierp. nprb. org BSIERP Integrated modeling

BEST-BSIERP Bering Sea Project bsierp. nprb. org Goal • Provide forecasts of future population trajectories of walleye pollock under anticipated climate warming; complement the vertically-integrated modeling effort

BEST-BSIERP Bering Sea Projections of Key Environmental Parameters for the Bering Sea: Empirical Downscaling from Global Climate Models bsierp. nprb. org Nick Bond 1, Franz Mueter 2, Muyin Wang 1, Anne Hollowed 3, Carol Ladd 4 1 Joint Institute for the Study of the Atmosphere and Ocean, UW 2 School of Fisheries & Ocean Sciences, UAF 3 Alaska Fisheries Science Center, NOAA 4 Pacific Marine Environmental Laboratory, NOAA

BEST-BSIERP Bering Sea Project Spring conditions Timing of ice retreat Prey Spawning bsierp. nprb. org Spring SST Early larvae (spring) Biomass Consumption rate Prey composition (Late) summer conditions Summer SST Stability Wind mixing Prey Late larvae (fall) Age-1 recruits Predation Spatial distribution

BEST-BSIERP Bering Sea Project Variables considered • Timing of ice retreat – Date when average ice cover on southeast Bering Sea shelf first decreases below 20% (updated from Mueter et al 2006) • Spring transition bsierp. nprb. org – Proxy for onset of thermal stratification based on change in SST (NOAA Extended Reconstructed SST, Smith et al. 2008) • Late summer SST – Mean Jul-Sep SST (NOAA ERSST) over shelf • Summer wind mixing – Mean May-July wind mixing (NCEP reanalysis, Kalnay et al. 1996)

BEST-BSIERP Bering Sea Project bsierp. nprb. org Variables considered (cont’d) • Water column stability – Mooring 2 observations & proxy based on 1 -D model estimates • Predation pressure (1977 -2008) – Index of pollock predation based on consumption rate and average prey composition in early 1990 s estimated in Aydin et al (2007) • Spatial distribution (juveniles, adults) – Anomalies in spatial distribution based on summer trawl survey (Spatial EOF/PCA)

bsierp. nprb. org BEST-BSIERP Bering Sea Project

Prediction interval Low Med High bsierp. nprb. org BEST-BSIERP Bering Sea Project Estimated effects of summer SST & predation on log-recruitment R 2 =0. 44 P = 0. 001 Simulate effect of increase in average SST on recruitment at three levels of predation

BEST-BSIERP Bering Sea Project Median R Predicted distribution of pollock recruitment under 0. 5 o. C and 1 o. C warming (relative to 1977 -2001) o Assuming predation equal to 1977 -2007 mean (dashed): bsierp. nprb. org With 0. 5 o. C warming: à 10% decline in median R à 31% exceed ‘ 77 -’ 07 average R With 1 o. C warming: à 30% decline in median R à 19% exceed ‘ 77 -’ 07 average R o

BEST-BSIERP Bering Sea Project bsierp. nprb. org H 3: Winners and losers Piscivorous fish abundance will increase. Competition with abundant, piscivorous fish will lead to a decline in murres, kittiwakes and fur seals. Growing populations of humpback and fin whales increasingly will both consume and compete with forage fish (including juvenile pollock) for zooplankton.

bsierp. nprb. org BEST-BSIERP Bering Sea Project

BEST-BSIERP Bering Sea Project bsierp. nprb. org 2000 s 2010 s 2020 s 2030 s 2040 s > 2050 Time of collapse based on 6 -year running mean of recruitment estimated from SST (constant predation) falling below 10% of present day median value

BEST-BSIERP Bering Sea Project bsierp. nprb. org Summary and Next Steps • Late summer/early fall conditions during larval stage are best predictors of recruitment • Simulations suggest a considerable reduction in average recruitment of walleye pollock under moderate warming • Recruitment model will be revisited; potential predictors such as stratification may be added • All of the sources of uncertainty (model, regional factors, global climate) will be quantified • Projections will be made available for MSE