Juvenile Salmon Growth as Proxies for Prerecruitment Ocean Productivity and Subsequent Recruitment of Groundfish in the Gulf of Alaska and Bering Sea
Figure 1. Linear regression model output describing age-1 pollock recruitment (millions) in the eastern Bering Sea as a function of juvenile growth measured on scales of age 2.2 sockeye salmon from Naknek River, Alaska.
Figure 2. Linear regression model output describing log transformed recruitment (millions) of age-1 pollock recruitment in the Gulf of Alaska as a function of juvenile growth measured on scales of age 2.2 sockeye salmon from Karluk River, Alaska.
Figure 3. Multiple regression model output describing the recruitment of age-1 pollock (t) in the eastern Bering Sea as a function of the temperature change index and the juvenile growth (t-1) measured on scales of age 2.2 sockeye salmon from Naknek River, Alaska.
Figure 4. Multiple regression model output describing the recruitment of age-1 Pacific cod (t) in the eastern Bering Sea as a function of the temperature change index and the juvenile growth (t-1) measured on scales of age 2.2 sockeye salmon from Naknek River, Alaska.
The Ecosystem Monitoring and Assessment (EMA) Program recently determined the value of using salmon growth time series as an ecosystem indicator and forecast tool for groundfish recruitment. The concept is that juvenile Pacific salmon growth on the continental shelf is a proxy for the overall productivity of waters above the continental shelf--an important rearing area for age-0 and
age-1 groundfish.
We used juvenile Pacific salmon growth during their first year in salt water (SW1) in year t to predict the recruitment at
age-1 (eastern Bering Sea (EBS) Pacific cod and walleye pollock and Gulf of Alaska (GOA) pollock) and at age-2 (GOA sablefish) in year t+1. Recruitment indices were taken from groundfish stock assessment reports. The measurement of growth during the juvenile life stage was made along the radial axis of the scales collected from adult Pacific salmon.
Juvenile marine growth (SW1) of sockeye salmon from Naknek River in western Alaska was used to predict the recruitment of
age-1 pollock and cod in the EBS. Juvenile marine growth of sockeye salmon from the Karluk River on Kodiak Island in southcentral Alaska was used to predict the recruitment of age-1 walleye pollock in the GOA. Juvenile marine growth of chum salmon from Fish Creek in southern Southeast Alaska was used to predict recruitment of age-2 sablefish in the GOA.
Estimating SW1 from scales of adult salmon rather than from scales of ocean-caught juvenile salmon may be biased due to size selective mortality of smaller fish during the marine life stage. To verify our findings, SW1 was estimated as the average body length of pink salmon captured in the surveys in the southern EBS from EMA Program's BASIS/BSIERP surveys. We also developed a temperature change (TC) index calculated as the difference between the late summer temperature (energy density hypothesis) and subsequent spring sea temperatures (oscillating control hypothesis) in the EBS. We hypothesized that the combined effects of cold summers (energy density hypothesis) followed by warm spring (oscillating control hypothesis) favor the recruitment of groundfish in the EBS.
For the EBS and GOA groundfish stocks, a change in the SW1 and recruitment relationship occurred near the 1988-89 shift in the Arctic Oscillation (AO). The AO is an index of the speed of the counterclockwise atmospheric circulation over the Arctic to lat. 55°N. During the 1977-89 negative AO phase, the tighter and faster circulation of counterclockwise winds in the Arctic stratosphere acted to retain stable cold air in the Arctic. During this phase, the SW1-recruitment relationship was positive, and TC was not significant in the model. During the 1990-present positive AO phase, a looser, slower, and weaker stratospheric circulation pattern results in less stable Arctic air, fewer winter storms, and moves cold air farther south into the Subarctic and GOA.
During this AO phase, the SW1-recruitment relationship was negative, and TC was significant in the model, suggesting that temperature may change quality or quantity of prey and increase competition for food among juvenile pelagic fish in the EBS. The importance of the TC index in determining year-class strength of groundfish indicates that temperature-related processes were more important in driving recruitment during the positive phase AO regimes. Positive AO phases impacted fish stocks differentially in the GOA and EBS. The prediction of a more positive and variable AO indicates that the recruitment of groundfish will increase in the GOA and decrease in the EBS.
During the 1989-2004 period, SW1 was a negative predictor for EBS groundfish and a positive predictor for GOA groundfish. For the EBS, SW1 explained 52% of the annual variability of age-1 pollock recruitment (Fig. 1). For GOA stocks, SW1 explained 32% of the annual variability in the log transformed age-1 pollock recruitment (Fig. 2) and 59% of the annual variability in age-2 sablefish abundance. For the EBS, the TC index and SW1 explained 75% of the annual variability in age-1 walleye pollock abundance the subsequent year (Fig. 3) and 87% of the annual variability in age-1 Pacific cod in the subsequent year (Fig. 4).
To verify our findings we used a short time series of the mean body length of ocean-caught juvenile pink salmon from the BASIS survey in the southern EBS, an area south of lat. 60°N and between long. 161°W and 168°W. For years 2003-07, the model with the average annual length of pink salmon and the TC index explained 80% of the annual variability in the estimated recruitment of age-1 Pacific cod and 95% of the annual variability in the estimated recruitment of age-1 pollock.
Juvenile Pacific salmon body length and marine growth (SW1) were important ecosystem indicators and predictors of subsequent recruitment of groundfish in the EBS and the GOA. Juvenile marine growth helped explain additional variation in recruitment not accounted for by climate indices. Colder summer sea temperatures and shorter juvenile salmon during age-0 life stage and warmer spring sea temperatures during age-1 corresponded with higher subsequent recruitment of age-1 pollock and cod in the EBS. The long-term monitoring of the biology of adult Pacific salmon provides a historical perspective of how the dynamics of marine species respond to low-frequency changes in climate. Monitoring juvenile salmon and age-0 groundfish at sea provides a real-time and more precise tool for predicting the recruitment of commercially important species prior to the age of determination of year-class strength.
