From ADF&G, Nov. 19, 2013                                    

World's biggest sockeye salmon run Credit:

World’s biggest sockeye salmon run

A total of 27.20 million sockeye salmon are expected to return to Bristol Bay in 2014. This prediction is 30% lower than the previous 10-year mean of total runs (39.13 million; range of 27.31 million to 46.33 million), and 16% lower than the long-term mean of 32.29 million. The forecast range is from 18.97 million to 35.45 million. All systems are expected to meet their spawning escapement goals.

A run of 27.20 million sockeye salmon can potentially produce a total harvest of 18.54 million fish with escapements near the midpoint of their escapement goals and industry is capable of taking the surplus fish. The projected harvest includes 17.48 million fish in Bristol Bay and 1.06 million fish in the South Peninsula fisheries. A Bristol Bay harvest of 17.48 million would be 35% lower than the previous 10-year mean harvest (26.86 million; range of 17.03 million to 31.10 million), and 11% lower than the long-term mean of 19.74 million.

The run forecast to each district and river system is as follows: 10.50 million to Naknek-Kvichak District (5.29 million to Kvichak River; 1.72 million to Alagnak River; 3.48 million to Naknek River); 4.64 million to Egegik District; 1.89 million to Ugashik District; 9.42 million to Nushagak District (7.24 million to Wood River; 1.40 million to Nushagak River; 0.78 million to Igushik River); and 0.75 million to Togiak District (Table 1).

The total run forecast of 27.20 million sockeye salmon is expected to be comprised of 10.07 million age-1.2 fish followed by 9.37 million age-1.3 fish, 4.94 million age-2.2 fish, 2.70 million age-2.3 fish, 0.121 million age-1.4 fish, and 0.006 million age-0.3 fish (Table 1).


Prediction or forecasting is very difficult, especially if it is about future salmon returns. We have used similar methods since 2001 to produce the Bristol Bay sockeye salmon forecast. These forecast methods have performed well when looking at the overall Baywide forecast. The forecast in 2013 was 4.7% below the total run and forecasts since 2001 have averaged 6.5% below the actual total run. Run forecast differences have ranged from 25.8% below actual run in 2007 to 20.6% above actual run in 2011. Forecasted harvests have averaged 19% below actual harvest since 2001 and harvest differences have ranged from 23% below actual harvest in 2009 to 35% above actual harvest in 2011.

There is a much greater amount of uncertainty in our forecasts of returns to individual rivers. Since 2001, on average, we have under-forecast the returns to the Alagnak (-35%), Togiak (-20%), Wood (-9%), Kvichak (-5%), and Naknek (-2%) rivers and over-forecast returns to Igushik (61%), Egegik (33%), Ugashik (11%), and Nushagak (1%) rivers.

Even though there is large amount of variability around the forecasts to the individual rivers, the overall Bristol Bay forecasts have been fairly accurate since 2001. This appears to have been the

result of over-forecasting returns to some rivers and under-forecasting returns to other rivers. The forecasts to individual rivers have been offsetting each other such that the overall Bristol Bay forecast has been more accurate than the individual forecasts.

We anticipate the 2014 run will be dominated by age-1.2 sockeye salmon (37%), followed by age-1.3 (34%), age-2.2 (18%), and age-2.3 (10%). There is always some uncertainty in our forecast of returns by age class. However, we expect the overall uncertainty in 2014 to be similar to what occurred in prior years. In 2013, we under-forecasted age-1.3 (39% forecast compared to 49% observed) and age-2.3 (13% forecast compared to 21% observed) sockeye salmon. We over-forecasted age-1.2 (23% compared to 16% observed) and age-2.2 (25% forecast compared to 12% observed) sockeye salmon in 2013. In general, there is more uncertainty in 2-ocean returns because there is less reliable information (jack returns from the previous year or brood year spawning abundance) available for which to build a dependable forecast model.

Historically, total runs of sockeye salmon to Bristol Bay have been highly variable. The 2014 forecast of 27.20 million is below the long-term historical average of 32.29 million from 1963 to 2013, and below the more recent historical average of 39.13 million from 2004 to 2013. We had seven consecutive years from 2004–2010 where total run exceeded 40 million sockeye salmon. In 2011, total run dropped to 31.91 million sockeye salmon and we are not sure if this recent trend of lower productivity will continue. We expect the 2014 run to be similar to the total run in 2013.

Chuck Brazil, Fred West, and Greg Buck

Alaska Department of Fish and Game

Division of Commercial Fisheries

Bristol Bay Research Staff