Research Document - 2002/116
Pre-season run size forecasts for Fraser River sockeye and pink salmon in 2003
By A. Cass
Abstract
Sockeye production from the 2003 cycle line has been dominated by returns to Chilko Lake and Lower Adams River (Shuswap Lake). Average sockeye returns for all stocks on the
cycle were 6.3 million sockeye/yr compared to an all-year average of 10.3 million fish/yr (1980-2000). At nearly equal proportions, Chilko and Late Shuswap sockeye together accounted for 61% of the total sockeye returns on the cycle since 1980. Forecasts are made for each of 18 individual sockeye stocks and four run timing groups (Table 1). Together the 18 sockeye stocks accounted for 96% of the estimated escapement to the Fraser River in brood year 1999. Escapement estimates for the remaining 4% are extrapolated based on mean recruits-per-spawner for combined stocks with escapement and recruitment data to forecast total returns for all spawning populations.
Fraser pink salmon forecasts for all spawning populations combined are also provided. Pink returns in brood year 2001 were near record levels at 21 million fish. Average pink returns in odd-numbered year was 14 million (1981-2001). Pink escapement in 2001 was well beyond recent historical levels.
Forecasts of returns are made using a variety of explanatory variables. For most stocks, forecasts are based on regression models that use spawning escapement to predict returning age-4 and age-5 sockeye in 2003. Additional explanatory variables are available for some stocks and include smolt and fry data. Model performance was evaluated in a retrospective analysis by comparing forecasts to estimated (observed) run sizes for years that estimates are available. The root-mean-square error criteria was used to select the best model from several candidate models.
Forecasts are provided at various probability levels of achieving specified run sizes by stock and run-timing group. The forecast of sockeye at the 50% level for all stocks combined is 5.5 million fish (89,000 Early Stuart, 412,000 Early Summer, 3.4 million Summer and 1.6 million Late run). The Summer Run forecast accounts for 61% of the total forecast with Quesnel and Chilko stocks in nearly equal proportion at 1.1 and 1.3 million sockeye respectively. The remainder is almost entirely Late run sockeye with the Late Shuswap forecast of 1 million sockeye accounting for 60% of the Late run component. The Fraser pink salmon forecast at the 50% level is 17 million fish.
Forecasts are associated with high uncertainty. Although forecasts are presented as probability distributions, they are based on models that for most stocks assume average survival conditions. Improvements to pre-season abundance forecasts are unlikely without a better understanding of environmental factors affecting survival. Reliability of forecasts ultimately depend on understanding processes that affect survival in both freshwater and the marine environment. Migratory conditions in the Fraser River in 1999 were poor for many sockeye stocks as a result of high river discharge. The effect of stress on survival of the progeny from sockeye that spawned in 1999 is not known. Indicators of sockeye freshwater survival throughout the watershed for the brood were variable. Low egg-to-fry survival was evident for Early Stuart sockeye at one of three sites sampled as well as for Nadina sockeye. Lower than average freshwater survival to the smolt stage is evident in the two sockeye populations where smolts are enumerated (Chilko and Cultus). Cultus sockeye have experienced long-term declines exasperated by the recent high pre-spawning mortality of Late-timed runs generally.
Oceanographic and meteorological conditions in the northeast Pacific and coastal British Columbia in 2000 and 2001 reflect moderate La Niña conditions. These years are the main ocean
entry years for age-5 and age-4 sockeye returning in 2003. Ocean survival conditions were favourable for some sockeye stocks and pink returns respectively in 2002 and 2001. Correlations of survival and specific oceanographic variables, however, have not been demonstrated.
Accessibility Notice
This document is available in PDF format. If the document is not accessible to you, please contact the Secretariat to obtain another appropriate format, such as regular print, large print, Braille or audio version.
- Date modified: