Science Response 2017/016
Pre-season run size forecasts for Fraser River Sockeye (Oncorhynchus nerka) and Pink (O. gorbuscha) salmon in 2017
Context
Low numbers of Fraser Sockeye salmon are expected to return in 2017 compared to both the cycle average (8.4 million) and the all-year average (7.9 million). The median (50% probability level) total Fraser Sockeye forecast of 4.4 million is close to half of the 2017 cycle average. The 2017 forecast ranges from 1.3 million to 17.6 million at the 10% and 90% probability-levels. Fraser Sockeye returns have fallen at or below the median forecasts for the past 12 years, excluding 2010, indicating average to below average survivals. In the past two return years (2015 and 2016), total returns corresponded to the 10% probability level forecasts, indicating poor survival for the Fraser Sockeye aggregate. Although the more abundant Summer-run stocks exhibited notably poor survival in these years, individual stock survival ranged from poor to above average. Most Fraser Sockeye that returned in 2015 and 2016 entered the Northeast Pacific Ocean in 2013 and 2014, respectively. This timing coincided with the development of unusually warm temperatures in the Northeast Pacific Ocean in late 2013, referred to as the ‘warm blob’. During this period, ocean temperatures were 3°C to 4°C above the seasonal average, and extended down to depths of 100 m. Given the persistence of extremely warm ocean temperatures through 2016, and the known association between warm ocean conditions and reduced survival of British Columbia (BC) salmon populations (Mueter et al. 2005), below average survivals are anticipated for the 2017 total Fraser Sockeye returns.
The effects of extremely warm water temperatures on survival have been incorporated quantitatively into the forecasts for seven stocks where temperature covariate models historically perform well. Although the effect of the warm coastal temperatures on Fraser Sockeye survival is highly uncertain, forecasts for these stocks using temperature covariate models were consistently smaller than forecasts produced by models that exclude these covariates. However, these stocks account for only 15% of the total forecast at the median forecast level. Given that the warm ocean conditions that developed in late 2013 have persisted, median (50% probability level) forecasts based on models that do not include indices of environmental conditions may overestimate returns. Therefore, for the remaining stocks, emphasis on the 25% probability level of the 2017 forecasts is recommended. This is particularly important for the Summer-run timing group because the forecasts for key Summer-run stocks such as Chilko, Late Stuart and Stellako, which are expected to contribute a high proportion (63%) to the total 2017 Fraser Sockeye forecast, were not produced with temperature covariate models. Furthermore, these Summer-run stocks exhibited particularly low survivals in the last two return years (2015 & 2016) (Table 2, column F). Thus, the overall Summer-run return may more closely align with the 25%, rather than the 50% probability level of the forecast.
For some stocks, other factors, such as survival due to delayed-density dependence in freshwater rearing areas, are better at explaining variation in historic returns than models that include environmental covariates. For 2017, a number of stocks (Gates, Scotch, Seymour, Stellako, and Late Shuswap) are expected to exhibit reduced survival due to delayed-density dependence, as represented by the Larkin model. For the Shuswap Lake complex, in particular, the extremely large escapements in 2010 continue to affect the forecasted survival of stocks that rear in this system (Scotch, Seymour, Late Shuswap, and miscellaneous Early Shuswap).
Chilko dominates the 2017 forecast of Fraser Sockeye (48% of the total median forecast); however, this forecast is highly uncertain, especially compared to previous years when smolt data were available. In the absence of reliable smolt abundance data, the forecast uses the brood year escapement to predict returns. The brood year escapement for Chilko in 2013 was the second highest on record (the largest escapement occurred in 2010). Since only one data point exists above the 2013 brood year escapement, the model used to generate the Chilko forecast is not well informed by data in this high abundance range. Additionally, the Chilko forecast was generated by the Larkin model, which includes the unprecedented 2010 brood year escapement as a predictor of delayed-density dependent effects. This data point has a strong positive effect on the forecast distribution, which is more pronounced at the higher probability levels. The large Chilko component of the total Fraser Sockeye forecast (48%) is therefore associated with high uncertainty, due to the lack of data at high escapements to inform the Larkin model.
The non-Chilko Summer-run stocks combined are expected to contribute 28% (Late Stuart: 8%; Stellako: 8%; Harrison: 6%; Quesnel: 4%) to the total median forecast. Although Quesnel historically dominated total Fraser Sockeye returns on the 2017 cycle-line (42%), the 2017 Quesnel forecast is extremely low compared to the cycle average. Quesnel is expected to contribute only 4% of the total 2017 median forecast, due to the combined effects of very low escapement in the 2013 brood year, and poor environmental conditions. The Late Run is expected to contribute 13% of the total Fraser Sockeye median forecast (Late Shuswap: 4%, Weaver: 4%, and Birkenhead: 3%) and the Early Summer Run is expected to contribute 8% to the total. The Early Stuart Run is expected to contribute only a small proportion (2%) to the total median forecast. Note that in the Late Run, Cultus sockeye, which are listed as ‘endangered’ by the Committee on the Status of Endangered Wildlife in Canada (COSEWIC), have a 2017 forecast that falls below the WSP abundance benchmark of 12,000 wild spawners (Grant and Pestal 2012) across all probability levels.
The total 2017 forecast of Fraser Pink Salmon ranges from 4 million to 16 million at the 10% and 90% probability-levels, with a median (50% probability level) forecast of 8.7 million. This median forecast is below average (12.4 million). Fraser Pink Salmon forecasts are extremely uncertain given the shifts in enumeration methodology over time, particularly with regards to the recruitment data (changes in escapement and catch methods). Pink Salmon fry abundance in the 2013 brood year was 230 million, which was almost half of the long term average (441 million).
This Science Response Report results from the Science Response Process of December 14, 2016 on the Pre-season abundance forecast for Fraser River Sockeye and Pink Salmon returns in 2017. The 2017 forecast relies on methods of past CSAS processes and publications (Cass et al. 2006, DFO 2006, 2009, 2011, 2012, 2013, 2014a, 2014b, 2015a, 2015b, Grant et al. 2010, Grant and MacDonald 2011, 2013, MacDonald and Grant 2012). To support the 2017 Fraser Sockeye forecast, an additional Science Response process occurred on January 17-18 to summarize data and information on fish condition and/or survival from the 2013 spawners and their offspring.
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: