Science Response 2015/014
Pre-season run size forecasts for Fraser River Sockeye (Oncorhynchus nerka) and Pink (O. gorbuscha) salmon in 2015
Pre-season run size forecasts of returning Fraser River adult Sockeye and Pink Salmon in 2015 were requested by Fisheries and Oceans Canada (DFO) Fisheries Management. Forecasts are used for pre-season planning purposes and for in-season management. They are most useful early in each stock-group’s return migration, before in-season test fisheries are able to provide reliable in-season run size estimates. Forecasts are produced by DFO as stipulated in Annex IV, Chapter 4 of the Pacific Salmon Treaty and are presented by stock and run timing group.
Forecasts are presented as cumulative probabilities (10%, 25%, 50%, 75%, and 90%) to capture inter-annual random (stochastic) uncertainty in Fraser Sockeye and Pink returns. A stock’s forecast probability distribution represents largely the range of survival it has exhibited historically (although other factors such as observation error also influence the distribution). Forecast values at the lower p-levels represent lower stock survival and, conversely, values at the higher p-levels represent higher stock survival. The 50% (median) probability level (p-level) is the mid-point of the forecast distribution, indicating a one in two chance that returns will be at or below this value (or conversely a one in two chance returns will be above this value), assuming that stock survival is similar to past observations. Since not all stocks will exhibit similar survival, current methods used to estimate the total Fraser Sockeye Salmon return forecast distribution, which sums the individual stock return forecasts for each p-level, over-estimate the range of potential total returns. It is therefore more appropriate to reference individual stock’s return forecasts versus the total Fraser Sockeye forecast. Alternative forecasts, generated from different model forms (e.g. Ricker, Power, Larkin, etc.), are also presented for each stock to capture the structural uncertainty in the forecasts.
The 2015 Fraser Sockeye forecast is dominated by the following stocks: Chilko: 35%; Harrison: 21%; and Late Shuswap 7% (Table 1A). The Summer Run timing group contributes the most to the total return forecast at 69%, followed by the Late Run (18%), the Early Summer Run (12%), with the Early Stuart Run contributing the smallest percentage (<1%) to the total return forecast in 2015.
Forecasts for 2015 that are particularly uncertain due to either additional uncertainty that is not quantified in the forecast distribution, or, different model forms that indicate notably different return forecasts (and, therefore survival), include the following:
Chilko Sockeye: median forecasts of four year olds (based on preliminary in-season three year old returns in 2014 used in alternative sibling models) indicate that returns for this stock could be substantially lower than the predictions from the traditional set of models presented.
Harrison Sockeye: this stock has exhibited dramatic increases in survival and abundance starting in 2000, and exceptional escapement in 2011 (four year old Harrison Sockeye returning in 2015). Little data exists to predict this stock’s survival following the record escapement, and therefore, this forecast is highly uncertain.
Scotch, Seymour, Late Shuswap, Portage, Gates: these stocks exhibited record escapement in the 2010 brood year (or in the case of Gates the 2011 brood year), which will contribute to the 2015 five year old returns (or four year old returns in the case of Gates); as a result five year old forecast models (or four year old in the case of Gates) are being extrapolated beyond the observed stock-recruitment range for these stocks, which adds additional uncertainty to these forecasts. Note: Scotch forecasts are additionally uncertain since the stock-recruitment time series is short (brood years 1980-2007) and this stock has been increasing in abundance since the start of the time series.
Fraser Pink Salmon: the Pink fry time series is considered an index of abundance only, and observation error remains unquantified. In addition, there have been considerable methodological changes in the recruitment time series (both in escapement and catch) over time. As a result, the Fraser Pink 2015 forecast distribution of returns does not reflect the full range of uncertainty.
This Science Response Report results from the Science Response Process of November 25, 2014 on the Pre-season abundance forecasts for Fraser River Sockeye and Pink Salmon returns in 2015. The 2015 forecast relies methods of past CSAS processes and publications (Cass et al. 2006; DFO 2006, 2009, 2011, 2012, 2013a, 2014a, 2014b; Grant & MacDonald 2013; MacDonald & Grant 2012).
To support the 2015 Fraser Sockeye forecast, an additional Science Response process occurred on January 27 and 28, 2015 to summarize data and information on fish condition and/or survival from the 2011 spawners and their offspring. This Science Response will be posted on the Fisheries and Oceans Canada (DFO) Science Advisory Schedule as it becomes available.
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: