Improvements to Predicting En-route Loss Estimates for Fraser Sockeye Salmon

The Fraser Panel has identified as a priority for 2015 SEF Projects to examine factors that may explain or predict en-route losses in Fraser River sockeye. This is in response to impacts associated with the growing uncertainty in predicting migration mortality of Fraser sockeye as overall levels of en-route losses increase (mainly in association with higher river temperatures  Patterson et al. 2007 a,b). Because fishing plans are adjusted to account for expected levels of en-route mortality, this has affected not only the achievement of spawning escapement targets but also the achievement of catch goals for U.S. and Canadian fishers in all sectors. Furthermore, harvest and escapement goals are made at the level of the four management groups but en-route losses can vary considerably at the stock level. These problems became acute and costly in 2013, when models used to predict en-route losses were accurate for a single-stock management group, Early Stuart, but not for the more abundant multi-stock Summer-run management group.

The en-route loss model that relies on simple inputs of temperature and flow from a single location grossly overestimated en-route loss of Summer-run fish in 2013. This overestimation of loss led to missed harvest opportunities for First Nations, commercial and recreational fishers from both Canada and the United States. In 2014, extensive debate continued over the use of simple environmentally-based models for predicting loss for all management groups, reminding us that this issue will not abate. The recent challenges faced by the Fraser Panel to make decisions regarding harvest adjustments to compensate for expected levels of en-route loss have highlighted the pressing need to re-examine the factors used to understand and describe en-route loss and the modelling work used to predict these losses.

We propose to address this priority by focusing on three main areas: 1) update and refine the current DNA-based stock assessment information needed for future modelling; 2) improve the existing management group-based en-route loss models (i.e., Management Adjustment models); and 3) expand the work to include new research on migration biology for individual stocks and explore alternative methods for applying this information to management.