CR-331978 (BPA Project 1993-056-00; Biomark Advance Hatchery Reform Research) seeks to contract Biomark, Inc. to improve the functionality of a statistical model developed in the last contract period. Briefly, the objective of the analysis is to understand the carryover effects of smolt age at hatchery release (1 vs. 2 years old) on cumulative life-cycle survival, and how these effects vary from year to year. In particular, the model uses PIT-tag detection data to determine how smolt age affects survival during subsequent life history stages, both directly during downstream migration and early ocean residence, and indirectly during upstream spawning migration if smolt age is related to age at maturity and the latter affects adult survival.

Objectives for a new contract include expanding the existing model to investigate the effect of smolt size at release, and its interaction with smolt age, on stage-specific and cumulative survival to adulthood. Previous work showed that smolt size is related to downstream survival and plays an important role in mediating survival differences between S1 and S2 smolts (Tatara et al. 2017), but these effects have not yet been incorporated into the full life-cycle model. Although size at tagging is available for all PIT-tagged fish, size at release is available only for a subset. To deal with this, we will model the relationship between size at tagging and size at release as a component of the broader model, using the “partially observed covariate” framework. This approach will fully propagate the uncertainty in imputed size at release into the estimated size effects on stage-specific survival and ocean age. We will use model selection to quantify the weight of evidence for size effects, and results will be summarized graphically.