Parken et al. 2002

Parken, C. K., J. R. Irvine, R. E. Bailey, and I. V. Williams. 2002. Habitat-baed methods to estimate spawner capacity for chinook salmon in the Fraser River watershed. Fisheries and Oceans Canada, Research Document 2002/114, Nainaimo, British Columbia, Canada.

     DFO requires escapement goals for chinook salmon populations to achieve objectives established by international agreements and domestic policy. In the Fraser River watershed, a stock-recruitment approach was used to generate an escapement goal for Harrison River chinook,, however insufficient data exist for other Fraser chinook to use this method. In this report we focus on the development of habitat models to produce estimates of spawner capacity for chinook salmon returning to the Fraser River watershed. We plan to develop escapement goals from estimates of spawner capacity and management objectives.
     To estimate spawner capacity, we used a stepwise process and initially stratified spawning systems by their biophysical characteristics. Within each stratum, we developed spawner density biostandard and spawner density-habitat relationships. Then, predictive relationships for numbers of spawners considering, and then alternatively ignoring habitat quality estimated spawner capacity for spawning systems within each stratum. These stochastics models were applied within two of eight population strata as case studies to assist with evaluating the habitat-based approach.
     Spawner habitat capacity models performed well overall, however, additional information is required before these models will consistently generate realistic estimates of spawner capacity for chinook in high gradient and confined-channel spawning systems. Spawner-density gradient relationships were significant for several population strata but spawning capacity estimated from relationships with moderate coeficients of etermination had high uncertainty. Estimates of spawner capacity were sensitive to scaling factors and high when scaling scaling factors had positive bias. Additional work is required to improve the precision of the estimates of spawner capacity and to develop accurate scaling factors. Accordingly, the reported models and spawner capacity estimates are preliminary and baed on a developing methodology.