A Bayesian Analysis of Daily Weir Counts for Estimating Chinook Salmon Escapement

• Activity Number: 80
• Type: Contributed
• Date/Time: Sunday, July 29, 2012 : 4:00 PM to 5:50 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Margaret Short*+ and James Jasper and Chris Shelden
• Companies: University of Alaska Fairbanks and Alaska Department of Fish and Game and Alaska Department of Fish and Game
• Keywords: salmon escapement ; weir ; Bayesian model ; Markov chain Monte Carlo ; Kogrukluk River

Abstract

Estimation of salmon escapement is important in managing this vital component of the Alaskan economy. Fish that ``escape'' the fishery upstream to spawn are counted as they pass a weir (barrier) on their way upstream. On days when the weir is operational, the fish may spend several days before passing; on days when it is inoperational, they pass freely. We model the daily counts (supplemented with radio-telemetry data), incorporating missingness due to breaches in the weir, using a Bayesian hierarchical model, fitted using Markov chain Monte Carlo. We illustrate using Kogrukluk River data for the years 1976-2011. This represents joint work with Jim Jasper and Chris Shelden, both at the Alaska Department of Fish and Game.