Language selection


Research Document - 2011/081

Hierarchical Bayesian Model to Estimate the Spawning Stock of Striped Bass (Morone saxatilis) in the Northwest Miramichi River, 1994 to 2010

By G. Chaput and S. Douglas


A Bayesian hierarchical model is described and applied to the catch, mark and recapture data to estimate the spawning stock of Striped Bass in the Northwest Miramichi River for 1994 to 2010. The hierarchical structure of the model considers the variation in catchability among individual commercial gaspereau trapnets within years as well as the variation in catchabilities among years for individual traps. The estimated catchabilities of individual traps show large variation among years. The estimated sizes of the spawning stock of Striped Bass have varied from 3,700 to 92,000 spawners during 1994 to 2010. The estimates of population size derived from mark and recapture experiments are of good precision, annual coefficients of variations (CV) range from 7% to 25%. When only catches and effort are available, as was the case for 1994, 1996, 2006 and 2010, the estimated spawning stock sizes are highly uncertain, both in their location (median) and their precision (CV of 63% to 265%). Because of the highly variable catchabilities estimated among traps and among years, precise estimates of population size require the collection of annual mark and recapture data.

Date modified: