Assessing the Precision and Accuracy of Counting Towers Used to Estimate Escapement of Sockeye Salmon on Two Small Streams in Southwestern Alaska

Accurate and precise escapement estimates of Pacific salmon, i.e. the number of adults returning to a particular drainage to spawn in a given year, are critical metrics for effective management and conservation throughout North America.  Measurements of salmon escapement are used to track population trends over time, set harvest rates, estimate stock-recruit relationships  and in some cases to establish environmental baselines.  A myriad of methodologies have been designed and implemented over the past century to correctly estimate escapement of salmon stocks, ranging from simply counting fish at a weir to elaborate mark-recapture studies to counts from high-tech sonar stations.  While existing peer-reviewed literature regards periodic counts from towers as one of the most accurate methods for estimating escapement, empirical studies utilizing counting towers have been somewhat restricted to large rivers.  In this study, we evaluated the precision and accuracy of counting towers on two small streams in southwest Alaska.  We found that in streams where channel morphology is conducive to protection from bear predation tower counts generate very precise estimates of escapement (CV = 0.06).  Conversely, streams that are very shallow and have minimal habitat complexity offer little refuge from bear predation, resulting in erratic migration patterns that consequently generate poor, imprecise estimates of escapement (CV = 1.35).