Effects of observer efficiency, arrival timing, and survey life on estimates of escapement for steelhead trout (Oncorhynchus mykiss) derived from repeat mark–recapture experiments

Abstract

Estimation of escapement for steelhead trout (Oncorhynchus mykiss) from periodic visual counts of spawners is complicated by extreme changes in observer efficiency over the migration period, low numbers of fish, pulsed arrival timing, and variable survey life. We present a maximum likelihood method to compute escapement and uncertainty that accounts for these difficulties using mark–recapture data from radiotelemetry and snorkel surveys. Estimates of escapement were highly sensitive to assumptions about arrival dynamics and survey life, moderately sensitive to the assumed ending date of the run, and insensitive to assumptions about the form of observation error. Discharge and diver visibility explained between 69 and 78% of the variation in observer efficiency. Simulations revealed that declines in observer efficiency over the duration of the run increased bias and variability in escapement estimates but that this can be mitigated to a limited extent by increasing the number of surveys. The simulations also provided evidence that our likelihood approach was superior to the standard area-under-the-curve method for computing escapement when estimates of the numbers present over time are affected by substantial sampling error.