A ROBUST PROCEDURE FOR ESTIMATING SALMON ESCAPEMENT BASED ON THE AREA-UNDER-THE-CURVE METHOD

Salmon spawning escapements are estimated using the area-under-the-curve (AUC) method by dividing the integral of the escapement curve by the average residence time of fish in the survey area, We present two forms of the basic AUC method which differ in the procedure used to estimate residence time from the observations of tagged fish during stream surveys. AUC estimates based on "observed residence times" were sensitive to variability in survey timing, observer efficiency, and tag detection, while those based on "total residence times" were more robust. For two coastal streams, escapement estimates based on "observed residence times" were between 1.1, and 6.8 times larger than an independent escapement estimate (from fence counts and mark-recapture data), while estimates based on "total residence times" were generally closer to the independent estimate (0.74-1.51 times the estimate, and within 26% six times out of seven). The consistency of our results under a wide variety of survey conditions, combined with their strong theoretical basis, indicates that an AUC method based on annual estimates of total residence time and observer efficiency is robust. However, the level of survey effort required for these methods would limit their application to high-priority streams or populations.