ESTIMATING COHO SALMON (ONCORHYNCHUS-KISUTCH) SPAWNING ESCAPEMENTS BY CONDUCTING VISUAL SURVEYS IN AREAS SELECTED USING STRATIFIED RANDOM AND STRATIFIED INDEX SAMPLING DESIGNS

Spawning escapement estimates were obtained for adult and jack (precocious male) coho salmon (Oncorhynchus kisutch) by conducting visual surveys on foot in stream reaches selected by stratified random and stratified index sampling designs. The approach was evaluated for three years in two streams with widely varying escapements; estimates of adult escapements using the stratified index design were always similar to estimates obtained through an independent mark-recapture program. The stratified random design underestimated escapement in every case but one. Because the distribution of fish was aggregated, with random sampling there was a higher probability of sampling low fish abundance areas than high-abundance areas. Numbers of jack coho were underestimated with each sampling design, probably because we overestimated our efficiency of seeing these fish. When using this approach, the efficiency of observing fish during surveys and the length of time salmon are alive in the survey area should be estimated; the latter for our survey areas ranged between 13 and 17 d with one exception (8 d). Labour costs associated with estimating escapements to our streams by visual survey indexing techniques were less than by mark-recapture or by operating fish counting fences.