Performance of Spawner Survey Techniques at Low Abundance Levels

Abstract

Population monitoring is essential to know whether coastal California's Endangered Species Act–listed Chinook salmon Oncorhynchus tshawytscha, coho salmon O. kisutch, and steelhead O. mykiss stocks are progressing toward recovery. In coastal California, salmonids are at the southern edge of their range, and this one of many reasons they are not abundant. This provides unique challenges for monitoring, as different survey methods will result in estimates with different levels of accuracy and precision, which are important for evaluating population trends. For this study we intensively monitored three Mendocino County watersheds to evaluate the reliability of two-stage data for monitoring regional escapement. Under this scheme, regional spawning surveys (stage 1) were calibrated with data from intensively monitored watersheds (stage 2), where escapement was estimated using capture–recapture methods, redd counts, and fish counts. The objective of the study was to evaluate the quality of the stage 2 data for calibrating regional surveys. We evaluated the precision of live-fish capture–recapture estimates and compared these estimates with estimates derived from spawning survey data using carcass capture–recapture, area under the curve (AUC), and redd counts. Live-fish capture–recapture produced escapement estimates with narrower 95% confidence bounds where permanent structures were used to capture fish. Redd counts converted to fish numbers using spawner: redd ratios were chosen for the regional salmonid monitoring method because they were reliable, economical, and less intrusive. Converted redd counts were statistically and operationally similar to live-fish capture–recapture estimates but required fewer resources. The AUC estimates were less reliable than converted redd counts and live-fish capture–recapture methods due to the sensitivity of the estimates of residence time and observer efficiency. Finally, we found that carcass capture–recapture methods were operationally unsuccessful in coastal California streams. On the basis of our results, we recommend that annual spawner: redd ratios from intensively monitored watersheds be used to calibrate redd counts for regional status and trend monitoring of California's coastal salmonids.