An assessment of trap efficiency to estimate coho salmon smolt abundance in a small Alaskan stream

Abstract

Smolt abundance is commonly estimated using trap efficiency-based methods; however, few studies have investigated the accuracy of trap efficiency estimates. The objectives of this study were to: (1) test the hypotheses that (i) trap efficiency is not affected by release timing nor release distance, (ii) trap efficiency-based estimates of smolt abundance are concordant with smolt-adult mark-recapture estimates, and (2) evaluate if water level and turbidity influence trap efficiency. In Deep Creek, Alaska, during 2001 and 2002, coho salmon Oncorhynchus kisutch smolt abundance was estimated using trap efficiency-based methods and compared to independent smolt-adult mark-recapture estimates. Marked smolts were released at two times of day (1200 hours and 0000 hours) and two release distances upstream of the trap (400 m and 1500 m) every 2 to 4 d throughout each year. Trap efficiency estimates were highly variable (range 0%-55%) and trap efficiency-based estimates of abundance were not concordant with smolt-adult mark-recapture estimates. Release timing and turbidity significantly influenced trap efficiency, whereas release distance did not. Several assumptions of the trap efficiency approach were not met, which produced biased estimates and conflicting results among years when comparing estimation techniques. These results suggest that assumptions of the trap efficiency-based methods be fully assessed to accurately estimate smolt abundance.