Survey of Coho salmon runs on the pacific coast of the Alaska Peninsula and Becharof National Wildlife Refuges, 1994 with estimates of escapement for two small streams in 1995 and 1996

Hetrick, Nicholas; Nemeth, Matthew

The primary objectives of this study were to describe the run timing and to evaluate methods for estimating escapement of coho salmon (Oncorhynchus kisutch) in small streams located on the Pacific side of the Alaska Peninsula. In the fall and winter 1994, aerial counts of adult coho salmon were conducted on several streams located along the Pacific Coast of the Alaska Peninsula and Becharof National Wildlife refuges. In 1995 and 1996, efforts were focused to estimate escapement of coho salmon on Clear and Sandy creeks. Escapement estimates were also generated opportunistically for pink (O. gorbuscha), chum (O. keta), and sockeye salmon (O. nerka). We assessed the applicability of using carcass counts to index escapement by exploring the relationship between escapement estimates and cumulative carcass counts between years and between streams. Live fish counts were extrapolated to estimate escapement using the area-under-the-curve method. Observer efficiency and species-specific residence times were measured on Clear Creek in 1996, and resulting averages were used in escapement calculations for Sandy Creek in 1995 and 1996, and for Clear Creek in 1995. Aerial survey counts of adult coho salmon in late September 1994 were highest for Yantarni (1,815), Nakalilok (1,790), Sandy (1,070), and Clear (655) creeks of the 22 streams surveyed. Eleven streams were resurveyed from the air on 1 December 1994. Counts of coho salmon were lower in December than in September for all 11 streams resurveyed, despite excellent survey conditions. Cumulative carcass counts were not significantly (r2 = 0.44) related to escapement estimates as the difference between these values varied greatly between years and between streams. In areas with high densities of bears like the streams we studied on the Alaska Peninsula, methods that rely on carcass counts may not be reliable indices of escapement because of predation on dead and dying fish. Observer efficiency averaged 74% for 13 trials conducted on Clear Creek in 1996, and was dependent on lighting, wind, and turbidity. Residence times were estimated to be 13.8 d for coho, 8.7 d for pink, 7.9 d for chum, and 18.1 d for sockeye salmon. Escapement estimates on Clear Creek, calculated using the average residence times and counts corrected for observer efficiency, were relatively consistent between 1995 and 1996 for coho (4,068 and 3,118), sockeye (122 and 338), and pink salmon (4,239 and 5,041), but were higher in 1996 (3,851) than in 1995(1,885) for chum salmon. Estimates of escapement on Sandy Creek were higher in 1995 than in 1996 for coho (4,057 and 2,205), pink (17,969 and 2,676), and chum salmon (1,867 and 780). We suggest that observer efficiency and residence times be assessed concurrently with foot surveys as both parameters may vary between surveys, among streams, and between years, thereby influencing escapement estimates generated using the area-under-thecurve method. The number of surveys necessary to estimate escapement may, however, be minimized by setting the survey interval equal to or slightly less than the expected residence time specific to that survey period, without a substantial loss of accuracy.

Survey of Coho salmon runs on the pacific coast of the Alaska Peninsula and Becharof National Wildlife Refuges, 1994 with estimates of escapement for two small streams in 1995 and 1996

Series