Estimating the Spawning Escapement of Hatchery- and Natural-Origin Spring Chinook Salmon Using Redd and Carcass Data

Authors:
Publisher website
Google Scholar
Abstract
Redd abundance and the demographic characteristics of salmon carcasses are frequently used (in combination) to estimate the abundance and characteristics of populations of Pacific salmon Oncorhynchus spp. However, there is uncertainty about the quality of these estimates because they rely on untested assumptions. We evaluated the quality of spawning ground estimates of Wenatchee spring Chinook salmon O. tshawytscha by comparing the number and origin of fish migrating upstream of Tumwater Dam with estimates of (1) redd abundance and (2) spawner abundance partitioned by sex and origin. Redd counts were well correlated with the abundance of fish passing Tumwater Dam (r2 = 0.84) between 1999 and 2007, indicating that such counts are a suitable index of population abundance. However, between 2004 and 2007, when carcass data were used to partition escapement estimates, the abundance of hatchery spring Chinook salmon was underestimated (mean = 6%, range 0–19%) and that of naturally produced fish was overestimated (mean = 19%, range = 1–43%). In addition, male spring Chinook salmon were underestimated by 20% (range = 18–22%) and females were overestimated by 32% (range = 15–65%). These biases were related to size-related carcass recovery bias and differences in the sex ratio between fish of different origins. The length of male spring Chinook salmon was significant in explaining the variation in the probability of carcass recovery but that of females was not. The sex ratios of carcasses adjusted for recovery bias were not significantly different from that for the population measured at Tumwater Dam. Spawning escapement estimates based on uncorrected carcass data underestimated those based on corrected carcass data 13–39%. We conclude that data collected during spawning ground surveys can be used to estimate the abundance and composition by origin of the spawning population if the biases in carcass recovery data are corrected.
Funding Information
  • Bonneville Power Administration and Chelan County Public Utility

This publication has 25 references indexed in Scilit: