Residual effects from fish wheel capture and handling of Yukon River fall chum salmon

Since 1996, U.S. Fish and Wildlife Service biologists have annually used fish wheels to capture migrating adult fall chum salmon Oncorhynchus keta in the main-stem Yukon River, Alaska, and estimated their abundance via mark-recapture methods. In each year of the study, the mark rate of captured fish at a site near Rampart has been substantially greater than rates observed at numerous locations upriver of that site. The factors most likely to cause the observed reduction in the mark rate are violations of mark-recapture model assumptions or the mortality of marked fish between the Rampart site and upriver locations. Results of studies conducted through 2000 were most consistent with the hypothesis of mortality. We investigate potential explanatory factors for the apparent reduction in mark rates at upriver locations using data collected during additional studies from 2001 to 2003. Results document that holding fish in submerged pens at the marking site negatively affects their ability to migrate for at least some time. No evidence of tag loss or spatial segregation within the mark-recapture study area was observed. A conclusion that some aspect of the capture and handling of fish elevates their mortality upriver of the mark-recapture study area seems well founded. However, holding of fish does not solely explain the reduction in mark rates at upriver locations, and other contributing factors remain unidentified. Researchers using fish wheels should be aware that the gear may be more harmful to fish than was previously thought and may bias estimators of some parameters such as abundance or migration speed.