COS 168-4 - Evidence for density-dependent avian predation on migrating juvenile salmon

Thursday, August 9, 2012: 2:30 PM
C120, Oregon Convention Center
James R. Faulkner, William D. Muir and Steven G. Smith, Northwest Fisheries Science Center, Seattle, WA
Background/Question/Methods

Predation is a significant source of mortality for endangered salmonids (Oncorhynchus spp.) during their seaward migration. Avian piscivores are abundant along the Columbia River downstream of the confluence with the Snake River, with large nesting colonies of Caspian terns (Hydroprogne caspia) and double-crested cormorants (Phalacrocorax auritus) concentrated on a few islands in the McNary Dam reservoir. Bird colony sizes exhibit minimal interannual variation and their total consumption of salmonids is relatively constant. Migrating juvenile salmonids can exhibit substantial interannual variation in abundance, which is largely dependent on production, juvenile survival, and removal at upstream dams by smolt transportation programs. In the absence of compensatory mechanisms, we would expect that a constant number of smolts consumed would result in decreasing predation rate with increasing smolt abundance. Our objective was to test hypotheses about the relationship between smolt abundance and predation rate. Individually-coded electronic tags were implanted in juvenile Snake River yearling Chinook salmon and steelhead at upstream sites for detection at downstream monitors and recovery on bird colonies for 14 migration seasons. We used capture-recapture methods and tag recoveries to estimate survival and avian predation rates for those fish.   

Results/Conclusions

The percentage of tags found on bird colonies was higher for steelhead than for Chinook, and highly variable from year to year. After accounting for colony size, the recovery percentages adjusted for detection efficiency were negatively associated with smolt abundance for both species, but relationships were strongest for steelhead. Likewise, higher survival probabilities were associated with greater smolt abundance after accounting for fish travel time, flow, temperature, and dam operations. These results support the hypothesis that predation rates are density-dependent, although we have yet to test how the relationship holds under lower densities. Abundance of migrating juveniles is lowest in years when large numbers are collected for transportation in barges to sites downstream of the dams. Survival of smolts migrating in river is therefore expected to increase if transportation rate is reduced. However, effects on overall smolt-to-adult returns for transported and non-transported fish combined will need to be determined before such management actions are considered.