Assessment of the in-river survival, migration rates and fishery impacts on spring-run Chinook salmon and summer-run sockeye stocks using radio-telemetry and near-shore species composition in the lower Fraser River using fishwheels.

In recent years, the Mission hydroacoustic program has been met with criticism, and errors have been attributed to biased species composition estimates from nearby test-fisheries, which result in incorrect partitioning of the acoustics signals among species. The primary objective for the fishwheel component of the 2009 study was to provide species composition data that could be combined with similar data from the Whonnock test-fishery to determine how best to partition Mission acoustic signals among species. Two fishwheels were continuously operated from early July to mid-September in a relatively fast-flowing section of the Fraser River near Crescent Island. All captured fish were identified to species, and species composition was calculated daily. Including jacks, 15,611 pink, 3,824 sockeye, 1,287 Chinook, 635 steelhead, 43 coho, and 1 chum salmon were captured at the fishwheels, along with 14 other species. The PSC Mission hydroacoustic counts were split into near-shore (<50 m from shore) and off-shore strata. The species composition of the Crescent Island fishwheels and the Whonnock gillnets were applied to the near-shore and off-shore counts, respectively, to derive daily estimates of the number of sockeye passing Mission which were consistent with the PSC’s ‘best judgement’ in-season sockeye abundance estimates.

A second objective was to use mark-recapture techniques to provide an independent assessment of Mission escapement. Fish were marked (radio-tagged) near Crescent Island, and 239 radio-tagged sockeye were known to pass Mission, of which 186 passed Qualark. Mark rates were estimated using the sonar-derived sockeye abundances at Qualark.

A third objective was to radio-tag and track Spring-run Chinook in order to estimate travel speeds, and to identify problematic areas along their migration paths. Capture was attempted using drift and set tangle-nets near Yale and the Harrison-Fraser confluence. Catch rates were very poor, and only 6 Spring-run Chinook were tagged despite 179.5 hours of set netting and 25 hours of drift netting effort between 21 April and 10 June.

The primary objective for the radio-telemetry component of the 2009 study was to provide estimates of in-river survival, migration rates and the impact of fisheries on the survival of adult sockeye and Chinook salmon. In all, 307 sockeye and 86 Chinook were radio-tagged near Crescent Island in approximate proportion to the run size. Each radio-tagged fish was also measured and spaghetti-tagged, and a small adipose tissue sample was taken for microsatellite stock identification. Radio-tagged fish were tracked using 21 fixed-station receivers in 18 locations along the Fraser River and within major tributaries. Eighty-five percent of the radio-tagged Chinook were detected at least once after release, and 59% were known to pass the Mission site, of which 11.1% were returned from in-river fisheries. Ninety-two percent of the radio-tagged sockeye were detected at least once after release, and 88% were known to pass Mission, of which 7.0% were returned from in-river fisheries, and 37.8% were tracked to the vicinity of spawning areas. ‘After-harvest’ survival to spawning areas was significantly higher for Late-run sockeye (94.5%) than for all other run timing groups. Survival estimates for Early Stuart (55.9%), Early Summer (50.9%) and Summer-run sockeye (56.1%) did not differ statistically. Early Stuart, Early Summer and Summer-run sockeye exhibited median travel times that were faster than those for Late-run sockeye and Chinook in the Fraser mainstem. Median Summer-run sockeye travel speeds in 2009 were faster than those measured in 2005, 2006, 2007 or 2008.

The highest rate of en-route loss was observed in the reach between Seton and Quesnel. One probable cause of en-route loss was a period of higher water temperatures, which occurred during the migration of Early Stuart sockeye, and was concurrent with a period of poor survival. En-route losses may also have been fishery-related: 23% of radio-tagged sockeye that were last-detected between Thompson and Quesnel disappeared during periods of fishing activity. Also, most of the en-route loss in 2009 occurred in reaches that are affected by difficult passage points. The data compiled from the 2005-09 studies provide compelling evidence that cumulative effects of elevated water temperature, in-river fisheries and difficult passage points are consistent with the timing and location of en-route losses in 2009 and other years. While there is little that can be done to affect annual water temperatures and flow, it is possible to reduce the amount of stress associated with in-river fisheries. The degree to which en-route losses are associated with intensive fishing periods requires a substantial reduction in the potential for tagging-related effects and more detailed assessments of the timing in-river fisheries above Sawmill Creek. To address this question, we recommend that the tags be applied at times and locations where water temperatures are low (in-river tagging in early July and marine tagging from mid-July through early September) and future studies include a greater number of fixed-station receivers in the Thompson to Chilcotin area, coupled with increased mobile tracking efforts. By conducting additional focused monitoring of selected key fishing areas, it may be possible to better determine the fate of the radio-tagged sockeye entering these fisheries.