Historically, adult summer steelhead Oncorhynchus mykiss returning to hatcheries on the lower Cowlitz River were sometimes transported and released in the river (recycled) to provide additional angling opportunity for the popular sport fishery in the basin. However, this practice has not been used in recent years because of concerns associated with interactions between hatchery fish and wild fish. Fishery managers were interested in resuming recycling but lacked information regarding effects of this practice on wild steelhead so we conducted a study during 2012–2013 to: (1) enumerate recycled steelhead that returned to the hatchery or were removed by anglers; and (2) determine if steelhead that were not removed from the river remained in the system where they could interact with wild fish.

During June–August 2012, a total of 549 summer steelhead were captured at the Cowlitz Salmon Hatchery, tagged, and released downstream near the Interstate 5 Bridge. All recycled steelhead were tagged with a white Floy^® tag and opercle-punched; 109 (20 percent) of these fish also were radio-tagged. All adult steelhead that return to the hatchery were handled by hatchery staff so recycled steelhead that returned to the hatchery were enumerated daily. A creel survey and voluntary angler reports were used to determine the number of recycled steelhead that were caught by anglers. We established three fixed telemetry monitoring sites on the mainstem Cowlitz River and eight additional sites were deployed on tributaries to the lower Cowlitz River where wild winter steelhead are known to spawn. We also conducted mobile tracking from a boat during October 2012, November 2012, and January 2013 to locate radio-tagged fish.

A total of 10,722 summer steelhead were captured at the Cowlitz Salmon Hatchery in 2012, which was the largest return since 2008. River flows during much of the study period were similar to 2008–2011 average flows, however, high-flow periods in July and November 2012 were nearly twice as high as the 2008–2011 average flows. We determined that 50 percent (273 fish) of the recycled steelhead returned to the hatchery and 18 percent (102 fish) of the recycled steelhead were caught by anglers. Most (243 fish; 89 percent) of the recycled steelhead that returned to the hatchery were recollected during July–August. The average elapsed time from release to recapture at the hatchery was 9 days (d) and 72 percent (182 fish) of the fish returned to the hatchery within 14 d of release. These trends were similar for recycled steelhead that were caught by anglers. Most fish were caught during July–August and the median time from release to capture was 10 d. We determined that 65 percent (70 fish) of the angler-caught fish returned to the hatchery within 14 d of release. River flows appeared to affect both hatchery returns and angler catch. The daily number of recycled steelhead that were recollected at the hatchery were low during periods when river flows were decreasing and high during periods when river flows were increasing. Conversely, daily angler catch of recycled steelhead generally was low when flows were increasing and high when flows were decreasing.

We determined that 32 percent of the recycled steelhead (174 fish) were not removed from the lower Cowlitz River, based on observations from hatchery returns and angler reports, but results from the radio-tagged fish were insightful for understanding what may have happened to these fish. By comparison, we determined that 24 percent of the radio-tagged fish were not known to have been removed from the river. We determined that 12 percent of these fish were actively moving in the lower Cowlitz River during October 2012–January 2013. None of the radio-tagged fish were detected in tributaries during the study period except for a single fish that spent approximately 7 d in the Toutle River during early September. During October 2012–January 2013, 10 percent of the radio-tags from recycled steelhead were detected near popular fishing areas, and 2 percent of the radio-tagged steelhead were never detected during the study period. We suspect that a large proportion of these fish may have been harvested and not reported, or died.

Detection patterns of radio-tagged steelhead showed that most fish (82 percent) moved upstream from the release site and were detected at the Trout Hatchery and the Barrier Dam sites. The median time from release to detection at these sites was 3.7 d and many of these fish made multiple trips between the two sites. Nearly one-third (29 percent) of the recycled steelhead that were detected at the Trout Hatchery and the Barrier Dam made at least two trips between the sites and some fish made as many as six trips. Radio-tagged fish that remained in the lower Cowlitz River during the spawning period (December 2012–January 2013) were observed in the river reach between the mouth of Ostrander Creek (river mile 10) and the Trout Hatchery (river mile 44).

During this study, we collected data on opercle punch regrowth rates to understand the temporal effectiveness of this marking technique. We took opercle measurements for a total of 190 fish during the study. Fresh opercle punches were measured for 63 fish at the time of marking, and the remaining 127 fish were measured when fish returned to the hatchery. We determined that opercle punches remained open for about 30 d. The holes appeared to regrow slowly in the first 20 d after marking, but regrowth accelerated during the 20–30 d post-marking period. After 30 d, all opercle punches that we observed had completely closed due to tissue regrowth.

Our study showed that a large proportion (68 percent) of the recycled steelhead were removed from the lower Cowlitz River. These fish primarily entered the hatchery or were caught by anglers within 14 d of release, which suggests that they present minimal risk to wild fish in the system. However, the remaining fish (32 percent) could not be accounted for, which may complicate fisheries management decisions associated with recycling summer steelhead. Findings from the radiotelemetry study suggest that unreported harvest or mortality could explain a large proportion of those fish that were not reported as having been removed from the river. Furthermore, intensive monitoring of the key spawning tributaries failed to detect a single fish during the spawning period. These findings were supported by observations from weir traps operated by the Washington Department of Fish and Wildlife. Our findings indicate that additional research may be warranted to further examine the effects of recycling hatchery summer steelhead in the lower Cowlitz River.