McMichael, Geoffrey, A., Todd N. Pearsons, and Steven A. Leider. 2000. Sustainable Fisheries Management: Pacific Salmon. Eds. E. Eric Knudsen, Cleveland R. Steward, Donald D. McDonald, Jack E. Williams, and Dudley W. Reiser. Lewis Publishers.


Adverse ecological effects on wild fish resulting from releases of hatchery-reared fish are increasingly being scrutinized and balanced against benefits afforded by hatchery programs.  To improve understanding of the potential ecological effects of hatchery steelhead trout (anadromous form of Oncorhynchus mykiss) on wild trout (resident form of Oncorhynchus mykiss)  populations, we studied releases of 23,000 to 38,000 hatchery-reared steelhead trout smolts into a Yakima River, Washington, watershed from 1991 through 1994.  In this chapter we synthesize results from many aspects of these studies as they relate to minimizing ecological risks to wild trout.  We snorkeled in control and treatment streams to observe behavioral interactions between hatchery steelhead trout with wild salmonids.  Movement of residual hatchery steelhead trout was examined using traps and direct underwater observation and relative abundance of hatchery and wild fish were estimated by electrofishing. Instream enclosures were used to determine whether  residual hatchery steelhead trout impacted growth of wild rainbow trout or spring chinook salmon (O.tshawytscha).  Potential for adverse impacts resulting from ecological interactions among wild salmonids and hatchery steelhead trout was greatest when (1) hatchery fish did not emigrate quickly; (2) water temperatures were over 8 C; (3) hatchery fish were the same species as the wild salmonids; (4) hatchery fish were larger than the wild salmonids; (5) habitat and/or food were limiting; and (6) numbers of fish released was over about 30,000.   Ecological interactions with wild salmonids could be reduced or minimized by releasing (1) only actively migrating smolts (no residuals); (2) hatchery fish that are smaller than wild fish; (3) the minimum number necessary to meet management objectives; (4) fish that are less likely to engage wild fish in agonistic encounnters; (5) when water temperatures are relatively cold (less than 8 C); (6) in areas where wild salmonid populations are absent; and possibly (7) in areas where habitat diversity is complex.  Management actions that encourage angler harvest of hatchery steelhead trout residuals while protecting coexisting wild species may also help minimize negative ecological  impacts.  Implementing these strategies may reduce the number of returning hatchery-origin adults of the target group but will reduce risk to the sustainability of wild fish  populations.


When smolt-sized  (over 175mm fork length) hatchery steelhead trout failed to promptly migrate toward the sea and instead residualized in freshwater, they behaviorally dominated the typically smaller, coexisting wild salmonids.  Larger salmonids typically dominated smaller ones, regardless of hatchery or wild origin or species.  The wild rainbow trout within our study area averaged 15mm fork length while hatchery steelhead trout we released averaged between 179 and 201 mm fork length. Despite this size disparity, rainbow trout and residual hatchery steelhead trout were often observed occupying  the same or similar habitats.  In 84% of the contests we observed between 1991 and 1994, larger fish were judged to be dominant.  Hatchery steelhead trout also dominated over 75% of the contests between steelhead trout and wild rainbow trout. 

Interactions observed in 1994, in streams where residual hatchery steelhead were present, more often involved physical contact (nips and butts; 25%) than those observed in control streams.  In contrast, the most common type of behavior observed in contests in control streams was a threat (47%), while only 20% of the behaviors displayed by fish in treatment streams were classified as threats.  Threats are a more energy efficient way to maintain established hierarchies than those more violent physical interactions commonly observed in treatment streams.  Residual steelhead were displacing wild salmonids more often than wild fish were displacing residual steelhead.

In addition to higher interaction rates, the incidence of disease in both residual hatchery steelhead trout and wild fish was greater when water temperatures were warmer.  For example, in 1991, when water temperatures were very high in late May and June (often exceeding 20 C), Saprolegnia infections were commonly observed on hatchery fish and, though less common, were also observed on wild fish within the release stream.  Infection sites were generally on the lateral body surface near the base of the dorsal fin which corresponds  with the area where most of the violent agonistic attacks we observed were targeted.  It is likely that the combination of high densities of residual hatchery steelhead trout, elevated rates of agonistic interactions, and warm water temperatures combined to increase stress levels in many salmonids to the point that they were vulnerable to fungal infection by Saprolegnia.

The presence of hatchery steelhead trout reduced the abundance of rainbow trout where they were commingled.  Hatchery steelhead trout residuals apparently had a greater impact on conspecfics than on other species, such as spring chinook salmon

To minimize the potential for hatchery fish to socially dominate wild fish, the size of the hatchery steelhead trout released should be smaller than their wild counterparts.

We suggest that only the minimum number of hatchery steelhead trout smolts necessary to meet management objectives should be released in any given situation.  The likelihood of adverse ecological interactions resulting from hatchery steelhead trout releases is positively correlated with hatchery fish abundance.

We estimated that 26 to 39% of the fish we released (23,000 to 39,000) did not emigrate from the study area during the first month after release.

Hatchery releases in areas containing depressed or critical stocks are not advisable simply because they may add ecological stress to precarious stocks.  In addition, any impacts on depleted wild populations will have greater per capita effects than for healthy populations.  In situations where a wild steelhead trout stock is abundant and healthy (e.g. reaching escapement goals), hatchery releases should be considered only on a case-by-case basis. We recommend hatchery fish only be released in areas where there is minimal ecological risk to wild stocks.

Angling regulations could be adopted to encourage the harvest of hatchery steelhead residuals incidental mortality to wild fish is possible when underutilized or unmarked fish are hooked and released.  This factor should be seriously considered before angling is used to reduce the abundance of residual hatchery steelhead.