Restoration Program

Program Administration

Session: Hatchery Impacts

Wednesday, February 7, 2007 - 4:15 to 5:45 PM

Influences of hatchery mating, rearing and release practices on life history traits of Chinook salmon at Trinity River Hatchery

David Hankin, Dept. of Fisheries Biology, Humboldt State University, dgh1@humboldt.edu, (707) 826-3683

Presentation [PPS - 2.6 mb]

Although habitat restoration in the Trinity River has been targeted at improving natural production of anadromous salmonids, it is important to recognize that returns of steelhead and Chinook salmon to the Trinity River have been very much dominated by hatchery fish for the past 30 years. Adoption of constant fractional marking for Chinook (25% with AD-clip + coded wire tag (CWT))and 100% Ad-clip marking for steelhead have allowed accurate assessment of the proportions of hatchery fish in spawning runs; these proportions have been very high at weirs and many hatchery adults spawn with wild fish on natural spawning grounds.

In addition to the obvious effect of dominating returns on a numerical basis, modest variation in hatchery mating, rearing and release practices can have profound impacts on the life history traits exhibited by returning fish. Using some old and new CWT recovery data for Chinook salmon released from Trinity River Hatchery and other hatcheries, I briefly show how month and size of release can influence size at age and maturation schedule of returning Chinook. I devote more attention to some new modeling results which show how "completely random mating" of hatchery adults, a policy that has recently been recommended by some geneticists, may lead to unintentional long-term selection for early age at maturity in Chinook salmon. I propose an alternative mating regime that more closely mimics natural spawning behaviors and does not lead to this result. Together, these analyses of existing data and hypothetical models, built on empirical data, argue that it is absolutely critical to consider the role of Trinity River Hatchery in assessment of the joint dynamics of wild and hatchery salmonids in the Trinity River system.

Presentation notes:

An overall question of his work: Are hatchery practices consistent with the goal of maintaining long-term success of natural production? Klamath Trinity fish are early maturing compared to other river basins. Maturing age of males is not correlated with age of females. Early spawning fish seem to produce earlier maturing fish (younger age at maturity). Yearling fish have higher survival but decreased size at 2 to 3 years of age and reduced egg size. Yearling fish held over in the hatchery could be selecting for a slower maturing fish if these yearlings are not selected from the full part of the run. Age at maturity is strongly inherited-jacks beget jacks, and maybe shouldn't be used in the hatchery, as they may be early maturing. Simple model runs suggest that random mating in the hatchery would end up producing younger fish at maturity. Mating males that are older than females would produce older fish at maturity.

^ TOP

Hybridization between spring-run and fall-run Chinook salmon returning to Trinity River, California

Andrew P. Kinziger, Department of Fisheries Biology, Humboldt State University, One Harpst Street, Arcata, CA 95521, Andrew.Kinziger@humboldt.edu , 707-826-3944
Eric J. Loudenslager, Department of Fisheries Biology, Humboldt State University, One Harpst Street, Arcata, CA 95521
David G. Hankin, Department of Fisheries Biology, Humboldt State University, One Harpst Street, Arcata, CA 95521
Eric C. Anderson, National Marine Fisheries Service, Fisheries Ecology Division, Santa Cruz , CA 95060
John Carlos Garza, National Marine Fisheries Service, Fisheries Ecology Division, Santa Cruz , CA 95060

Presentation [PPS - 1.3 mb]

Historically, it is believed that spring-run and fall-run Chinook salmon (Oncorhynchus tshawytscha) returning to the Trinity River were reproductively isolated, with spring-run fish spawning far upstream in early-fall and fall-run spawning further downstream in late-fall. Construction of Lewiston Dam in 1964 resulted in the creation of an impassable migration barrier, however, and spring-run and fall-run Chinook salmon presently overlap extensively in their spawning habitat. This extensive compression of spawning habitat, combined with overlap in spawning period, may facilitate hybridization, which may, in turn, threaten the long-term genetic integrity of these two unique run types. The Trinity River Hatchery (TRH), constructed as mitigation for this loss of upstream spawning habitat, cultures both spring-run and fall-run Chinook salmon and tagging data have shown that they occasionally interbreed fish assigned to the two runs despite substantial efforts to avoid such hybridization. Genetic analysis of samples taken from throughout the spawning season in 1992 indicate the presence of two genetically differentiated populations and that the proportion of fish from the two subpopulations gradually shifted through time, with weekly samples taken earlier in the spawning season having a higher proportion of presumptive spring-run and those taken later with a higher proportion of presumptive fall-run. Simulation analyses suggest extensive hybridization between spring- and fall-run Chinook salmon returning to TRH. Since it is unclear whether genetically distinct groups of spring- and fall-run Chinook salmon returning to TRH may be maintained in the future in the face of ongoing hybridization, we assayed samples collected in 1994 and 2004. Analyses of these data suggest the degree of hybridization has remained relatively stable over the time period studied. It is unclear to what extent this hybridization has been caused by hatchery operations or was occurring prior to construction of the dam. Further, it is unclear what factors are responsible for maintaining distinct runs of spring- and fall-run Chinook despite extensive hybridization.

Presentation notes:

Flow releases may be bringing in fall fish earlier. Possible problems with hybrid between spring and fall fish are have lower survival, genomic extinction, and genetic wastage. Genetic tests show low genetic divergence among fish, but genetic distinctions do exist. They noticed a weak cline through time. They see three possibilities: 1) hybridization will lose the spring and fall runs, 2) the runs are diverging, 3) there is a single ancestral population that is stable.

Questions: How could scales from 1920s collections be used? If there are small amounts of tissue still stuck to scales, this is testable. Any other places to study fish? South Fork and the Salmon are two places. The wild populations could be hybridizing more.

^ TOP

Integrating Hatchery and Natural Production in Restoration Planning - Current State of the Science

Eric J. Loudenslager, Department of Fisheries Biology, Humboldt State University
ejl1@humboldt.edu, (707) 826-3445

Presentation [PPS - 0.1 mb]

Hatcheries have been used to culture salmon and steelhead for over 100 years. Over this period the purpose of hatchery production has changed with an increasing understanding of the biology of salmon and the environments they inhabit. Initially hatchery production was intended to provide sufficient salmon to permit unregulated harvest. Later hatchery production was (is) used to mitigate for lost natural production because spawning habitat was blocked by dams. In this case the hatchery salmon substitute for naturally produced salmon and are intended to be harvested, not spawn naturally. Most recently, hatchery production is being proposed to be used to aid in natural production where the abundance of salmon is depleted. Where hatchery production is intended to substitute for natural production and provide a harvestable surplus, hatchery programs have been reasonably successful. The majority of salmon and steelhead in many rivers are of hatchery-origin and hatchery salmon and steelhead exhibit productivities that exceed natural productivity. There is however, no empirical evidence that natural spawning by hatchery-origin adults improves the demographic status of a natural population. There is conceptual, theoretical, and empirical evidence that the release of hatchery smolts from mitigation programs carries a cost for the natural populations co-inhabiting the watershed. The cost arises from ecological (competition, predation, disease transmission) and genetic (interbreeding between populations experiencing divergent selection regimes) interactions. Measuring the effects has been accomplished in a limited number of locations and appears to be highly variable. The magnitudes of the effects apparently are case specific and not yet sufficiently evaluated to provide prediction. The initial Trinity River Restoration Program's first goal was to Improve the capability of the Trinity River Hatchery to mitigate for salmon and steelhead fishery losses that occurred above Lewiston Dam , and the second was Restore natural (instream spawning) salmon and steelhead production in the mainstem and tributaries below Lewiston Dam to pre-dam levels . Given our current understanding of salmon life-history and the effects of hatchery production on natural populations, achieving both of these objectives might be irreconcilable. Decisions will need to be made about the priority of hatchery production to mitigate dam construction with the goal of supporting harvest, versus restoration of a natural population below the dam with abundance and productivity at pre-dam levels. The challenge of balancing natural and hatchery production to achieve both conservation and harvest management goals is not unique to the Trinity River, but common to most managed river systems in the Pacific Northwest . Under the auspices of the Puget Sound Hatchery Scientific Review Group, and the Northwest Power and Conservation Council Fish and Wildlife Program, at least two planning tools - the All H Analyzer (AHA) and Risk Assessment Modeling Project (RAMP) - have been developed that permit stakeholders, managers, and scientists to begin to quantify the options. Examples of evidence of the effects of ecological and genetics interactions and the modeling tools will be presented.

Presentation notes:

The interactions of hatchery fish with native populations was not really considered 50 years ago. A model was presented that attempts to balance hatchery versus natural production. Loudenslager recommends that trade-offs between competing interests (production for harvests versus natural fish) be done formally and the decisions be transparent. Scientists maybe should not be making these trade-off decisions.

Questions: How are hatcheries doing now? Grade is 1-2 on a scale of 10. 50 % or more of the fish on the spawning grounds are now of hatchery origin. There should be the same amount or proportions of wild fish in the hatcheries. There is notably less hatchery introgression in the Klamath tributaries.

^ TOP

Discussion Session: Hatchery Impacts

Panelists: David Hankin, Andrew Kinziger, and Eric Loudenslager

Where should focus be? It may be counter-productive to focus only on the upper reach where 50 % of spawning occurs. How would hatchery fish do in the wild? Can they add to the fitness of the wild? These are two life phases at equilibrium, the selection differential between them will be important. Natural productivity can decline compared to hatchery productivity, especially if hatchery fish have higher productivity than wild fish in the hatchery. Though the situation may seem bleak, it may not be quite so bad. The beginning stages are to know abundances and interactions. Are these data being presented to hatcheries? Not yet. There is expected resistance because once a decision is made to cut hatchery production, it is perceived as hard to recover this production.

^ TOP