9514 - Laserfiche WebLink

Home Browse Search

Corollary 1: Domestication selection can be avoided if there is no mortality in culture. This is a misconception because it fails to recognize that the genetic effects of fish culture can transcend the culture period. Even if all progeny survive in a hatchery until time of release, they will exhibit a range of values for traits such as size, morphology, aggressiveness, swimming speed, metabolic rate, etc., and these characteristics can have a profound effect on post-release survival and reproductive success. For example, some fish will be larger than others at the time of release because they hatch earlier, grow faster, or are better able to compete for food in the hatch- ery. In turn, fish that are larger at release may survive to adulthood at a higher rate than smaller fish (for example, because they are better at avoiding predation). This process will select for genotypes that produce large juve- nile fish under hatchery conditions, even if all fish survive until time of release. Therefore, domestication selection for hatchery-adapted traits can occur even in the absence of mortality during the culture phase. Furthermore, in addi- tion to any contribution to adaptation for hatchery condi- tions, high survival in culture represents a substantial relaxation of selection that would occur in the wild. Corollary 2: Domestication selection can be avoided if family size is equalized. Equalizing family size in cultured populations can help reduce domestication selection (Allendorf 1993), but limits exist to the effectiveness of this strategy. Although in some cases it may be possible to equalize family size during the captive phase, the key to reducing domestication selection is equalizing reproductive output into the next generation, and that is much more difficult. In a typical salmon hatch- ery, for example, 90% or more of the mortality occurs after release of the juveniles, so efforts to equalize family size in captivity can easily be nullified by events that occur later in the life cycle (e.g., Geiger et al. 1997). Even in cases where it is possible to equalize reproductive success across families, the result still will be genetic change relative to the natural population, which typically will experience strong sexual selection for reproductive success. Finally, equalizing family size does not address the within-family component of domestication selection. Corollary 3: Any effects of domestication will be reversed by natural selection that occurs after the fish are released. The process of domestication is easy to understand in species such as cattle and sheep, and few also doubt that most hatchery trout populations are domesticated. How- ever, many question whether domestication really occurs in Pacific salmon hatcheries, which typically culture fish for only 2-18 months of a life cycle that lasts several years. When comparing salmon to domesticated popula- tions using broodstock that spends its entire life in cap- tivity, it seems reasonable to ask, "How can hatchery salmon be domesticated when, after release into the wild, they migrate thousands of miles to the sea and back every generation?" February 1999 FISH CULTURE-PERSPECTIVE No doubt natural selection will operate on the post- release population to help eliminate individuals that are not well suited to survive in the wild, and this can help offset the effects of domestication selection. However, the traits exposed to selection in the post-release juvenile-to- adult phase will generally not be the same as the early- life-history traits for which selection was relaxed during the period in culture. It is mathematically possible for the mortality experienced after release to exactly compensate for genetic change that occurs in the captive phase, but the chances that this will happen are extremely small. Further- more, even if this did occur, it would cancel out any bene- fit of the hatchery program. Therefore, we are led to the conclusion that a successful hatchery program-one that produces more fish than would have been produced in the wild-will always result in some genetic change to the hatchery population. As noted below, whether these changes will affect natural populations depends on several additional factors. Corollary 4: Domestication selection can be avoided if fish are propagated for only a single generation. This belief appears to be surprisingly widespread among hatchery managers and fisheries managers, but I can find no theoretical or empirical support for it. Genetic change can occur at many points within a single genera- tion as well as between generations. There is no mecha- nism that automatically erases genetic changes that occur within a single generation. Theory and empirical studies agree that, in general, cumulative genetic changes will increase with the length of domestication, but the changes will not be zero for a single generation of culture. A variation of this theme is the belief that genetic changes can be avoided if only natural fish are taken into the hatchery each generation for broodstock, thus allowing all returning hatchery fish to spawn naturally. The idea here is that any effects of domestication from a single gen- eration in the hatchery will be erased in the ensuing gen- eration in the wild. This concept is similar to Corollary 3, except that the purifying selection does not occur until the subsequent generation. This would be convenient if it were true, but I am not aware of any empirical data to support it, and its theoretical basis appears to be weak. Genetic theory for populations that experience different selective regimes but are connected by gene flow (e.g., Karlin and McGregor 1972; Lythgoe 1997) indicates that the result of alternating generations between captive and natural environments could lead to adaptation of the overall population to the hatchery environment (and reduced fitness in the wild) as easily as it could preserve adaptations to the natural environment. In practice, the net result would almost certainly be some combination of the two effects. It may be that alternating hatchery and wild genera- tions is the best way to minimize effects of domestication selection-this topic deserves additional study. However, there does not appear to be any scientific reason to believe that this strategy will eliminate domestication entirely. Fisheries .. 15