Some steelhead scientists and managers, in recent years, have insisted that hatchery supplementation programs (releasing hatchery fish to augment the numbers of naturally spawning fish) have only a trivial effect on the productivity of the naturally spawning populations, particularly if naturally spawned wild fish are used exclusively for the hatchery brood stock. This is to say that if hatchery fish always are only one generation removed from the wild fish the adverse impact is minimal and tolerable.
In this article I explore this thesis using existing data and a simple model to evaluate the cumulative effects of supplementation over several generations. I have found the thesis to be false.
The existing data better support the hypothesis that supplementation of naturally spawning salmonid populations with hatchery fish will substantially decrease the productivity of the naturally spawning fish and will result in total production far less than anticipated if genetic changes are not considered.
Studies have shown that steelhead in the hatchery, from fry to release size, go through a domestication selection process. Those that do well in the hatchery environment survive and are eventually released, but there is also a known loss in genetic fitness, affecting the rearing and survival of these fish in the natural stream. This loss should reduce the productivity (adult progeny per spawner) for naturally spawning steelhead populations when hatchery fish augment these populations. The magnitude of the reduction depends, in part, on when the period of limiting, density-dependent mortality occurs for naturally rearing fish (density dependent means when there are too many juvenile fish for the available food and habitat).
The available data suggest progressively declining fitness for natural rearing with increasing generations in the hatchery: The reduction in survival from egg to adult may be about 25% after one generation in the hatchery and 85% after six generations. Such reductions for an entire population in stream systems where the amount of spawning habitat limits the naturally spawning populations would reduce carrying capacity (and, therefore, productivity) by these same amounts. Reduction in survival from yearling to adult may be about 15% after one generation in the hatchery, and 67 % after many generations. Such reductions for an entire population in stream systems where the limiting period for natural production occurs at or shortly before the juvenile fish become yearlings would reduce carrying capacity and productivity by these same amounts.
To investigate the cumulative effect of supplementation, I have considered a supplementation program where all hatchery fish spawn naturally and, after the first generation, all wild fish are taken into the hatchery as brood fish. The fish all mature at the same age (i.e., no overlap in generations), and year-to-year variation in environmental conditions is ignored. Each generation of hatchery rearing moves the relative survival for a population, relative to that for the original wild population, one-third of the way to the asymptotic (extreme low) value: approximately 0.55 for survival to yearling or 0.15 for survival to adult. I have assumed that natural selection on each generation of fish rearing in streams likewise will move relative survival one-third of the way to the asymptotic value for natural rearing: 1.0.
This supplementation program produces a saw-toothed decline in fitness or success for rearing in natural streams, until, by the ninth generation egg-to-adult survival is reduced by 50%.
If the naturally spawning population is limited by the amount of spawning gravel, the associated carrying capacity is reduced by 40% within four generations, and by 50% after nine.
Application of spawner-recruit models to a reasonable set of scenarios shows that the actual production (hatchery plus natural) with supplementation can be only one-third that expected if the genetic consequences of hatchery rearing are ignored. If supplementation is discontinued, the population may be in danger of extinction depending on how quickly it can regain fitness for natural rearing.
The cumulative effects of supplementation will vary with the proportion of wild fish brought into the hatchery each year, the productiveness of hatchery and natural rearing, and the period of density-dependent limitation in the natural stream. Although I have presented results for only a few combinations of these parameters, the results are sufficient to show that managers can seriously overestimate the benefits of hatchery supplementation by ignoring the genetic consequences. Modifying hatchery environments to reduce domestication selection may ameliorate the genetic problems, but the practicable level of amelioration is unknown and may be slight.