Sufficient direct information does not exist to definitively show that hatchery broodstock weirs placed on spring chinook salmon streams deleteriously affect natural production. However, inferences based upon limited knowledge suggest that weirs will, at a minimum, decrease natural production through both direct and indirect means. In addition, although effects may occur separately, the process for decrease in stock viability most likely occurs through some collective interaction. Long-term operation of weirs also jeopardizes the survival of wild fish above the weir. Finally, massive releases of hatchery fish may also jeopardize overall spring chinook salmon survivals.

The greatest direct effect of hatchery weirs occurs initially when 100% of the adult removed for hatchery production depletes natural production on a one-to-one basis. Subsequently, with continual removal of fish at the weir for hatchery egg take, unless only hatchery-produced fish are removed, a decrease in natural production will also occur. But since most hatchery fish are unmarked, origins of adult returns are generally unknown, and thus the proportion of hatchery to wild removed or released show a weir is also unknown. Although a small decrease in natural production might seem reasonable if expectations are high for hatchery success, subtle unrecognizable effects from this strategy are potentially insidious and, via a number of pathways, may lead toward further large decreases in natural production and, potentially, overall production as well. These effects fall into two broad categories: 1) consequences of decreased natural population size, and 2) consequences as a result of hatchery practices.

Spring chinook salmon from the Snake River Basin already have low genetic diversity compared with other Columbia River salmon stocks (Winans 1989). Additional removal of wild fish from headwater river stretches with low populations may eventually reduce the size of the effective breeding population show a weir to a point where further decreased genetic diversity would limit the long-term viability of that segment of the natural population (see abstract by Waples in Session I). Another possible effect from decreases in the number of spawners may occur if the nutrient load to the upper reaches of streams decreases (Cederholm et al. 1989). It certainly decreases nutrients for predators in the ecosystem, and it may also affect aquatic invertebrate fauna. Since nearly all aquatic invertebrates have some relation to trout and salmon and most are eaten to some extent by fishes (Shapovalov and Taft 1954), changes in nutrient loads, although potentially only subtle, may affect the quality or condition of fry and parr during their freshwater rearing and thus their potential long-term survival.

In addition to removal of wild spawners at weirs, two additional deleterious effects could occur as a result of hatchery fish releases above the weir. First, hatchery fish may not migrate far enough upstream of the weir to take advantage of available spawning habitat. Over time with continual removal of wild fish and replacement by hatchery fish, all spawning activity except in the near vicinity of the weir could cease. This would lead to underutilization of previously productive habitat, possibly replaced by less successful production from less favorable habitat.

Secondly, transmission of hatchery diseases to wild fish could occur. For example, in 1988 the incidence of bacterial kidney disease (BKD) infection in spring chinook salmon adults at Sawtooth Hatchery was 95% (Pascho and Elliot 1989). Some of theft fish had much higher BKD infection levels than others. If hatchery-origin adults released above the weir have high BKD levels, they may spawn with wild fish, pass on the high BKD levels, and thus potentially lower offspring fitness levels.

Outside of problems related to which adults are taken at weirs, another major problem use relates to the increased number of smolts produced as a result of hatchery practices. Between 1964 and 1968, when four dams existed on the lower Snake and Columbia Rivers, 1.3-2.0 million Snake River wild spring chinook salmon smolts produced 50,000-80,000 adult fish (Raymond 1988). Snake River hatcheries now produce upward of 10 million spring chinook salmon smolts. If historical population sizes were related to the carrying capacity of the freshwater environment, then the massive increase in migrating smolts, which feed during the river migration, will, at best, tax the river's carrying capacity, but more likely create severe food competition between smolts, decreasing their overall chance of survival. Other problems also occur with increasing fish density. Shapovalov and Taft (1954) found that as the density of salmonid smolts increased, their survival decreased. They speculated this resulted from an increased susceptibility to predation with increased fish concentrations. Murphy and Shapovalov (1951) and Fagan and Smoker (1989) additionally argued that density-dependent factors operate on stocks during early ocean entry. Fagan and Smoker proposed further that large hatchery releases cause fluctuations in stocks they seek to enhance. They extended the modeling results of Schaffer at al. (1986) and suggested that large-scale, high-production hatcheries can expect many years in which returns we virtually nil.

Whatever the mechanism, past experiences with hatcheries and weirs indicate that wild stock survivals and overall production are lowered with increased hatchery production. Chinook salmon production was low from Oregon coastal streams between 1930 and 1960 when many chinook hatcheries were in production. Because of lack of hatchery success and low production, hatcheries switched to coho salmon. Since the early 1960s, wild stocks of chinook salmon from Oregon coastal streams have rebounded and overall chinook salmon production has increased (Jay Nicholas, pers. commun.)¹.

Finally, speculations regarding problems of removals at weirs for hatchery production from unknown adult stock origins presume that hatchery fish contribute significantly to returning fish populations. If wild fish dominate the returns, however, then maintenance of hatchery production would be at the expense of the wild stock it was designed to supplement, and at times when conditions are favorable for wild stocks to potentially increase, recovery of the stocks in hatchery-influenced areas might not occur. Based upon Idaho Fish and Game index redd counts (White and Cochnauer 1989), it appears this scenario possibly exists. Between 1959 and 1963, the average number of spring chinook salmon redds on what are now classified as wild and natural streams was 1,656, and on what are now classified as hatchery-influenced streams was 2,494. Between 1979 and 1983, the average number of redds in both classification uses dropped substantially to 134 and 360, respectively. However, during the recent rebound in spring chinook salmon returns between 1984 and 1988, the average number of redds in the wild/natural areas increased by 350% to 463, which was 27% of historical levels; whereas the estimated average number of redds in hatchery. influenced areas decreased further to 341, which was only 14% of historic levels. Greater increases in redd counts should have occurred in hatchery-influenced areas if hatcheries were truly supplementing wild stocks.

¹Jay Nicholas, research biologist, Oregon Department of Fish and Wildlife, Corvallis, OR 97333. Pers. commun., September 1989.

Cederholm, C. J., D. B. Houston, D. L. Cole, and W. J. Scarlett. 1989. Fate of coho salmon (Oncorhynchus kisutch) carcasses in spawning streams. Can. J. Fish. Aquat. Sci. 46:1347-1355.

Fagan, R., and W. W. Smoker. 1989. How large-capacity hatcheries can alter interannual variability of salmon production. Fish. Res. 8:1-11.

Murphy, G. I., and L. Shapovalov. 1951. A preliminary analysis of northern California salmon and steelhead runs. Calif. Fish and Game 37:497-507.

Pascho, R. J., and D. G. Elliott. 1989. Juvenile fish transportation: Impact of bacterial kidney disease on survival of spring/summer chinook salmon smolts, 1988. Report to the U.S. Army Corps of Engineers, Contract E86880047, 89 p. (Available from Natl. Fish. Res. Cent., Bldg. 204 Naval Station Puget Sound, Seattle, WA 98115.)

Raymond, H. L. 1988. Effects of hydroelectric development and fisheries enhancement on spring and summer chinook salmon and steelhead in the Columbia River Basin. N. Am. J. Fish. Manage. 8:1-24.

Schaffer, W. M., S. Ellnor, and M. Kot. 1986. Effects of wise on some dynamical models in ecology and epidemiology. J. Math. Biol. 24:479-523.

Shapovalov, L., and A. C. Taft. 1954. The life histories of steelhead trout (Salmo gairdneri) and silver salmon (Oncorhynchus kisutch)with special reference to Waddell Creek, California, and recommendations regarding management, Calif. Dep. Fish and Game Bull. 98. 375 p.

White, M., and T. Cochnauer. 1989. Salmon spawning ground surveys. Unpubl. manuscr., 47 p. Idaho Fish and Game, 600 S. Walnut, Boise, ID 83707.

Winans, G. A. 1989. Genetic variability in chinook salmon stocks from the Columbia River Basin. N. Am. J. Fish. Manage. 9:47-52.

Q:

You say that we shouldn't place weirs to collect wild fish—what options are there to obtain hatchery broodstock?

A:

You should determine if you really need these fish for hatchery production.

Q:

Many wild stocks cannot meet demands for harvest rates (both commercial and sport)—what then might the options be?

A:

It is possible to release 10 million fish and the returns may be low, to add more to it may cause it to get even worse. If you need hatchery fish at one hatchery, get them from another hatchery.

Q:

We have used nonendemic stocks and it hasn't worked. Therefore, we have gone to endemic stocks, but do you imply that this may be a dead end?

A:

Yes; you can see this in Waples's data for certain groups. You can move hatchery fish to a nearby hatchery.

Q:

Is there really 5% survival of Cowlitz stock at Ringold?

Q:

In 1968, the Cowlitz wild spring chinook were introduced to the basin, and there was a 5% return—the returns for the Cowlitz River were probably not much higher, based on counts at Mayfield Dam. Why were there such dramatic differences in Idaho?

A:

I have no idea. The Sawtooth Hatchery may be unique in its low returns. In the cam of spring chinook salmon, with the long outmigration, perhaps as distance increases them is a greater probability of problems. When you look at the 5 billion chum salmon released by the Japanese and the 4 billion chum salmon released by the Soviets, you don't see this. The Cowlitz stock may do well, and some others do well, too. But some need to go a long way, and this may lower the chances of all fish traveling along with them.

Q:

The carrying capacity of other areas (e.g., mid-Columbia) may provide ideas for outplanting densities, so that they may not be exceeded. Should we not stop this altogether, just stay within limits?

A:

Yes, one should go slowly. In the Middle Fork, there were 1,600 redds, then one saw 400 redds. Therefore, that should have a greater potential. But as we supplement more and more, we have to consider the whole system when there is a long migration time. It is an ecological problem that needs to be dealt with. Barging fish to the estuary within 24 hours may help. One should use caution and not consider adding great numbers over what there is now. Comment: Scales from Sawtooth Hatchery and McCall Hatchery fish were looked at and apparently there was a difference between the males of hatchery and wild fish. These scales have been sent to the Oregon Department of Fish and Wildlife (ODF&W) for further analysis. It looks as if more than 80% of the fish were hatchery returns to the Sawtooth weir.

A:

I hope so.

A:

Perhaps it would be better not to mark fish.

Q:

Whether the fish are tagged or not (and about one-third are tagged), only 20-80% arrive at the first downstream dam—why?

A:

That has to be a rhetorical question. Fewer releases may lead to better quality fish. High numbers of releases seem to have suppressed both hatchery and wild fish.

Q:

What is the release strategy at Sawtooth Hatchery? Is it a true supplementation (if so, it is different from past practices) or a conventional culture/release? In the conventional mode I would expect survival to be poor.

A:

The system is already overburdened, so the results would be the same in either case.

Q:

It has been said that there is not full seeding, so why wouldn't there be room for supplementation?

A:

I would say proceed very cautiously with very low numbers. Take into account the status and capacity of individual systems, but go slow.

Table of Contents