Coho salmon Oncorhynchus kisutch from the Strait of Georgia were used to test the hypothesis that slower growing fish in their first ocean year had lower survival over the late fall and winter than faster growing fish. The Strait of Georgia provided a suitable area for this study because it is a semi-enclosed rearing area for juvenile Pacific salmon that is distinct from the open marine rearing areas off the west coast. Coho salmon that survived the winter had significantly larger spacing between circuli on scales, indicating that brood year strength is related to growth in the first marine year. Other studies have shown that smaller fish of a cohort are less able to survive periods of energy deficit than larger fish. Thus, size-related mortality in the first marine fall and winter may be an important determinant of brood year strength of some coho salmon stocks and stocks of other species of Pacific salmon.

The critical size and critical period hypothesis identifies two distinct periods of mortality in the first ocean year of Pacific salmon. The first period is immediately after salmon enter the ocean and is believed to be primarily the result of predation. The second period occurs in the late fall and winter and is related to the ability of juvenile salmon to grow during the summer to a size that will allow them to survive in the ocean after the fall equinox. In 2001, coho salmon returned to the Strait of Georgia in the spring of their second ocean year for the first time in 6 years. The scales from these coho salmon had an average circuli width in the area corresponding to the early marine period in the previous summer (2000) that was significantly wider than observed on the scales of coho from the same brood year sampled in September and November, 2000. This indicated that mortality in the fall and winter of 2000/2001was size related with more of the larger fish surviving the winter. The survival of larger coho supports the hypothesis that growth of coho during the first marine summer is an important component of the natural processes that regulate brood-year strength.

We hypothesise that salmon year class strength is determined in two stages during the first year in the ocean. There is an early natural mortality that is mostly related to predation, which is followed by a physiologically-based mortality. Juvenile salmon that fail to reach a critical size by the end of their first marine summer do not survive the following winter. In this study we describe our initial tests of this critical size and critical period hypothesis using data from ocean surveys of juvenile salmon and from experimental feeding studies on coho. Conservative swept volume abundance estimates for juvenile coho, and possibly chinook, indicate that there is high mortality in fall and winter during their first year in the sea. Studies of otolith weight show that the length and otolith-weight relationship for young coho changes in the early fall of their first ocean year. Studies of growth and associated hormone levels in feeding studies show that slow growing juvenile coho are stunted and deficient in an insulin-like growth factor-I (IGF-I). Juvenile coho sampled in September had low IGF-I values, indicative of poor growth. The results of these studies provide evidence for the general hypothesis that growth-related mortality occurs late in the first marine year and may be important in determining the strength of the year class (brood year). The link between total mortality and climate could be operating via the availability of nutrients regulating the food supply and hence competition for food (i.e. bottom–up regulation).

For several decades fisheries biologists focused their attention on the taxonomy, life history, and population dynamics of single species of fishes. Most attention went to the preferred commercial species, but less popular fishes were not totally ignored. There was always the intention to piece together the species relationships into some integrated big picture, but the incentive was never strong enough to bring enough people together for a long enough time to begin to understand how whole ecosystems could be understood and protected. Several recent events have now provided this incentive. The most serious is the recognition that climate im pacts must be understood both for fisheries management and for the detection of global warming impacts. The lessons from recent fisheries management issues such as East Coast cod, Atlantic salmon management, and coho problems, clearly have shown that there are some problems with previous concepts and that it is cost effective to study marine ecosystems. It is also good politics. It is easier to make difficult decisions when people are well informed about what is known and what is not known. It is time to manage and protect whole ecosystems. This will not be a linear extension of single-species thinking. A more abstract concept is needed in which the single species is seen in relation to the processes that affect ecosystems and less in terms of numbers of individuals. The timing of copepod production, the condition of juveniles at certain times of the year, and the abundance trends of associated species may all become ways of assessing fishing impacts. Ecosystem management requires an under standing of the influences that regulate species naturally. For salmon, we propose a new concept of natural regulation that we call the critical size-critical period hypothesis. According to this hypothesis, the abundance of salmon is determined both in the early marine period and in the first fall and winter of ocean residence. The amount of mortality late in the first marine year is related to the rate of growth during the summer. Like all difficult but essential tasks, it is important to get started with ecosystem management. It is also important to recognize that the communication and coordination of relevant information for ecosystem management may be as challenging as acquiring the understanding of how to do it.

Pacific salmon have been fished extensively for at least a century and artificial production of Pacific salmon has been a management strategy to improve production for almost as long. Hatchery production is considered important because it is commonly believed that the carrying capacity for salmon in the ocean has remained more or less constant and is underutilized as a consequence of limited production of smelts in fresh water. Since the mid-1960s, there has been an increase in hatchery production, partly as a response to a desire to increase catch and partly because of improved hatchery techniques. Since the late 1970s there has been a dramatic increase in the total Pacific salmon catch. The increases in catch are now known to result from a large-scale shift in the productivity of the sub-arctic Pacific and management and enhancement actions. Environmental indices changed about 1989-1990 and may indicate that the productive regime of the 1980s has ended. This would imply that under natural conditions Pacific salmon abundance would decline. There are no clear indications of what to expect in the new regime, but it is apparent that the massive production of artificially reared Pacific salmon would not be necessary in a less productive regime. OF concern is the impact that the large numbers of artificially reared salmon will have on wild salmon stock levels. (C) 1997 International Council for the Exploration of the Sea.

This paper develops a hypothesis for the variations observed in adult contributions resulting from size/time of release experiments conducted at West Coast hatcheries over the past decade. Subsamples of hatchery coho salmon were transferred to seawater net-pens and observed for up to 6 months, during which time morphologically distinct forms — parr (stunt and revertant), transitional and smolt — were identified. Both parr forms exhibited little or no growth in seawater net-pens, but maintained plasma sodium concentrations slightly above normal. False smolting occurred for 1–2 months after seawater entry, followed by reversions to parr (desmoltification) of the smaller fish in the population. Critical size — the size of the largest parred individual — is a good indicator of the minimum size necessary for survival and growth in seawater, and a critical growth path must be maintained in seawater to avoid reversion to parr. Critical size increased between the summer and winter solstice and was dependent on fish size, time of year, and age at seawater entry. Critical size in fresh water was the best predictor of mortality after transfer to seawater netpens. In May and June, over 79% of downstream migrants from one hatchery recaptured in the Columbia River estuary were above the critical size; in July, only 23% of the migrants were above critical size. Virtually all migrants captured in the ocean were above critical size: those below critical size disappeared between the estuary and the ocean. Fish released in May and June near the peak of gill Na+K+-ATPase and thyroxine surges grew more rapidly in the ocean than those released in July. Fish released in June produced a greater adult contribution to the fishery than fish released before or after. For any given time of release, the greater the mean size, the greater the contribution. Smoltification and desmoltification are initiated by photoperiod and mediated by size, and may be events independent of migration. Reversion to parr may be a delayed and unrecognized source of ocean mortality. It is recommended that fish be released from hatcheries at sizes substantially larger than at present.

The rate of deposition of carotenoids in pen-reared coho salmon was investigated by the addition of known carotenoid levels to diets. The carotenoids added to the diets were derived from red crab (P. planipes), and a process is described for the preparation of a soya oil carotenoid concentrate. Using a 3-stage counter-current extraction process, extracts containing 155 mg/100 g oil were prepared from red crab (P. planipes). Oregon moist pellets containing 3, 6, and 9 mg carotenoid/100 g were prepared using these extracts and were fed to coho salmon (Oncorhynchus kisutch) for 120 days. The amount of carotenoid deposited in the flesh of the fish was related to the carotenoid content of the diet and to the weight of the fish. Fish fed diets containing 6 and 9 mg carotenoid/100 g for the same length of time contained 60% more flesh carotenoids than those fed 3 mg/100 g. In general, after 120 days of feeding, only those fish feeding on diets containing 6.0 and 9.0 mg carotenoid/100 g and weighing over 215 g were assessed as having good-to-excellent coloration. Analysis of the flesh showed that there was no correlation between its carotenoid and fat contents.