Ocean Conditions Symposium - Managing for salmon as if the ocean mattered

Managing for salmon as if the ocean mattered

Daniel L. Bottom
Oregon Department of Fish and Wildlife
Corvallis, Oregon

Introduction

The topic of this symposium?the effects of the ocean and climate on salmon production?marks a dramatic break with traditional ideas in fisheries conservation. Through most of its history, fisheries science has assumed that salmon populations are in a stable balance that is maintained by biological interactions in freshwater. According to this idea, nature produces a vast excess of salmon eggs and fry each year, which are subjected to "density-dependent" mortality from competition, predation, and disease. Left undisturbed, each salmon population presumably reaches a stable threshold or "carrying-capacity" level that is determined by the most limiting resource in the environment, typically thought to be the amount of food or habitat available during the earliest juvenile stages.

Upon this density-dependent and freshwater foundation, fisheries science constructed a framework of ideas that has long guided salmon conservation in North America. Included in this framework are the following principles and assumptions:

Nature produces a predictable surplus of salmon for harvest.

Since density-dependent factors limit the production of each brood of salmon, any spawners returning above the minimum number required to seed the available freshwater habitat will be "wasted" if they are not harvested. Because populations are in a perfect balance with their available resources, the annual surplus available for harvest can be precisely calculated from equilibrium population models.

The goal of management is to remove limiting factors to salmon production.

The stable balance of populations with their environment allows resource managers to manipulate salmon and their ecosystems to achieve management goals. By controlling the sources of freshwater mortality, the harvestable surplus can be increased to claim for people what would otherwise be "wasted" in nature. Large estuarine and marine environments, which cannot be readily controlled and are presumably neutral factors in salmon production, for all practical purposes can be ignored.

The effects of management manipulations are ecologically benign.

Natural systems, as evidenced by the stable balance of populations, are like machines with separable and replaceable parts. These parts can be modified or substituted to more efficiently produce salmon without adversely affecting the ecosystems that support salmon.

Management success is measured by salmon abundance.

The goal to increase salmon production and yield dictates that abundance measures (e.g., total pounds and numbers of fish harvested or annual return of spawning adults) are the key indicators for evaluating management results. Differences in the geographic origin or behavior of local populations (e.g., run timing, duration of estuarine rearing, age of returning adults, etc.) are irrelevant to the total quantity of salmon produced.

The appropriate scales for understanding salmon are defined by the jurisdictional boundaries of the agencies that manage salmon.

The jurisdictions of government agencies determine the geographic extent of their interests in salmon. These proprietary interests, in turn, dictate the relevant scales of information each agency needs to properly manage salmon.

Recent evidence that Columbia Basin salmon production may be controlled by changes in the atmosphere, ocean, and estuary undermine these traditional assumptions. The results show that salmon populations are not in a steady-state equilibrium, that annual returns are not entirely determined by freshwater conditions, and that salmon variability is at least partially explained by remote physical factors that can affect survival independent of the density of fish. Moreover, the vast scale and uncertainty of oceanic and atmospheric processes call into question traditional harvest and hatchery programs that depend on steady-state population models and artificial control of freshwater environments. The presenters at this symposium raise important issues that require changes in the way we think about salmon and how we might go about conserving them. Here I briefly summarize the history of ideas about the ocean and salmon, new information and ideas presented at this symposium, and the implications of these results for salmon conservation in the Columbia River Basin.

Changing Views of the Ocean

Throughout the history of resource management, ideas about the importance of the ocean to salmon production have evolved slowly to accommodate new understanding of salmon biology and changing attitudes toward fishery regulation. This history of thought can be characterized in three major stages (Figure 1):

The Ocean is Irrelevant (1870s to 1920)

During the second half of the nineteenth century, state and federal fish commissions were established to address concerns about the decline of fisheries in New England and to promote new hatchery technology that promised to make the nation's most valued food fishes widely available to all citizens (Bottom 1997). Hatchery development was consistent with a prevailing cultural ideal that promoted efficiency and control of natural resources for human benefit. The desire to increase fish production emphasized the freshwater phase of salmon life history where limitations to survival seemed readily apparent and could be artificially controlled. The success of hatcheries, though unsubstantiated, seemed self evident because 70 to 90% of the eggs survived in a hatchery environment compared with only a few percent or less in nature. These results led to wildly optimistic claims of potential hatchery benefits based on the untested assumption that total fish production would increase in direct proportion to the number of eggs that were saved by raising them in a controlled environment (Lichatowich et al. 1996; Bottom 1997). The presumed success of hatcheries, in turn, supported a policy by the U.S. Fish Commission to make fish so abundant that harvest regulations would be unnecessary (Goode 1886).

During this period, estuarine and ocean environments were not considered relevant to salmon management. Before the turn of the century, little was known about the ocean life of salmon species, which were generally assumed to rear within a few miles of the mouths of their local streams until they matured (Lichatowich et al. 1996). The high natural mortality of eggs and fry and the ability to control these effects, focused exclusive attention on the freshwater phase of salmon life.

After 1900, experience with marine fisheries, particularly in the North Atlantic, convinced many biologists that overfishing was becoming a serious problem (McEvoy 1986). In the Pacific Northwest, these concerns raised some doubts that salmon abundance could be maintained through hatchery production alone (Higgins 1928). New fishing technology increased the efficiency of salmon harvest in the Columbia River, and boats outfitted with gasoline motors began to establish troll fisheries in the ocean. By 1920, as many as 2,000 trollers worked the mouth of the Columbia River (Smith 1920 cited in Lichatowich et al. 1996). These changes expanded the distribution of fisheries into adjacent coastal waters, a development that would have far-reaching implications for future conservation efforts. Concerns about overharvest and the need for better information to support conservation stimulated early research on the migrations of salmon (Gilbert 1912). But resource managers remained steadfast in their belief that hatcheries could maintain or increase fish production. Within a framework of presumed freshwater population control, estuarine and ocean factors simply had little meaning.

The Ocean is Benign (1920s to 1970s)

By the turn of the century, public anxiety over the effects of unrestrained economic development became the focus of new conservation policies of the Progressive Era. Progressive ideals emphasized the principle of scientific management by experts to insure efficient production and equitable allocation of natural resources. These ideas were widely accepted in fisheries conservation by the 1930s (Larkin 1977) as fish and game regulations expanded and data collection increased to provide a scientific basis for management. New tagging studies documenting extensive ocean migrations by salmon and the movement of salmon fisheries offshore meant that the ocean was indeed becoming relevant to fisheries management. But despite new information about salmon life history, interpretations of salmon production continued to rely on traditional freshwater assumptions. The ocean was like a vast, inexhaustible pasture designed to accommodate all those salmon that survived the rigors of freshwater life.

A major focus of scientific management during this period was the improvement of hatchery technology to more fully control salmon mortality in freshwater. Throughout this period, fish runs in the Columbia River steadily declined. But despite the failure of hatcheries to maintain salmon abundance, hatchery production continued to expand (Lichatowich et al. 1996). With the construction of mainstem dams on the Columbia River, new hatchery programs were established to mitigate effects. Continued research after World War II yielded substantial improvements in fish nutrition and disease prevention and allowed fish to be reared to a yearling ("smolt") stage before release. Throughout this rapid expansion of hatchery programs, the productivity of the estuary and ocean seemed unlimited; fishery managers simply assumed that the vast ocean reservoir downstream could absorb any desired quantity of salmon smolts released from Columbia River hatcheries.

By the 1920s, understanding the fluctuations of economically important species had become a primary focus of many subdisciplines of ecology including fishery science (Kingsland 1995). The entire science of population regulation developed from a new conceptual framework that viewed species as collections of populations rather than as a single homogenous and unchanging group (Mayr 1982). Studies of Atlantic herring and other marine fishes (Heincke 1898 and Hjort 1914) had revealed a geographic structure of populations within species that accounted for year to year fluctuations in abundance. Prior to this understanding, variability in the landings of many marine species were assumed to result from changes in the distribution of a single species group across its geographic range (Sinclair 1988). The population ideas that first developed out of studies of marine fishes were similarly applied to anadromous salmon. By the 1930s, studies had confirmed the "home stream theory," which held that salmon return to their natal streams to spawn (Lichatowich et al. 1996). These results revealed a complex geographic structure of populations within salmon species. Rich (1939) thus described in detail the importance of protecting local breeding populations of salmon as the fundamental units for conservation of species.

Biologists immediately recognized that the population structure of wide-ranging marine and anadromous fish had a critical meaning for fisheries management: Harvest of a population in one part of its range would affect the same population across the rest of its distribution (Sinclair 1988). In the late 1920s, chinook salmon tagged from the Columbia River in ocean waters west of Vancouver Island demonstrated the international conflicts that could arise when a population harvested at sea had spawning grounds in another country (Lichatowich et al. 1996). When Willis Rich (1939) later summarized the stock concept for Pacific salmon, he warned that any conservation efforts within the Columbia River could be negated by the activities on distant ocean fishing grounds. Thus, interest in the ocean life of salmon first developed around the definition of property rights and the allocation of migratory fish that crossed state and national boundaries.

Ironically, this same understanding of population structure that raised concerns about harvesting salmon in the ocean had little influence on thinking about ecological effects of the ocean on salmon production. The discovery of dominant year classes in marine fish populations had clearly demonstrated that ocean conditions during early life stages could account for year-to-year fluctuations in their production (Sinclair 1988). But an important distinction was made when population ideas were applied to anadromous fish: If conditions during early life stages are most important to recruitment success, then the critical factors in salmon production must remain in freshwater. W. F. Thompson (1919) concluded that "The salmon is a highly localized anadromous species, for which artificial propagation is carried on very extensively, and its fresh water life is perhaps more critical than its marine. It is therefore not comparable to purely marine species." Willis Rich (1928) proposed a research program for the International Pacific Salmon Investigation Federation that included "for the sake of completeness" research on various biological factors influencing the marine survival of salmon. He noted, however, that "since [these factors] are not, apparently, subject to any control by man, these problems do not appear . . . to be of prime practical importance." Thus, the "practical" purpose of management?to control salmon production?and the assumption that freshwater was the critical limiting environment (and conveniently, an environment that could be controlled) were mutually reinforcing ideas.

The idea of a constant and benign ocean environment for salmon became formalized in the sustained-yield concept, which proposed an objective, scientific methodology for setting harvest levels. Although biologists were very familiar with the concept by the 1930s, the theory was not developed fully until the work of Beverton and Holt in 1957 (McEvoy 1986). Stock/recruitment models began from the assumption that abundance of a fish population is regulated primarily by density-dependent factors during early life stages. Maximum Sustained Yield was based on a logistic growth curve developed from animal populations held under constant food supply and environmental conditions (Barber 1988; Botkin 1990). It assumed that natural populations reach a stable equilibrium level (carrying capacity) set by available resources in the environment as defined by the logistic curve. In practice, however, spawner-recruit relationships using empirical data relied on many years of observation to show an "average" relationship between population size and the resulting recruitment. Thus, rather than contribute to a better understanding of the effects of environmental change, population models assumed that change was insignificant by averaging conditions over the period of observation (e.g., Cushing 1995). Population models legitimized the benign ocean.

The Ocean is Dynamic (1980s to 1990s)

Within the last two decades, traditional ideas about the ocean and salmon production have undergone a dramatic change in the Pacific Northwest as a result of regional fishery collapse, the increasing risk of extinction of many populations, and new information about the effects of variable ocean conditions on salmon survival. The view of the ocean as a stable and tranquil pasture for salmon has been replaced with the idea of a dynamic, unpredictable, and sometimes hostile ecosystem.

As recently as the 1960s, the assumption of freshwater control of salmon production seemed well supported by the apparent success of new hatchery technology. As Oregon coastal and Columbia River hatcheries began producing large numbers of yearling coho salmon, both the survival rate of hatchery fish and the total return of adult salmon measurably increased (Figure 2). But after 1976, coho populations unexpectedly collapsed despite continued increases in hatchery output, providing the first convincing evidence that mortality factors outside the freshwater environment could be responsible for fluctuations in salmon abundance (Bottom et al. 1986). Successful prediction of adult returns from the previous year's run of precocious males (jacks) further indicated that survival of juvenile coho salmon sometime during their first six months in the ocean could control the production of an entire year class of adult salmon (Gunsolus 1978).

For the last twenty years, scientists have been documenting ocean effects on salmon production involving physical processes over a wide range of spatial scales. Oregon researchers first examined local upwelling processes (Gunsolus 1978; Scarnecchia 1981), which were known to increase nutrient levels and biological productivity at about the time salmon smolts first enter the ocean. Nickelson (1986) found a positive correlation between the percent survival of hatchery coho salmon released off Oregon and average upwelling intensity during the spring and summer. In recent years, the importance of large-scale climatic changes have become obvious as a result of unusually frequent El Ni? activity in the tropics, including two very strong events in 1982-83 and 1997-98. During the 1982-83 event, researchers documented range extensions of marine fishes, birds, and plankton (McClain and Thomas, 1983; Pearcy et al. 1985; Mysak 1986); reduced reproductive success of Oregon seabirds (Graybill and Hodder 1985); and reduced size, fecundity, and survival of adult coho salmon off Oregon (Johnson 1988). These and other climatic effects create an entirely different view of the Pacific Ocean as an interconnected, basin-wide ecosystem in which the background conditions (e.g., species composition, circulation patterns, and biological productivities) continually shift among geographic regions in response to the global heat budget (Barber 1988). Wholesale shifts in the ecological condition of regions around the North Pacific, in turn, alter the environmental context of local salmon populations, whose paths of entry into the ocean are fixed by the location of their home streams.

Factors Affecting Salmon Production

This symposium summarizes recent results that further support the idea that the ocean and estuary are dynamic ecosystems that can produce year-to-year and decade-to-decade variations in salmon production. Among the important findings in this discussion are the following:

Estuary and Plume Effects

Estuarine rearing may improve ocean survival of salmon

Estuaries provide important habitats for juvenile salmon for rearing, adaptation to salt water, and refuge from predators. Size at migration as influenced by estuary rearing conditions may be an important factor affecting salmon survival in the ocean. For example, although 5 different life history types were identified among chinook salmon poulations in Sixes River (Oregon), those juveniles that reared in the estuary for an extended period in late summer and grew to a relatively large size before their ocean migration accounted for 90% of the returning adult spawners (Reimers 1973).

The relative benefits of estuarine rearing may vary from year to year

Manipulation experiments have been used to compare the survival of different groups of chinook salmon that were given or denied access to the Campbell River estuary (British Columbia) before they entered the ocean (e.g., Levings et al. 1989). An intriguing part of this research is that only two of three brood years of chinook salmon showed higher survival as a result of estuarine rearing, suggesting that the relative importance of estuaries to salmon production may vary from year to year as a result of environmental changes. These findings support the notion that estuarine rearing may be just one of a variety of alternative life-history strategies that salmon have acquired to minimize the risk of brood failure in a variable environment.

Nearshore environmental conditions during the first few weeks of ocean life may be critical to salmon survival

The recruitment success of each year class of salmon appears to be established sometime soon after the juveniles enter salt water. Nearshore conditions within the Columbia River plume therefore may be critical to salmon production.

The specific mechanisms affecting salmon survival in the Columbia River estuary and ocean are poorly understood

Understanding the mechanisms of estuarine and marine survival in the Columbia River Basin is limited by the lack of research on basic salmonid ecology within these environments. The inability to readily distinguish wild from hatchery fish complicates assessment of the effects of hatchery production on natural patterns of residence, migration, and habitat use by juvenile salmon in the estuary and plume.

Flow regulation in the Columbia Basin has altered the seasonal hydrograph with effects on salinity, density, and sediment transport

Impoundment of summer flows and releases during the winter by Columbia River dams alter the physical properties and distribution of the plume. The shift in seasonal hydrograph has decreased the volume of Columbia River water transported off the Oregon coast during summer and increased the volume off Washington in winter with effects on salinity distributions north and south of the Columbia River mouth (Williams et al. 1996). Substantial decreases in spring freshet flows from the Columbia River may reduce concentrations of nutrients and food resources in the plume, decrease turbidities, and increase predation pressures to the detriment of salmon survival.

Flow and hatchery release schedules have reduced established patterns of salmon migration and rearing in the Columbia River estuary

By dampening seasonal fluctuations in the hydrograph, dam operations have reduced the diversity of freshwater habitats and variety of flow conditions available to salmon. At the same time, hatchery production and release strategies have been narrowly programmed to fit the schedules of water releases through the dams. River operations thus constrain the historical diversity of rearing behaviors and concentrate salmon migrations through narrow "windows of opportunity" prescribed by the management system (Williams et al. 1996). Such changes may limit the flexibility of Columbia River salmonids to withstand variable estuarine and ocean conditions.

Creation of impoundments and removal of tidal wetlands may have enhanced pelagic food chains and reduced detrital sources that support Columbia River salmon

Loss of vegetated wetland habitats has reduced emergent plant production and availability of macrodetritus in the estuary to the detriment of food chains that support salmon (Sherwood et al. 1990). At the same time, impoundments created by the mainstem dams have increased phytoplankton production and sources of microdetritus available to pelagic food chains. The tremendous expansion of American shad populations in the Columbia River is consistent with this apparent shift in estuarine food chains.

Ocean and Climate Effects

Climatic processes at a Pacific Basin scale can have an overriding influence on local and regional biological production

Changes in global atmospheric circulation, ocean currents, and thermal regimes that can last for decades may set broad limits of salmon productive capacity. Although managers certainly cannot control these large-scale processes, natural variability must be understood to correctly interpret the response of salmon to management actions in the Columbia Basin. The capacity of local populations to realize the full productive potential of any particular climatic state will depend in large part on local habitat, harvest, and hatchery decisions.

The Pacific Basin ecosystem does not move toward a steady state

Changes in the ocean basin ecosystem undermine assumptions of a steady-state background upon which biological interactions take place. The factors affecting salmon production include both density-dependent and density-independent processes. Variability of the Pacific Basin ecosystem suggests that equilibrium population models are not very useful for making long-term conservation decisions.

Climatic changes affecting salmon are often nonlinear and unpredictable.

Decades-long shifts in ocean and climate regimes occur unexpectedly. The unanticipated decline in Oregon salmon production that began after 1976 (Figure 2), for example, coincided with a large-scale shift in oceanic regime. Sufficiently conservative standards of salmon protection are necessary even during a high productivity state to maintain the genetic and life-history diversity needed to withstand subsequent productivity troughs.

Shifts in climate regime change the carrying capacity "rules" for salmon

Changes in climate regime alter the distribution of species, structure of marine food chains, and physical processes. Salmon populations that enter the ocean during different climatic regimes therefore experience an entirely different suite of physical and biological interactions. It should not be surprising, therefore, if simple correlations between salmon production and selected variables (e.g. upwelling, temperature, etc.) established for one climatic state no longer apply during another.

Biological responses in different regions of the North Pacific oscillate out of phase

Atmospheric and oceanic processes in the North Pacific create opposing regional patterns of productivity in the central North Pacific and the California Current region (Oregon, Washington, and California). However, managers should be cautious in assuming that recent decreases in salmon production in the central North Pacific necessarily indicate a return to decades of improving salmon survival off Oregon. It is unclear, for example, how the predicted warming of global climate from steadily increasing concentrations of atmospheric carbon dioxide might alter future productivity oscillations in the North Pacific.

Stream flows and temperatures are affected by the same large-scale processes that control ocean circulation and productivity.

It is not possible to partition the freshwater, estuarine, and ocean factors that affect salmon in part because these environments are embedded within the same regional and global climate systems. In Oregon, stream and ocean conditions that affect salmon survival tend to oscillate in phase with one another: the same climatic conditions that produce warm ocean temperatures and low coastal productivity often coincide with periods of low precipitation, reduced stream flow, and increased river temperatures (Greenland 1994).

Unique local geography and disturbance histories may establish different biotic potentials and responses to large-scale climate change.

Although there is considerable synchrony in patterns of marine survival for southern stocks of salmon (e.g, British Columbia, Washington, Oregon, California), unique geographic influences and local population behaviors may produce complex patterns of environmental change and salmonid response within this broad region. For example, the direction of storm tracks and effects of varied topography on patterns of rainfall and snowmelt may yield different environmental and populations responses to the same regional climate.

Is the River Irrelevant?

The fact that large-scale climatic changes may regulate regional patterns of salmon production has led some people to argue that protection and restoration of freshwater habitats will provide little benefit to salmon. Like some management approaches that ignore the ocean in favor of the stream (because it is not practical to control the ocean), some now claim that the river has become irrelevant (because variability in the ocean is the critical limiting factor)(Figure 1). Opposing management views of river versus ocean dominance flow from the same conceptual framework, which defines resource management as the active removal of production constraints within the single most limiting environment of the salmon life cycle. This view is a simple extension of Leibig's Law of the Minimum, which holds that food or nutrients in least supply control the productivity of a population. In this version of Leibig's Law, the entire environment where the apparent limiting factor occurs becomes the one critical area of management concern above all others. Thus, some argue that the stream environment becomes inconsequential to salmon production if recruitment variations can be associated with ocean conditions.

Solazzi et al. (in review) recently completed stream restoration experiments in several Oregon coastal basins that illustrate some of the flaws in this argument. In the Nestucca River Basin, the number of yearling coho migrants substantially increased in a treatment stream (East Creek) after it received extensive habitat improvements throughout a 2.4 kilometer reach. The treatments, which were designed to improve the quality of overwinter habitat for coho salmon, included construction of 23 dam pools, 8 off-channel rearing ponds, and additions of large wood. Following habitat restoration, the number of coho migrants leaving East Creek increased relative to the number of migrants in the adjacent reference stream (Moon Creek), which received no habitat improvements (Figure 3). Whereas the mean number of salmon migrants in the reference stream steadily declined to near extinction levels during the post-treatment period, the mean number doubled in East Creek following habitat improvements. Moreover, these changes were the direct result of improved overwinter survival of coho salmon. Following treatment, mean overwinter survival for salmon in East Creek increased by 250% (0.11 to 0.39) while the background mean survival in the reference stream declined from a mean of 0.19 to 0.10 (Solazzi et al. in review). These results illustrate that by increasing smolt output, appropriate stream restoration activities may help buffer populations from the additional mortality that also occurs in the estuary and ocean. In fact, not only is the quality of freshwater habitat relevant to salmon conservation, it becomes all the more important during periods when the ocean exerts strong control over salmon populations by decreasing the rate of marine survival.

A principal problem with the conceptual framework of traditional salmon management is that it assumes that the environments at each salmon life stage are separable and independent. Yet the capacity of salmon to reproduce depends upon the connectivity of an entire chain of aquatic habitats from headwater streams to estuary and ocean. The rate of return from a particular brood depends on the cumulative mortality across all these habitats and life-history stages. But this cumulative mortality is not a simple sum of independent mortalities at each successive life-history stage. Mortality from egg to adult and the "carrying capacity" of the environment for salmon are properties of the larger freshwater-estuarine-marine ecosystem (Figure 4). For example, extreme flows, habitat quality, and other factors that contribute to mortality in freshwater determine the ranges of size, times of emigration, and physiological condition of the surviving juveniles that migrate downstream. Slightly different migration times through the estuary and ocean, in turn, may be advantageous in different years, depending upon variations in the timing of coastal upwelling, near-shore ocean temperatures, the location of the Columbia River plume, and so on. Thus, selection pressures at each life stage, which determine the biological characteristics and migration times of the surviving population of migrants, may be as important to the subsequent survival of salmon in the ocean as are the particular environmental conditions within the ocean itself.

Implications of Estuarine and Ocean Variability for Salmon Conservation

Although resource managers cannot control environmental variations in the estuary and ocean, this does mean that they can afford to ignore them. First, the failure to account for natural fluctuations may lead to unwarranted conclusions about the success or failure of restoration efforts. Changes in climate may cause substantial increases or decreases in fish abundance unrelated to management efforts. Many population increases that fishery managers originally attributed to hatchery programs, for example, were in fact the result of environmental changes that naturally increased salmon survival (Bottom et al. 1986; McEvoy 1986; Lichatowich and Nicholas in press). Second, estuary and ocean dynamics that regulate salmon productivity require management responses involving all other aspects of the salmon life cycle that are under human control (Williams et al. 1996). Conservation decisions that are appropriate under one ocean and climate regime may not be appropriate under another. For example, harvest levels must be adjusted to account for changes in survival during periods of low ocean productivity. Furthermore, opposing cycles of salmon abundance between the Central North Pacific and the California Current region (off Washington, Oregon, and California) underscore the importance of stock-specific management of fisheries. Even during periods of high salmon survival off Oregon, harvest limits must ensure that Columbia Basin stocks are not overexploited by northern fisheries trying to compensate for coincidental decreases in Alaska and British Columbia stocks.

Diverse life histories of salmon provide resilience to species in a fluctuating environment. For example, northern and southern coastal chinook stocks in Oregon exhibit different ocean migration patterns such that all fish may not be equally vulnerable to an El Ni? event or local upwelling collapse. Similar migratory differences may explain the substantial decline of tule fall chinook stocks during the 1982-83 El Ni? compared with other stocks from the Columbia Basin that had a more northerly ocean distribution (Johnson 1988). Loss of freshwater habitats, regulation of river flows, and the shift to large-scale production of very few hatchery stocks have all reduced the diversity of salmon life histories in the Columbia River basin and may limit the variety of migratory pathways of salmon into the estuary and ocean. Thus, management manipulations that alter population structure, life histories, or habitat diversity in freshwater may directly alter the capacity of salmon to withstand fluctuations in the estuary and ocean (Williams et al. 1996). Efforts to "stabilize" conditions in freshwater through flow regulation or hatchery programs, for example, may unwittingly eliminate behaviors that buffer salmon production in a variable estuary and ocean.

There are two principal strategies that resource managers can adopt to better accommodate variability in the estuary and ocean and support salmon recovery:

Restore and maintain life-history diversity among salmon populations

A primary goal of restoration in the Columbia Basin should be to promote the greatest possible re-expression of life-history diversity among salmon populations. The impact on salmon diversity of existing management programs for habitat, harvest, hatcheries, and flow regulation should be explicitly evaluated. This review should include an assessment of the potential effects of hatchery programs on historical salmonid migrations, residence times, and habitat use in the estuary and Columbia River plume. At the same time, new restoration activities should be designed to expand flow variations and the variety of quality habitats needed to support diverse salmon life histories. Stream habitat restoration should encompass both tributaries and mainstem areas. Williams et al. (1996) recommend use of water storage and natural runoff to re-establish peak spring flows as a strategy for restructuring and revitalizing mainstem habitats (Williams et al. 1996). Restoration of the spring freshet plume could have additional downstream benefits in the estuary but these potential effects should be evaluated. Within the estuary, tideland marsh and swamp habitats should be restored through dike and tidegate removal to re-establish productive marsh-channel rearing areas and to promote macro-detrital production.

Develop an integrated monitoring and research program that incorporates the entire chain of salmon habitats

Monitoring and research activities in the Columbia River basin should include the estuary and nearshore ocean to evaluate salmonid food webs, growth and residence times, and habitat use. This work should be developed as part of a larger monitoring/research design to provide basic information about salmonid ecology throughout the salmon life cycle and to monitor effectiveness of recovery efforts throughout the basin. Key physical and biological variables should also be identified and monitored to provide indicators of estuarine and near-shore variability needed to guide management activities in the rest of the basin. Finally, research is needed to examine whether specific patterns of salmonid migration and habitat use in the estuary and plume can be linked to specific tributary habitats and stocks upriver. These results would guide habitat protection and restoration efforts upriver toward the areas needed to maintain diverse life histories in the estuary and nearshore ocean. By protecting upstream-downstream habitat linkages, managers can insure that the available habitats and productive capacities of the estuary and Columbia River plume are fully utilized. From this perspective, the diversity of patterns of salmonid use expressed in the estuary and plume could become a system-wide indicator of the success of restoration activities throughout the basin.

References

Barber, W.E. 1988. Maximum sustainable yield lives on. N. Am. J. Fish. Mgmt. 8:153-157.

Botkin, D.B. 1990. Discordant harmonies: A new ecology for the twenty-first century. Oxford University Press, New York, New York.

Bottom, D.L. 1997. To till the water: A history of ideas in fisheries conservation, p. 569-597. In Pacific Salmon and their Ecosystems: Status and Future Options. Edited by D. A. Stouder, P. A. Bisson, and R. J. Naiman. Chapman and Hall, New York.

Bottom, D.L., T.E. Nickelson, and Johnson, S.L. 1986. Research and development of Oregon's coastal salmon stocks. Oregon Department of Fish and Wildlife, Fishery Research Project NA-85-ABD-00111. Annual Progress Report, Portland, OR.

Cushing, D. 1995. Population production and regulation in the sea: A fisheries perspective. Cambridge University Press, Cambridge.

Gilbert, C.H. 1912. Age at maturity of the Pacific Coast salmon of the genus Oncorhynchus. Bull. U.S. Bureau of Fisheries, Vol. XXXII. Government Printing Office, Washington, D.C.

Goode, G.B. 1886. The status of the U.S. Fish Commission in 1884, p. 1139-1180. In Report of the Commissioner for 1884, Part XII. U.S. Commission of Fish and Fisheries. U.S. Government Printing Office, Washington, D.C.

Graybill, M., and Hodder, J. 1985. Effects of the 1982-83 El Ni? on reproduction of six species of seabirds in Oregon, p. 205-210. In El Ni? North: Ni? Effects in the Eastern Subarctic Pacific. Edited by W.S. Wooster and D. L. Fluharty. Washington Sea Grant Program, University of Washington, Seattle, WA.

Greenland, D. 1994. Salmon populations and large-scale atmospheric events, p. 103-114. In Salmon Ecosystem Restoration: Myth and Reality, Proceedings 1994 Northeast Pacific Chinook and Coho Salmon Workshop, November 7-10, 1994, Eugene, OR. Edited by M. Keefe. Oregon Chapter American Fisheries Society, Corvallis, OR.

Gunsolus, R.T. 1978. The status of Oregon coho and recommendations for managing the production, harvest, and escapement of wild and hatchery-reared stocks. Oregon Department of Fish and Wildlife, Clackamas, OR.

Heincke, F. 1898. Naturgeschichte des Herings I. Die Lokalformen und die Wanderungen des Herings in den europaischen Meeren. Abhandlungen der deutschen Seefischereivereins, vol. 2. O. Salle, Berlin.

Higgins, E. 1928. The policy of the Bureau of Fisheries with regard to biological investigations, p. 588-598. In Progress in Biological Inquiries 1926, Including Proceedings of the Divisional Conference January 4 to 7, 1927. Edited by E. Higgins. Bureau of Fisheries Document No. 1029. Annual Report of the Commissioner of Fisheries to the Secretary of Commerce for the Fiscal Year ended June 30, 1927. Washington, D.C.

Hjort, J. 1914. Fluctuations in the great fisheries of northern Europe. Rapp. Proc. Verb. Res. Cons. Perm. Int. Expl. Mer 20: 1-228.

Johnson, S.L. 1988. The effects of the 1983 El Ni? on Oregon's coho (Oncorhynchus kisutch) and chinook (O. tshawytscha) salmon. Fish. Res. 6:105-123.

Kingsland, S.E. 1995. Modeling nature: Episodes in the history of population ecology. University of Chicago Press, Chicago, IL.

Larkin, P.A. 1977. An epitaph for the concept of maximum sustained yield. Trans. Am. Fish. Soc. 106:1-11.

Levings, C. D., C. D. McAllister, J. S. Macdonald, T. J. Brown, M. S. Kotyk, and Kask, B.A. 1989. Chinook salmon (Oncorhynchus tshawytscha) and estuarine habitat: A transfer experiment can help evaluate estuary dependency, p. 116-122. In Proceedings of the National Workshop on Effects of Habitat Alteration on Salmonid Stocks. Edited by C.D. Levings, L.B. Holtby, and M.A. Henderson. Can. Spec. Publ. Fish. Aquat. Sci. 105.

Lichatowich, J.A., and Nicholas, J.W. In Press. Oregon's first century of hatchery intervention in salmon production: Evolution of the hatchery program, legacy of a utilitarian philosophy, and management recommendations. In Proceedings of the Symposium of Biological Interactions of Enhanced Wild Salmonids, June 17-20, 1991. Department of Fisheries and Oceans, Pacific Region, Nanaimo, British Columbia.

Lichatowich, J.A., L.E. Mobrand, R.J. Costello, and Vogel, T.S. 1996. A history of frameworks used in the management of Columbia River chinook salmon. Mobrand Biometrics, Vashon Island, WA.

McClain, D.R., and Thomas D.H. 1983. Year-to-year fluctuations of the California Countercurrent and effects on marine organisms. CalCOFI Rep. 24: 165-181.

Mayr, E. 1982. The growth of biological thought: Diversity, evolution, and inheritance. The Belknap Press of Harvard University Press, Cambridge, MA.

McEvoy, A.F. 1986. The fisherman's problem: Ecology and law in the California fisheries. Cambridge University, New York, NY.

Mysak, L.A. 1986. El Ni?, interannual variability and fisheries in the northeast Pacific Ocean. Can. J. Fish. Aquat. Sci. 43: 464-497.

Nickelson, T.E. 1986. Influences of upwelling, ocean temperature, and smolt abundance on marine survival of coho salmon (Oncorhynchus kisutch) in the Oregon Production Area. Can. J. Fish. Aquat. Sci. 43: 527-535.

Pearcy, W.G., J. Fisher, R. Brodeur, and Johnson, S. 1985. Effects of the 1983 El Ni? on coastal nekton off Oregon and Washington, p. 188-204. In El Ni? North: Ni? Effects in the Eastern Subarctic Pacific. Edited by W.S. Wooster and L. Fluharty. Washington Sea Grant Program, University of Washington, Seattle, WA.

Reimers, P.E. 1973. The length of residence of juvenile fall chinook salmon in Sixes River, Oregon. Res. Rep. Fish Comm. of Oregon 4: 3-43.

Rich, W.H. 1928. General review?The International Pacific Salmon Investigation Federation, p. 633-637. In Progress in Biological Inquiries 1926, including Proceedings of the Divisional Conference January 4 to 7, 1927. Edited by E. Higgins. Bureau of Fisheries Document No. 1029. Annual Report of the Commissioner of Fisheries to the Secretary of Commerce for the Fiscal Year ended June 30, 1927. Washington, D.C.

Rich, W.H. 1939. Local populations and migration in relation to the conservation of Pacific salmon in the western states and Alaska. Contribution No. 1. Fish Commission of Oregon, Salem, OR.

Scarnecchia, D.L. 1981. Effects of streamflow and upwelling on yield of wild coho salmon (Oncorhynchus kisutch) in Oregon. Can. J. Fish. Aquat. Sci. 38: 471-475.

Sherwood, C.P., D.A. Jay, R.B. Harvey, P. Hamilton, and Simenstad, C.A. 1990. Historical changes in the Columbia River estuary. Prog. Oceanog. 25: 299-352.

Sinclair, M. 1988. Marine populations: An essay on population regulation and speciation. Washington Sea Grant Program. University of Washington Press, Seattle, WA.

Smith, E.V. 1920. The taking of immature salmon in the waters of the state of Washington. Washington Department of Fisheries, Olympia, WA.

Solazzi, M.F., T.E. Nickelson, S.L. Johnson, and Rodgers, J.D. In Review. Effects of increasing winter rearing habitat on production of salmonids in two coastal Oregon streams.

Thompson, W.F. 1919. The scientific investigation of marine fisheries, as related to the work of the Fish and Game Commission in Southern California. U.S. Fishery Bulletin No. 2. State of California Fish and Game Commission, Sacramento, CA.

Williams, R., L.D. Calvin, C.C. Coutant, M.W. Erho, J.A. Lichatowich, W.J. Liss, W.E. McConnaha, P.R. Mundy, J.A. Stanford, R.P. Whitney, D.L. Bottom, and C.A. Frissell. 1996. Return to the river: Restoration of salmonid fishes in the Columbia River ecosystem. Northwest Power Planning Council, Portland, OR.

Additional illustrations from this presentation

< paper: preparing for the future | paper: changes in size at maturity >