Sarah K. Gaichas of the Alaska Fisheries Science Center, Seattle, Washington, for her paper (coauthored with Kerim Aydin and Bob Francis) entitled "Trade-offs between fisheries and marine mammals in Gulf of Alaska and Eastern Bering Sea"

Teresa Ish of the University of California, Santa Cruz, for her poster "Environment, krill, and squid: from life history to fisheries"

Abstracts
(alphabetical by author)

Because oceanographic processes are important in determining distributions of larvae, understanding oceanographic processes is important in determining which locations are likely to be "sources" and "sinks" for fish larvae. However, whereas delivery of larvae by oceanographic processes is important for considering what areas and habitats are most in need of management as nurseries, for many species benthic features must be of paramount importance, because of benthic-associated processes (e.g., territoriality, feeding and refugia) and species-specific ontogenetic shifts in habitat selection. Habitat management must identify and include essential nursery habitat that is connected to habitats associated with subsequent life stages. Habitat-associated features of nursery habitats have a powerful influence on recruitment, with effects on movement, growth, and survival through the postsettlement transition and juvenile phase of life history. The demographic consequences of these benthic processes project forward to influence abundances of adult populations of fishes and reflect backward to influence settlement (through habitat selection). Additional implications for our understanding of community structure, and for proper system-level management will be revealed by research on the importance of habitat connectivity to marine and estuarine fishes.

In the preparation of the paper it was necessary to review a number of background papers and documents regarding ecosystem principles, properties and management. The exercise gave the author a sense of ecological rejuvenation, but I found myself struggling with the growing ecological jargon. After reading the Ecosystem Based Management report prepared for the US Congress some of the terms used and stated views regarding ecosystems I became demoralized and decided that I was just not with it, politically incorrect, consumed with an educational bias, and senile--perhaps all of the above. In an effort to find a sense of reality the author distributed a number of the stated definitions and scientific concepts to a number of associates. He found that he was not alone and that uncertainty of terms and concern regarding promoted ecological concepts was wide spread. The paper discusses some of these terms, concepts as well as ecological principles and conclusion put forth by the Ecosystem Principle Advisor Panel and the potential confusion it may evoke as well as its positive contributions. It concludes with comments on the nature of trade-offs anticipated and their evolution.

The increased examination of multispecies trophic modeling in Large Marine Ecosystem (LME) fisheries has led to speculation that we can tune these ecosystems through the manipulation of trophic cascades. For example, it has been suggested that culling marine mammals may greatly increase fish available to fisheries; conversely, a decrease in fishing might be seen as increasing food supply for endangered species. However, the discussion of such trade-offs presupposes that we have a reasonable degree of command and control over population-scale trophic cascades and that, in cases where we don't, the simple expedient of improving data collection will improve our control. I review the current "state-of-the-art" multispecies models (both biomass-dynamics and age-structured) that have been used to examine some North Pacific fisheries, focusing on strategic (long-term) projections. The review suggests that, although local targeted management of trophic flows may be effective, the population-scale trade-offs lie not in selecting a "desired" set of biomass components but in selecting levels of uncertainty. We should phrase alternatives as trade-offs between preserving our short-term control and abdicating such control by increasing chances of "system flips" or regime changes. Precaution has long been a watchword for single-species management, suggesting that the development of ecosystem management should be less about directing production and more about finding quantitative measures of structural and dynamic integrity. The development of multispecies modeling is necessary for such quantification, as we should control ourselves away from causing uncertain trophic cascades rather than "controlling" marine ecosystems into them.

In this paper, we describe some recent experiences with ecosystem trade-offs in managing New England fisheries. Conflicting legislative mandates to conserve fishery resources while sustaining fisheries and protecting essential fish habitat and threatened species are underlying themes. For sea scallops, spatial management approaches are promising but require trade-offs with groundfish closed areas and protection of essential fish habitat. Rebuilding groundfish stocks and protecting threatened species have conflicted with traditional policies of no catch quotas and unrestricted access to fishing grounds. For Atlantic mackerel and herring, we describe trade-offs between maintaining these abundant pelagic stocks versus minimizing forgone yields and collecting information needed for assessment and understanding trophic interactions. Although applying a holistic ecosystem approach to manage New England fisheries should reduce conflicts and improve stakeholder satisfaction, such an approach will in practice be tempered by political interests and the public's willingness to pay for increased management and resource costs.

Laguna Madre, in southeastern Texas, comprises a portion of the largest coastal hypersaline lagoon system in the world. Extreme variability in precipitation and evapotranspiration and tropical storms or hurricanes may cause major changes to the system. No rivers drain into the lagoon. The ecosystem is dominated by sea-grass habitat in the lagoon and the highly productive wind-tidal flats. The latter convert plant biomass to animal biomass where the water meets the land, filling the role marshes play in less arid climates. Sea-grass meadows are extremely valuable habitats providing complex structure in the habitat, supplying nursery areas, refuge, and rich foraging grounds for a variety of estuarine fish and invertebrates, including species of commercial and recreational importance. Shorebirds and wading birds abound over large undisturbed wetland complexes. Fish biodiversity comprises 79 species. The historically highly productive commercial fisheries have now given way to some of the best recreational fishing for red drum, black drum, and spotted sea trout in North America. Thirty-seven species of fish and invertebrates have ecological, commercial, or recreational value or are indicator species of environmental stress. Conservation issues and evaluation of long-term effects for protection or sustainable use of the Laguna Madre were examined using a mass-balance trophic model (ECOPATH), where fluxes among the main functional groups of the ecosystem were determined. Particular emphases were directed toward maximizing yields, conservation of biodiversity, areas of concern, human issues, and their trade-offs in the management of this ecosystem. Maximum annual average yield, changes in fishing mortality (F) of the sports fisheries and shrimp stocks, as well as for redhead duck hunting, were simulated with the aid of Ecosim and Ecospace, the time-dynamic and spatial versions of the model, which was run for 50-year periods, allowing evaluation of changes in overall biomass composition. Changes in shrimp biomass resulting from increases in F produced important perturbations in the system. Artificial increases in F over some nonexploited predators produced significant increments in shrimp and predator biomasses with economic importance. The model outputs allowed identification of management options.

Numerous trade-offs must be considered for management of marine ecosystems; especially crucial are those that relate to economical, social, and existence values. An important aspect of the problem is how we chose between having depleted systems with valuable invertebrate fisheries and restored systems with high standing fish stocks. Empirical evidence indicates that, as the fisheries in an area develop, the catch levels increase while their trophic level decreases. Eventually, catches may or may not decrease, population crashes may occur, and the risk of alternate stable states looms. The present study in this context examines current knowledge of the effect of predator removal, asking the question: would we be better off with restored ecosystems?

In multispecies fish communities, predation levels change dynamically in response to changes in the abundance of predator and prey species, as influenced by the fisheries that exploit them. In addition to community-level metrics, the abundance of each species must be tracked relative to its biological reference point. In situations with many interacting species exploited by multiple fishing fleets, it can be complicated to illustrate how the effort of each fleet will affect the abundance of each species. We have adapted the AMOEBA approach to graph the reference levels of multiple interacting species exploited by multiple fleets. This method is illustrated with 10 species and eight fishing fleets in the North Sea. We fit a relatively simple response-surface model to the predictions of a fully age-structured multispecies model. The response-surface model links the AMOEBA for fishing effort to separate AMOEBAs for spawning stock biomass, fishing mortality, and yield. Ordination is used to give the shape of the AMOEBAs functional meaning by relating fish species to the fleets that catch them. The aim is to present the results of dynamic multispecies models in a format that can be readily understood by decision makers. Interactive versions of the AMOEBAs can be used to identify desirable combinations of effort levels and to test the compatibility of the set of single-species biological reference points.

Fishery management plans ignore the potential for evolutionary change in harvestable biomass, in part because proof that size-selective fishing mortality genetically alters the productive capacity of fish stocks is lacking. We approached this problem experimentally by subjecting populations of an exploited fish (Menidia menidia) to large, small, or random size-selective harvest of adults over four generations. Harvested biomass evolved rapidly in directions counter to the size-dependent force of fishing mortality. Large-harvested populations initially produced the highest catch but quickly evolved a lower yield and spawning stock biomass than controls. Small-harvested populations did the reverse. These shifts were caused by selection of genotypes with slower or faster rates of food consumption and somatic growth. Although our experiment involved captive populations, there is compelling new evidence that natural fish stocks display genetically variable rates of somatic growth and are capable of rapid life history evolution in the wild. Management regimes that account for evolutionary dynamics are necessary to preserve natural genetic variation in physiological traits that are the foundation of long-term sustainable yield.

Harvesting of marine living resources in the Southern Ocean is managed by the Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR). CCAMLR is widely known for its ecosystem-approach to managing fisheries, which includesthe maintenance of food webs in the conservation objectives. CCAMLR has developed a precautionary approach for taking account of predators when setting catch limits for krill, the most important prey species in the Antarctic marine ecosystem. It has also established an ecosystem-monitoring program to help determine when harvesting is affecting predators. More recently, the Scientific Committee of CCAMLR has been exploring different approaches for evaluating the effects of harvesting on predators of krill, including predictive models as well as more refined monitoring tools. This paper will explore these new approaches, illustrating methods for incorporating the uncertainties in trophic linkages when formulating food-web models and when designing monitoring programs. It will also discuss the emerging framework for underpinning the monitoring and management of the effects of harvesting on food webs in the Southern Ocean. In particular, the paper will evaluate the types of conservation objectives for predators of fished species and the approaches that could be considered for achieving the conservation objectives.

An important goal of ecosystem modeling is to assess potential conflicts among alternative ecosystem management options. The Lake Superior ecosystem provides an excellent setting in which to test whether ecosystem models can detect ecosystem management conflicts. Overfishing and invasions by sea lamprey and rainbow smelt caused lake-wide collapses of the dominant piscivore (lake trout) and zooplanktivore (whitefish and lake herring) species. Restoration policies put a higher priority on restoring lake trout (the predator) rather than lake herring (the prey) and proceeded along a path of artificial enhancement and reduced fishing mortality for lake trout only. Although lake trout have since recovered to historical highs, the lake herring population remains at critically low levels. We developed an Ecopath with Ecosim model of the Lake Superior fish community (1929-1998) to determine the relative impacts of lake trout enhancement programs, fish community dynamics, and fishing mortality on the lake herring population. Although much of the observed variation in forage fish biomass was accounted for by the simulated dynamics of top trophic level groups, we found little evidence that any factor besides fishing could account for lake herring dynamics. Difficulty representing trophic dynamics at or near critically low population levels raises some doubt as to our ability to model and respond to important ecosystem trade-offs.

Identifying trade-offs in fisheries presumes some knowledge regarding interactions between harvested species, yet there is considerable uncertainly regarding the type and nature of food web interactions. I show some examples of this uncertainty from analyses of the Baltic Sea pelagic food web, and I explore the potential biases resulting from using models that do not account for these interactions. The Baltic Sea pelagic food web comprises a top predator, cod, and two prey species, sprat and herring. My analysis has indicated that cod-herring interactions are donor-controlled, whereas cod-sprat interactions follow the commonly assumed "mass-action"-based functional response. Moreover, both sprat and herring are important predators on cod eggs, potentially resulting in a cultivation/depensation effect. I used a series of simulations to evaluate the predictions based on naïve models that either assume mass-action functional responses or do not include egg predation. "Projection models" in which predictions are based on a mechanistic construction of food web interactions, are prone to very large biases when they do not fully or accurately describe the nature of predator-prey interactions. However, "inverse models," where parameter estimates are derived from fitting the model to time series data, have considerably less bias. Specifically, when egg-predation is important, a simple multispecies model correctly identifies the optimal effort allocation, whereas single-species models were much less accurate. However, both single-species and multispecies inverse models were not able to identify accurately the optimal effort allocation when the presumed functional response was not correct. A simple extension of the model to account for this interaction was able to predict accurately the true population dynamics. These results suggest that inverse-modeling methods are generally robust and that the inherent uncertainty in food web responses to harvesting is not a legitimate excuse for eschewing ecosystem-based approaches to fisheries management.

Most North Pacific kelp forest ecosystems are organized around a 3-tiered trophic cascade in which sea otter predation on sea urchins helps to maintain lush algal assemblages, in turn influencing other species through a labyrinthine and strongly connected interaction web. Here I consider 3 issues bearing on the nature and strength of these interactions. First, where did the key players come from and how long have they been together? The fossil record indicates that ancestral sea otters and strongylocentrotid sea urchins have co-existed in the North Pacific basin since at least the late Miocene. The evolutionary history of kelps is less clear owing to the near absence of a fossil record. However, a variety of evidence, including changes in the shape of limpet apertures, the distribution and abundance of herbivorous marine mammals, and increased body size in algivorous abalones, indicates that the kelps also have a North Pacific center of radiation. Second, do all sea otters contribute equally to the trophic cascade or are some individuals more important than others? Longitudinal records from tagged sea otters in Monterey Bay indicate that most animals specialize on one to several prey types, with little or no inter-individual dietary overlap in many instances. These findings indicate that per capita interaction strengths vary considerably among individual sea otters. Third, to what extent do in situ processes regulate kelp forest ecosystems? Sea otter populations in western Alaska abruptly collapsed during the 1990s, thus causing sea urchin numbers to rise and the kelp beds to decline. The likely cause of these events was increased predation by killer whales on sea otters. A variety of anthropogenic changes in the open ocean emanating from the overexploitation of living resources ultimately appear to be responsible for these changes. These various dimensions of complexity (time, space, and individuality) provide substance to our understanding of the dynamics of kelp forest ecosystems.

The issue of uncertainty and risk within the context of single-species assessment and management has attracted considerable attention, and it is now commonplace to have management advice framed in terms of a risk assessment. In particular, the issues of parameter uncertainty and intrinsic variability of marine populations have attracted considerable attention. Consideration of model uncertainty within the single-species context has been less extensively explored. Moving toward ecosystem-based management will necessarily entail additional consideration of uncertainty in model structures as well as strategies for coping with the increased uncertainty accompanying estimation of interspecific interactions, habitat-dependent productivity, and other factors. Strategies for assessing trade-offs in an ecosystem-based management framework and coping with uncertainty will require an adaptive-management approach. In this paper, the components of uncertainty and risk in an ecosystem context will be explored, and their implications for setting management objectives will be explored. Clear specification of these objectives will be essential in assessing management trade-offs in the ecosystem context.

How do we manage an endangered species of fish fed upon by mammals protected by the Marine Mammal Protection Act? How do we resolve the conflict between any effects of predators on prey populations and vice versa? This quandary exemplifies many dilemmas involving tradeoffs that are humanly impossible to resolve.

In systemic management the resolution to such dilemmas is found in natural systems themselves. Guidance is based on empirical information regarding balances achieved among the opposing forces of nature to account for all tradeoffs--all of reality. All consequences are accounted for, as are all aspects of the interconnected nature of the complexity of reality.

A key difference between systemic management and conventional management is the guiding information used. Conventional management assumes that we can guide ourselves through thought, models, meetings, lists, consensus, or stake-holder input. All are subject to irresolvable conflict and never fully account for complexity. By contrast, systemic management makes use of empirical information for guidance, using science to make measurements directly relevant to specific management questions. Resulting data are used to set goals, standards, and reference points for management that seeks to put humans into the normal range of empirically observed natural variation to achieve sustainability.

The continental shelves of the Gulf of Alaska and Eastern Bering Sea are highly productive marine ecosystems which currently support some of the most economically important fisheries in the United States. While the major commercial fish stocks in these regions are generally considered healthy and well-managed, some marine mammal populations have declined during the period that the fisheries have flourished. Management objectives for fisheries and marine mammals were developed separately from single-species points of view and are based on different legislation. Because some hypothesize that fisheries have indirectly affected marine mammals in the Gulf of Alaska and Eastern Bering Sea by removing their prey, completely separate management for fisheries and marine mammals is no longer possible in these regions. In the single -species context, trade-offs are daunting: maximizing marine mammal and fisheries objectives simultaneously is challenging, there is considerable uncertainty on both "sides," and the system as a whole is still not considered in the discussion (problems can be displaced elsewhere). Here, we explore whether shifting our perspective to the ecosystem level improves our ability to understand the relationships between fisheries, marine mammals, and other components of these systems to help determine which single-species objectives are achievable.

Legislation in New Zealand demands that fishing activities have a "no more than minor" impact on associated and dependent species. Hence, consideration on the effects of fishing on parts of the environment must at least be addressed. However, this legislative requirement has either been ignored, or when addressed, sparked debate and confusion. The same legislative requirement governs management of the significant inshore and offshore shellfish farming sector. By contrast to the wildstock fisheries, significant advances have been made in ecosystem management of shellfish culture activities. In some cases, this holistic approach has encompassed associated demersal wildstock fisheries. This approach is underpinned by a number of scientific tools, including Ecopath and in-house developed models. Details of our experiences with ecosystem management of inshore marine resources will be presented.

Numerous computer programs are now available to ecologists, corresponding to different theoretical perspectives on problems faced in the field. Despite having been at the forefront of the revolution in mathematical ecology in the 1970s, qualitative analysis, a.k.a. loop analysis , has received short shrift since then because of the difficulties of dealing with an analytical approach on computers as well as the limited mathematics of the approach.

We present novel advances in mathematics that greatly increase the predictive power of qualitative analysis, show applications in community ecology, and demonstrate a computer program that links an intuitive graphic representation of a model to a mathematical analysis. We suggest that these contributions will complement other tools available to ecologists and enhance experimental design and adaptive management.

While it is widely recognized that existing fisheries management systems have largely failed, the public and almost the entire scientific community believe this failure is due to overfishing, and the almost universal prescriptions for solving the problem are the precautionary approach, marine protected areas and ecosystem management. These prescriptions fail to recognize that the real problem is overcapitalization and the race-for-fish, and these "solutions" treat a symptom not the problem. Overfishing in the U.S. accounts for about a 15% loss of potential yield, a minor concern, while social and economic losses due to the race-for-fish are much greater. Solutions to the real fisheries problem do exist and have the common characteristic of changing the incentives in the fishery management system so that what is good for an individual or group is good for society. Various forms of these systems are in place in many places around the world, including community ownership of fishing grounds, cooperative fisheries, rights based fishing including ITQs. The failure of the existing system is illustrated by the fishery on the shelf of the west coast has recently been closed due to concern for several stocks classified as overfished. I show that this is irrational from almost any social and economic perspective and a true ecosystem based management would in fact likely perpetuate overfishing on some stocks.

The California market squid, Loligo opalescens, is an annual, semelparous species that spends most of its life offshore, growing, then moves inshore onto the continental shelf to reproduce. The California market squid is extremely important to the California Current ecosystem, providing food for a variety of marine birds and mammals, as well as commercially important fish species, such as rockfish, salmon, and anchovy. Furthermore, squid feed extensively on krill, another important developing California fishery. The fishery for squid is also the largest fishery in both tonnage and value in California. All fishing occurs inshore during the reproductive phase of the squid's life cycle with very little, to no, regulation. We will first review the history of this fishery and provide an analysis of recent catch information. In order to analyze the effects of environmental variation on squid life history and behavior, we developed a model to predict the pattern of growth and in-shore migration. We are able to predict the continuous movement of squid inshore, as opposed to a single wave, thus providing a mechanistic explanation of the statistical data of squid landings throughout the fishing season. We will then explore the management implications of these results.

The biology of the ocean is very rapidly changing state from complex to simple, from 3-dimemnsional to 2-dimensional, from heterogeneous to homogeneous, from food chains capped by large vertebrates to those capped by small invertebrates, and by explosive increases in microbial biomass. The human drivers are overfishing, pollution, introduced species, aquaculture, and climate change -- probably in that order of importance historically if not actually. Rates of change are accelerating and may be difficult to reverse. The rise of jellyfish and bacteria and demise of animals effectively erase half a billion years of Phanerozoic evolution, taking us back to the latest Precambrian before the explosion of metazoan life. What kinds of species will dominate the ocean? What are the most likely future scenarios, and what are the implications for our use of the oceans and our way of life? Fishers have found good markets for the jellyfish, but not yet for the bacteria. Do we even want to try?

Estuarine shrimp trawling in North Carolina is controversial due to the destructive nature of the gear and the high bycatch generated by the fishery. New evidence suggests that the discards of the shrimp fishery may contribute to the success of blue crab populations through food subsidization. Blue crab and shrimp are North Carolina's first and second most profitable commercial fisheries, respectively. Therefore, managers must balance the economic values of these two fisheries with the trawler-induced mortalities of fish, bivalves, and other estuarine flora and fauna. We constructed a trophic model of the Neuse River Estuary, NC, and used network analysis to evaluate system characteristics. Biomass and production values were measured within the estuary and fishery-influenced feeding relationships were determined through field experiments. We then altered the base model to generate estimates of changes in ecosystem characteristics and economic yield with and without shrimp trawling, with and without discarding, and at different levels of trawling effort. Thus, we can generate preliminary assessments of proposed management strategies for estuarine shrimp trawling. This work demonstrates that conceptual and numeric models can provide estimates of economic and ecological trade-offs and reveal unanticipated ecosystem consequences of fishing and fisheries management strategies.

There is broad consensus that fisheries cause major impacts to ecosystems and should be managed sustainably, which requires much more information than we now have. Few fisheries have the legal mandate for ecosystem-based management, nor to apply precautionary management when information is lacking. Thus there is little incentive for fishers to demand improved information. The California Marine Life Management Act (1999) requires the maintenance of ecosystem health and diversity. The complexity of California nearshore ecosystems and fisheries that exploit them make this particularly challenging. We present the key elements, scientific rationale, and implementation plan for a transition from information-poor, precautionary management, to information-rich, spatially explicit ecosystem-based management in the California nearshore fishery. Each stage of management is implemented via a set of monitored parameters, with explicit triggers for a range of interventions, from catch and gear restrictions to areal closures. Marine reserves serve as reference points in BACI experimental designs, a function often overlooked when discussing the conservation value of MPA's. The complexity of scientific monitoring, the statistical power of the monitoring design, the benefits to consumptive and non-consumptive values, and accountability all increase from Stage 1 (information-poor management) through Stage 2 (information-moderate) to Stage 3 (information-rich). The most significant scientific hurdle comes with Stage 3, when ecosystem and environmental variability effects are taken into consideration. The greatest socioeconomic hurdle is at the outset, in convincing all stakeholders that basing resource access on a richly detailed and wide open information system is in everyone's best interest.

Human activities have influenced ecosystems for thousands of years, but organisms created using the newest molecular biotechnologies have unprecedented potential to affect aquatic ecosystems in unintended ways. Genetically engineered salmon, now in the Food and Drug Administration review process, are proposed for deployment in commercial aquaculture. If released into ecosystems, these novel organisms would have uncertain potential effects on ecological processes. However, genetically engineered fish pose risk rather than certainty of harm, and their regulation must deal with trade-offs between caution and certainty, between open discussion and secrecy of data, and between acceptance and control of ecological risk. Based upon field interviews and regulatory materials, this study argues that the current regulatory approach is inadequate to address these uncertain ecological risks. Proponents of the technology argue that ecosystems are fundamentally balanced and resilient and that the risks are insignificant because of the fishes’ reduced ecosystem fitness. Skeptics maintain that ecosystems are characterized by instability and contingency rather than equilibrium, and that a small number of ecologically fit organisms may be enough to change the state of an ecosystem if conditions are favorable. The current regulatory approach hampers public discussion of data about the risks, limits the role of the agencies with the greatest expertise in fisheries and ecological sciences, and makes precautionary action more difficult in the absence of quantifiable harm that has already occurred.

Recently, NMFS was petitioned to list North Atlantic white marlin as an Endangered Species. If approved, other marlin species will likely soon follow. Although no highly migratory pelagic fish species has been listed by our ESA; the example requires attention and raises opportunity. The ESA requires assessment of status and trends for listed species. It also requires protection from exploitation which will severely restrict US commercial and recreational fisheries. The trade-off is clear--the species are rare, they can’t be caught by US fisheries, but must be assessed. We conducted analyses based on an alternative strategy that recognizes the value of tag-release programs for recreational fisheries such as those promoted through The Billfish Foundation and contrasts that with the cost of conducting assessments based on commercial catches. Bycatch by foreign longline fleets account for more than 90% of marlin mortality which might be reduced by up to 50% through a "buy back" program that compensates longliners for the catch lost due to eliminating hooks nearest buoys. We simulated ecological consequences of increased marlin abundance using an Ecosim model developed for the central Pacific. More marlin caused reduction in mahimahi which reduced predation on juvenile bigeye tuna. The ecological feedback fueled yields of highly valued species and provides a bases for estimating the economic benefits owing to complex food web interactions. This presentation will elaborate on the speculation above.

From the second law of thermodynamics we know there is a finite amount of energy and biomass. Thus, limits to biomass production lead to some hard choices about the particular species composition to be allocated from the total biomass available in a particular ecosystem. Here I examine biomass allocation using an aggregate biomass and functional guild approach for various scenarios from the northwest Atlantic. I also include both ecological and abiotic constraints in models of biomass production and allocation. Assorted model results all show greater stability of biomass at the guild or aggregate level when compared to the species level, irrespective of species composition within a guild. Fishing pressure and abiotic conditions are typically the more dominant factors changing total guild biomass when compared to internal ecological dynamics. We often assume that we can direct the response of an ecosystem toward a particular end when in fact it is unclear if doing so is possible for such complex, open ecosystems. Probably the best we can do now in an EBFM context is establish conditions in accordance with our understanding of how the ecosystem functions and consistent with our ecosystem goals so that the system has a greater probability of responding in the desired manner.

In the Strait of Georgia, British Columbia, complex trade-offs exist among values obtained from commercial and sport fisheries, marine-mammal conservation, salmon enhancement, and rebuilding of depleted groundfish stocks. Since the 1970’s, harbor seal populations have increased nearly 35-fold after predator-control programs have ceased. The resulting predation mortality on salmon populations is estimated to have increased 8-fold for coho and 10-fold for chinook. Salmonid enhancement programs have artificially increased juvenile salmon abundance 3-fold, which has probably caused increased competition for limited food supplies. Climate variability also plays a role in reduced marine survival rates for salmon. Intensive fisheries for lower trophic levels such as euphausiids pose additional threats for wild salmon stocks and depleted groundfish populations. We examined the potential magnitude and direction of these trade-offs using a dynamic ecosystem model that was fitted to historical time-series data. Time-series data were best supported by a declining trend in primary productivity. Uncertainty about future primary production dominated management tactics needed to respond to ecosystem trade-offs. If primary production remained low, hatchery programs and intensive fisheries for euphausiids reduced survival even further for salmon, and groundfish stocks failed to recover even in the absence of fishing. If primary production continued to decline, rebuilding salmon stocks would require a combined predator control program that reduced both Pacific hake and harbor seal abundance. Intensive euphausiid fisheries have a negative impact on Pacific hake but jeopardize the rebuilding of salmon and groundfish populations. Rebuilding groundfish stocks required reduced fishing mortality under all production scenarios.

Marine reserves are established to manage fisheries and conserve whole species assemblages. However, protecting one species may influence other taxa linked to the target species through predator-prey, competitive or positive interactions. Conversely, efficacy in protecting the target species may depend on the response of the species prey, competitors, or predators. Predator-prey models exploring the effects of different reserve designs show that adding species interactions can generate complex responses to protection, where trophic interactions change not only the magnitude but also the direction of the species response to protection. A review of empirical data is conducted to identify under what conditions species may decline following the establishment of marine reserves. Modeling and meta-analysis of monitoring data can help determine the range of possible responses of sets of interacting species to protection, and yield simple rules for reserve design.

Extraction of organisms from the sea limits energy for other parts of marine communities. Billions of humans remove food from oceans, thereby modifying ecosystems and other human opportunities. A fundamental reason for fisheries and ecological degradation is the reluctance to account for such trade-offs. I used a policy analysis routine in Ecopath with Ecosim to explore the interplay of "economic," "social" (employment), and "ecological" objectives in fisheries-policy alternatives for Prince William Sound. Maximization of economic or employment objectives made pinnipeds, halibut, and lingcod go extinct, while Pacific cod and sablefish increased in abundance. Maximizing ecological objectives made porpoise, pinnipeds, orcas, and seabirds increase, while Pacific cod and sablefish decreased. Increases of top predators with simulated decreases in fishing indicated competition between fisheries and predators (and reflected direct killing in the case of pinnipeds). The relative weighting of ecological considerations was increased considerably over those of economic and social considerations to achieve numerical stability in pinnipeds--the chosen assessment endpoint. Fisheries catches were predicted to decrease in almost all scenarios, indicating that current levels of fishing are unsustainable in the context of conservation policy objectives.

Marine reserves may be an effective way to simultaneously protect entire ecosystems. To date, however, many scientists and policymakers have uncritically embraced the beneficial effects of reserves. The stakes are potentially high. If reserves do not measure-up to expectations, future designation of reserves can be jeopardized, along with the ecosystems they were intended to protect. Recent studies and a meta-analysis suggest that reserves are a powerful and effective management tool; however, 85% of existing assessments rely on a Control-Impact design. At best, these designs document that reserves have local effects (within the boundaries of the reserve); at worst they overestimate the effect of the reserves by confounding the effect with pre-existing differences among sites (as would be expected if reserves are established in the best remaining habitat). The more controversial (and important) aspect of marine reserves (i.e., "spillover") has not been commonly assessed and demands a different approach. We evaluate weaknesses of existing assessments and then lay out the application of the BACIPS assessment design. Our goal is not to detract from the expeditious establishment of marine reserves, but rather to suggest ways in which their establishment can facilitate sound scientific assessment while embracing the concept of adaptive management.

[Abstract not available]

In the 25 years since the Magnuson Fisheries Conservation and Management Act was implemented, substantial agreement has been reached about how to manage single-species fisheries in the United States. Biological reference points, such as the biomass that will produce maximum sustained fisheries yield, are estimated from fairly standardized kinds of fisheries models, and management regulations such as quotas are set according to control rules wherein these biological reference points are important benchmarks. Debate about the specifics of fisheries management takes place with this basic framework. However, the objectives, principles, goals, and scientific methodology of ecosystem-based management are in an early stage of development and utilization, and no standardized definition or approach currently exists. Ecosystem-based management regimes may run the gamut from a suite of single-species reference points (e.g., all species within the system must be kept above some minimum biomass level) to that based upon reference points that measure some level of ecosystem function (e.g., measures of biodiversity or mean trophic level). Management regimes that do not rely upon quantitative reference points, such as systems of marine protected areas, gear restrictions, or community-based management, have also been referred to as ecosystem-based management. To include ecosystem values such as biodiversity and ecosystem function in fisheries management under U.S. fisheries law, it will be necessary to evolve consensus toward a standardized, practical approach to ecosystem management. This paper presents an overview of ecosystem-based approaches to fisheries management and forges a new direction for the evolution of such a consensus.

The potential impact of disease on population dynamics of marine species has been debated for some time. Although recent studies and syntheses have suggested that outbreaks of pathogens and parasites of marine invertebrate and vertebrate fisheries may be increasing, little consideration has been given to how disease may affect our ability to restore and mange exploited species. Unsustainable harvest and habitat degradation may not be the only reasons for current population reductions, for many species increased prevalence and/or virulence of pathogens may also be a significant contributor to population declines. Unfortunately, few empirical data sets exist to evaluate the potential consequences of disease on restoration and management of marine fisheries. Using examples primarily from marine invertebrate fisheries, I discuss how management and restoration strategies may be complicated by disease outbreaks. For example, if disease is identified as a current factor in the decline of a fishery, adoption of marine protected areas may contradict current strategies for managing around disease. Given that disease dynamics for many populations may be related to habitat degradation, species introductions, climate change, and expansion of aquaculture facilities, disease relationships need to be identified and addressed within the broad mandate of ecosystem management.

People matter for ecosystem based management - unless the socioeconomic impacts of marine management measures (e.g. marine protected areas, capacity reductions, bycatch measures, or individual quota) are addressed, there are likely to be obstacles to successful implementation. We present our work on a coast-wide analytical framework for assessing past trends and predicted impacts on coastal communities on the West Coast of the United States resulting from changes in fishing behavior in response to management measures. Harvesting a variety of databases containing ecological, fishery-dependent and socioeconomic data, we built an extensive relational database, standardized the data, and conducted a meta-analysis on them. We mapped trends over time and space in a geographic information system (GIS) that covers the length of the West Coast from Washington to California and covers the entire exclusive economic zone (EEZ). Using this framework, we analyzed various fleet reduction scenarios in terms of their effects on habitat areas and types, economic activity on shore, and likely implications for the remaining industry. The tools are both intuitive and analytically powerful to conduct decision-support analysis in the field, for example in participatory decision-making processes, and can accommodate qualitative data sources such as traditional ecological knowledge about resources and ecosystem linkages.

Fisheries management decisions are often portrayed as a trade-off between conservation and exploitation. The concept of the burden of proof illustrates this trade-off; one side presses for limiting exploitation to levels known to be safe, and the other presses for tolerance of exploitation until data identify a problem. This perceived trade-off is a myth. In reality, ecosystem-based management systems can perform as well as or better than conventional exploitation-based management systems by both conservation and socioeconomic standards. At their core, these superior systems rely on off-limits populations to buffer against uncertainty. Scientists and managers rejected such policies in the past because they can lead to large fluctuations in catch from year to year. I will illustrate how combining off-limits populations with caps on total catches can lead to a wide spectrum of socioeconomic properties while also addressing ecosystem concerns. Using such systems, managers must still face a socioeconomic trade-off between high but variable yields and lower but stable yields. Nevertheless, managers can achieve this full spectrum of socioeconomic options without sacrificing an ecosystem-based approach to fisheries management.

Scientists have given advice on how to manage marine resources, fishery resources in particular but also non-fishery resources that impinge on the former, since the middle of the 19th century. The world view suggested by the nature and context of the advice given has changed in some ways since then but has remained remarkably consistent in other ways. Those things that have not changed reflect the more fundamental aspects of the public world view in which scientists and fisheries managers, whatever their private world views, play out their roles. I will therefore explore whether there are real differences between how management advice on trade-offs in marine-resource management were given and received first when the context was "fishery biologists" applying "single-species approaches" and second in the present day context of "conservation biologists" applying "ecosystem approaches."

Two cephalopod species, northern shortfin squid and longfin inshore squid, play a pivotal role in the complex and valuable marine ecosystem of the U.S. mid-Atlantic fishery. Squid and their predators-summer flounder, silver hake, goosefish and bluefish-comprise more than half of the total commercial landings in the U.S. mid-Atlantic continental shelf ecosystem. Currently, all of these species are considered overfished and are managed under separate management plans, so trade-offs between these fisheries are not considered. This project will explore the structure and function of the food web to identify these trade-offs. This goal will be achieved through an extensive diet analysis and the development of bioenergetics models that estimate consumption rates of squid and their predators. The information gained from these two initiatives will be incorporated into a coupled bioenergetics and age-structured population model designed to evaluate the interactions between squid and fish. Ecopath with Ecosim, a multi-species ecosystem-based model, will then be used to simulate the dynamics of the entire food web under alternative harvest policies for analysis of the economic trade-offs and ecological impacts imposed through fishing. The results from this study will advance our knowledge of food-web dynamics and can be used to implement improved policies for fisheries management in this region.

Most of the world’s marine ecosystems have been deeply altered by direct and unintentional effects of fishing and by coastal development. There appears to be emerging consensus among scientists that the basic goal of marine ecosystem management today should be to restore ecosystems to a more natural "healthy" state, by reducing fishing effort in general, moving to selective fishing practices that reduce bycatch and habitat damage, and protecting and restoring habitats through zoning and protected area policies. The rhetoric from those of us who have particular values related to biodiversity has done little to discourage the public from believing that only more natural ecosystems will be sustainable. That rhetoric is deeply misleading: there is no sound scientific rationale or evidence for equating "natural" with "sustainable," and it is very probably possible to develop sustainable-use regimes for ecosystems that are very different from what was there naturally. For example, shrimp fisheries have thrived and show every sign of being sustainable in many areas where finfish have been depleted, and they are often much more valuable from an economic perspective. The Bering Sea ecosystem will quite probably remain very productive for fisheries even if marine mammals like the Steller sea lion disappear entirely. Scientists need to be much more honest about such things, and much more objective about informing the public particularly about the economic costs and trade-offs associated with the restoration of values related to biodiversity. Those of us who promote the restoration of natural ecosystem structure are not the ones who would bear the costs of doing it.

In the eastern tropical Pacific (ETP), five tuna fisheries land and discard different species assemblages. These fisheries are roughly separated by flag, so management actions that reallocate total fishing mortality among fisheries can excite competition among national interests. We assert that all national interests are justifiable, and there is a need to identify ecosystem management objectives that are flexible, allowing specific national objectives to be achieved through the political process. We use an ecosystem model to describe how arbitrary management objectives based on the "existence value" of various species groups can influence the optimal distribution of total fishing mortality among the five fisheries in the ETP. Such objectives can define "indifference regions" in which multiple fishing-mortality allocations are optimal. Large indifference regions may, in fact, be desirable. Large regions might allow political negotiating to continue while an ecosystem perspective is maintained. Management objectives based on indices of ecosystem health may not be flexible (having a small or no indifference region) and may therefore create difficulties for negotiation. We map the indifference regions of a few arbitrary objective functions developed from management issues pertinent to tuna fisheries in the ETP.