Ecosystem-based management of fisheries: is science limiting?

With the adoption of the Convention on Biological Diversity (CBD; UN, 1992), managing the environment in an ecologically sustainable manner has shifted from being an option to being a legal necessity: sustainability is now the overarching goal of environmental management policy. Ecosystems have natural resilience. Reproduction and adaptability are fundamental biological attributes which respond to human activities. This means that the real challenge for managing the system is determining the key limits, i.e. the types and levels of human activities that can be sustained without compromising the functioning of the ecosystem, and setting in place policies to obtain those goals. The latter is a socio-political issue, the former is very much a scientific issue. The ecosystem approach promotes conservation and equitable, sustainable management of land, water, and living resources. It relies on scientific understanding of ecosystem structure, processes, functions, and interactions (CBD, 2000). It recognizes that humans, with their cultural diversity, are an integral component of many ecosystems. It also emphasizes the fact that, by their very nature, ecosystems cut across traditional management sectors. Therefore, we are faced with developing, for example, fisheries management in an ecosystem context, while not necessarily being able to influence fertilizer applications in the hinterland that may impact fish nursery areas. In most jurisdictions the requirements of the CBD are being developed within existing management structures. The reality is therefore that each sector, within each jurisdiction, is trying to develop ecosystem-based management regimes. Here we consider how science is placed to contribute to this wider fisheries management agenda. Fishing activities influence marine ecosystems in a number of ways. These include (i) direct removal of target species, (ii) direct changes in size structure of target populations (Jennings and Kaiser, 1998), (iii) alteration in non-target populations and communities of fish and benthos (Rumohr and Krost, 1991; Camphuysen et al., 1995; Tuck et al., 1998), (iv) alterations in the physical environment (Churchill, 1989; Messieh, 1991; Riemann and Hoffmann, 1991; Auster et al., 1995; Schwinghamer et al., 1996), (v) alterations in the chemical environment (ICES, 1998; Percival et al., 2005), and (vi) food-chain effects such as trophic cascades (Carpenter et al., 1985) and altered predation pressure (Frid et al., 1999). It is impossible to harvest a species without causing (i), and because harvesting is selective, causing (ii). The challenge for ecosystem-based fisheries management is then to provide the maximum sustainable take of target organisms, with the minimum impact on other ecosystem components. A number of measures can be regarded as addressing ecosystem considerations in the management of the fisheries of the Northeast Atlantic. The ecosystem issues addressed in the present management regime are interactions between species (i.e. through the use of multispecies stock assessment procedures); fisheries effects on the supply of food for predators (e.g. cod, Gadus morhua, and capelin, Mallotus villosus; ICES, 2001, 2003a) and sandeels (Ammodytes spp.) for seabirds (ICES, 1999); and the additional fishing mortality of non-target species (e.g. pingers to prevent cetaceans entering set nets (EC, 2004b). There has been less progress in providing protection of habitat areas and the ecological functions they deliver, or in conserving genetic diversity. It is also recognized that the influence of the ecosystem on fisheries is not well addressed by current management procedures; for example, there is no mechanism for accounting the variations in climatic factors and hydrography (ICES, 2005). Ecosystem dynamics are controlled in part by external drivers such as hydrodynamic/atmospheric factors. These drivers influence the productivity of phytoplankton (Dickson et al., 1988; Richardson et al., 1998), and the dynamics of zooplankton communities (Marshal, 1949; Krause and Trahms, 1983). The survivorship of fish larvae and hence recruitment into the adult stock are related to plankton productivity, while climate variation affects the natural mortality, distribution, and migration of fish populations (Jennings et al., 2001). In future, fisheries management will need to incorporate these types of external drivers and their impact on ecosystem dynamics to improve the reliability of predictions, especially in the face of increasing climate variability linked to global warming. However, the level of uncertainty in our existing knowledge of the effects of external drivers on fish stocks will need to be addressed if the information on the state of external drivers available from “operational oceanography” is to be used for management purposes. While all habitats are likely to provide ecological services and the physical environment for certain species or life stages to exist, not all habitats are equally sensitive to the impacts of fishing. The Oslo–Paris Commission (OSPAR) provides the international regulatory framework for a number of human impacts in the Northeast Atlantic, although it does not have competence over fisheries management. OSPAR is, however, the lead body for the application of the UN Convention on Biological Diversity (UN, 2002) in the regional marine environment, and as such is developing management measures to ensure that CBD goals are met. OSPAR is in the process of identifying habitats that are sensitive to various human impacts, including fishing. This categorization obviously requires consideration of the different types of fishing activity and then a consideration of appropriate management measures (Table 1; ICES, 2002b, 2003b). Matrix of potential mitigating measure for various fishing métiers in OSPAR sensitive habitats (after ICES, 2002b). Shading indicates fishing activities thought to have no effect. Key to sensitive habitat types (taken from the draft OSPAR list): Deepwater biogenic habitats: Lophelia pertusa reefs, carbonate mounds, oceanic ridges with hydrothermal vents, seamounts, and deepwater sponge communities. Structural benthic epifauna: Sabellaria spinulosa reefs. Benthic infauna: Seapens and burrowing megafauna communities. Shellfish beds: Ampharete falcata sublittoral community, Ostrea edulis beds, Modiolus modiolus beds, and intertidal mussel beds. Nearshore communities: Zostera beds and littoral chalk communities. Key to mitigation measures: AC, area closures; R, reseeding/restocking; GM, gear modification. In one sense the management measure to protect physical habitat is straightforward: one simply needs to remove the damaging activity from the sensitive area. Therefore, the management measures are inherently spatial. Such closures to all, or certain damaging types of, fishing gears have now been implemented for some deepwater coral communities (EC, 2004a). The more widespread application of measures based on protected areas is limited by lack of agreement on the total fraction of a habitat that should be protected. For rare habitats one can envisage complete protection, but IUCN guidelines suggest that at least 20–30% of each habitat type is strictly protected (IUCN, 2003). The lack of highly spatially resolved maps also hinders this process, although some European countries have started to map their national seabed areas using a standard habitat classification scheme (http://eunis.eea.eu.int/habitats.jsp), so it seems likely that in the near future the first obstacle, namely knowing what we have and how it is distributed, will be overcome. Assuming that we have also agreed on a target proportion of habitat to protect, we must decide which part(s) of the whole should be protected. For example, should we distribute a prescribed protection of 20% of the habitat in one block, two blocks of 10% each, 20 of 1%, 40 of 0.5%, or some combination of areas of varying size? Nature reserve theory is much more advanced for terrestrial systems, and generalities have emerged (Diamond, 1975; Pressey, 1994; Etienne and Heesterbeek, 2000; Kingsland, 2002). Long-term conservation is best served by a large number of areas, providing that each is of a sufficient size to hold viable populations. In terrestrial systems the aim has been generally to provide protection for one or a few species, whereas in the marine environment we usually seek to provide protection to a functioning, multispecies assemblage. Nevertheless, many of the lessons are likely to transfer, but with additional provisos, occasioned by the open nature of most marine populations, that decisions need to be made to determine the optimum spatial distribution of closed areas on an effective biological scale that considers gene flow, fish movement, and the transport of planktonic larvae. The aim will be to have an ecologically linked network of protected areas. The spatial configuration of a network of protected areas will need to consider biological and physicochemical factors, such as patterns of larval dispersal, identification of where and when adults spawn, the biology and behaviour of the larvae, and the hydrography of an area (Stobutzki, 2001), in order to maintain the functional integrity of the ecosystem in those areas. However, there are no absolute figures as to how close reserves should be (Roberts et al., 2003), and hard evidence to support the design of reserve networks is lacking owing to the complexity of the task and a scarcity of examples. A model developed to investigate the design of marine reserve networks in the Gulf of California determined that the space between adjacent reserves should not exceed 100 km, because of uncertainty about dispersal patterns, and calculated that reserves needed to be sufficiently large (50 km2) to ensure the local retention of >90% of algal propagules and >45% of fish and invertebrate larvae (Sala et al., 2002). However, some argue that the reserves should not be placed too close to one another to reduce the risk that a local catastrophe will affect more than one of them (Roberts et al., 2003). Hydrographic information would also need to be incorporated into the design of networks, and it has been suggested that in areas where currents are strongly directional, the reserves sited in upstream locations will be more likely to supply recruits to the rest of the management area than those in downstream locations (Roberts et al., 2003), so would presumably be of greater value. However, where currents are complex or reversing, a more even spread of reserve locations would be preferable. Although we may now know the sensitive habitats and where some are, issues concerning the number and spatial distribution of protected areas remain a major challenge. Recent advances in understanding of the fine-scale genetic make-up of benthic populations (Todd, 1998) and in the dynamics of meta-populations (Grimm et al., 2003) do provide some pointers that may allow the development of appropriate modelling frameworks. Genetic diversity is, potentially, the product of very long periods of evolution, yet losses may be irreversible and virtually immediate (Nielsen and Kenchington, 2001; Kenchington, 2003). This diversity is important for the ability of a species to adapt to extrinsic factors such as pollution or climate change. Loss of populations (extirpation) most likely equates to a loss of adaptive variation in a species, and a loss of adaptive radiation may ultimately result in the loss of that species. Management units are rarely based on population genetic structure. The North Atlantic blue whiting (Micromesistius poutassou), for example, is managed as a single stock. However, population genetic studies have indicated that partially separated stocks of blue whiting exist in the Mediterranean and in the Northeast Atlantic (Giaever and Stien, 1998). Similarly, Ruzzante et al. (2001) reported on the decadal stability of the genetic differentiation of five cod populations on spawning banks off Newfoundland and Labrador, Canada. This genetic structure persisted through the recent population collapse, with only some suggestion of post-collapse mixing between populations from two of the spawning banks. This information is critical to recovery management, because it indicates that population re-growth will be the mechanism for rebuilding the stocks, as opposed to immigration from other areas. Commercially, genetic diversity is also very important for aquaculture, providing the raw material for selective breeding programmes and the revitalization of inbred broodstock. ICES suggests four general measures to mitigate against the loss of genetic diversity (ICES, 2002a): (i) fishing mortality should be kept sufficiently low to maintain large populations; (ii) the harvest should be widely geographically distributed among all recruited populations, to avoid local depletions and fragmentation; (iii) there should be a reduction of fishing effort rather than alternative management approaches that result in fisheries becoming even more selective; and (iv) there should be a case-by-case evaluation of the risks associated with the loss of genetic diversity vs. the benefits of imposed action. These were suggested as “common sense” approaches for managers to follow until the scientific community could recommend a more rigorous framework for specific stocks. Points (i), (ii), and (iii) seek to reduce the selective pressure exerted by fishing and the potential creation of a genetic bottleneck in either the species (i) or any subpopulation (ii). Point (iii) addresses concerns that more selective fishing technology may impose a greater selection pressure than the rather crude size selection imposed by simple nets. Point (iv) aims to provide a drive to find specific solutions matched to the circumstances of individual stocks. While these measures have formed part of formal ICES advice to the European Community, no fisheries regulations have been introduced yet specifically targeting conservation or protection of genetic diversity. Two types of barriers inhibit the implementation of ecosystem-based advice: (i) a deficiency in our scientific understanding of the ecosystem, including the human part of it, and (ii) methodological impediments to using the science as a basis for management. The preceding sections suggest that we currently lack critical understanding of links between hydrographic regimes and fish stock dynamics; the importance of habitat distributions; “design rules” for MPAs, how they should vary by regions/systems, and how to ensure that they are ecologically linked; ecological dependence/foodweb dynamics; predictive capabilities in complex systems; how to incorporate uncertainty into management advice; the genetics of target and non-target organisms; and the response of fishers to management measures. It is beyond the scope of this paper (and the expertise of the authors) to provide detailed descriptions of all the science programmes needed to address these knowledge gaps. It is, however, clear that a number of these gaps are at the boundaries of traditional discipline areas or are in need of multidisciplinary approaches. For example, genomic and molecular markers may throw light on the genetics of marine organisms and will inform our understanding of sub- and meta-population dynamics. They will contribute to our understanding of the scale of dispersal and linkages between spatially distinct populations, the latter being important for consideration of the potential effectiveness of MPAs. Similarly, understanding links between physical drivers, such as hydrography and climate, and biological responses will be important for improving predictions of future stock status. Additionally, how the ecosystem, as a complex set of feedbacks and linkages, buffers/mitigates these responses adds a further complexity to the predictions, and learning to manage uncertainty is yet another key challenge. It is clear that as we enter the age of operational oceanography, one of the key users of the information will be those seeking to provide ecosystem-based management advice for fisheries. However, humans are also part of the fishery ecosystem, and we need to give as much consideration to predicting the response of our own species to the fisheries management regime implemented (Pascoe, 2006). We cannot be regarded as some sort of simple, additional, predator. The response of fishers to fisheries regulations varies depending on many factors, including their degree of perceived ownership of the management regime, their perception of the behaviour of other players (fleets and individuals), and the economic and social circumstances. It is clear that many gaps in our knowledge can be identified, but it is also apparent that there are a number of issues that may restrict the use of the knowledge we have. In the US, Canada, Australia, the EU, and probably many other fisheries management regimes, management decisions are made, quite rightly, in a political arena. In a democracy this is the logical place for the balancing of the various, potentially conflicting, issues. However, in these cases it is also true that science travels through a number of filters before it is considered by the decision-makers. The internal reports, refereeing, distillation into briefing papers/advice, and the interpretation of this by civil servants all dilute the power of the science and, in our opinion, remove the ability to seek alternatives; indeed the debate is polarized. There is a need to ensure that science is presented directly to decision-makers in a manner that allows them to engage in dialogue. This allows misunderstandings to be avoided and allows for the scientific evaluation/consideration of alternatives that might emerge during the political debate. The reasons for the development of our current administrative frameworks and the development of new ones that see the removal of these barriers are issues that need to be addressed by our social and political science colleagues. In many cases, however, these studies need to proceed as joint natural and social science initiatives, because only in this manner can we expect the natural sciences to understand the nature and context for the advice they are being asked to produce. Effort reduction is widely recognized as being the single most effective measure that could be implemented to benefit it will not on own provide protection to all ecosystem attributes in need of areas, complete or certain métiers are the only protection for habitat and even then we must understand the scale and spatial distribution to provide whereas effort will reduce the selective pressure on the gene it will not remove it, and although approaches such as gear have a clear to in habitat and in mortality, these will not remove the genetic effects of fisheries. The ecosystem approach is inherently and must be for and to of the has a in so that the management set are and at consideration of the The science needs of an ecosystem approach are much than those of fisheries This will some of if the advice is to from the best and need to that science is inherently and, as there will be no single necessarily a the best is clear is that the of science needed to to this process is wider than that for present fisheries management. The and presented are the own and do not the of ICES or any other However, we the that within the ICES science community has In we the of the on the Ecosystem of Fishing and has to the development of an ecosystem approach to fisheries management in In we Kenchington, and for their on this and related This in part by the through Ecosystem The of the presented the of the and two we their and them

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.