Rigorously protect structural habitat

The close association of reef fishes with habitat through ontogeny is now well established (Appeldoorn et al. Reference Appeldoorn, Recksiek, Hill, Pagan, Dennis, Lessios and Macintyre1997; Reference Appeldoorn, Friedlander, Sladek Nowlis, Ussegilo and Mitchell-Chui2003; Cocheret de la Morinière et al. Reference Cocheret de la Morinière, Pollux, Nagelkerken and van der Velde2002; Nagelkerken et al. Reference Nagelkerken, Roberts, Van Der Velde, Dorenbosch, van Riel, Cocheret de la Morinière and Nienhuis2002; Christensen et al. Reference Christensen, Jeffrey, Caldow, Monaco, Kendall and Appeldoorn2003; Nagelkerken & van der Velde Reference Nagelkerken and van der Velde2004; Mateo & Tobias Reference Mateo and Tobias2004, Reference Mateo and Tobias2007; Aguilar & Appeldoorn Reference Aguilar-Perera and Appeldoorn2007; Monaco et al. Reference Monaco, Friedlander, Caldow, Christensen, Rogers, Beets, Miller and Boulon2007; Pittman et al. Reference Pittman, Christensen, Caldow, Menza and Monaco2007). A critical concept embedded within this realization is that the unit of management is not the species or community, but the habitat, and, as in terrestrial systems, that habitat can be used as a surrogate for demography (Hansen et al. Reference Hansen, Urban, Marks, Hansen and di Castri1992; Merriam & Wegner Reference Merriam, Wegner, Hansen and di Castri1992). Recognition of the importance of maintaining healthy habitats is explicit within current US management under the concern for EFH and the designation of Habitat Areas of Particular Concern (HAPCs). In the US Caribbean, the CFMC has broadly defined EFH for reef species and has already designated several HAPCs (CFMC 2004).

Protect water quality

The aqueous environment of coral reefs represents a specific component of habitat as well as an important mechanism for transporting materials and nutrients. It is the component of habitat most immediately impacted by terrestrial activities such as coastal development, poor land-use practices, offshore sewage outfalls and pollution events. Primary productivity of coral reefs is based on benthic production of algae, sea grasses and coral-symbiotic zooxanthellae, and these can be severely impacted when turbidity reduces light penetration (Kinzie et al. Reference Kinzie, Jokiel and York1984). In US Caribbean coastal waters, suspended sediments and eutrophication are the major factors responsible for the higher turbidities (Acevedo & Morelock Reference Acevedo, Morelock, Choat, Barnes, Borowitzka, Coll, Davies, Flood, Hatcher, Hopley, Hutchings, Kinsey, Orme, Pichon, Sale, Sammarco, Wallace, Wilkinson, Wolanski and Bellwood1988; Nemeth & Nowlis Reference Nemeth and Nowlis2001; Burke & Maidens Reference Burke and Maidens2004; Garcia-Sais et al. Reference García-Sais, Appeldoorn, Bruckner, Caldow, Christensen, Lilyestrom, Monaco, Sabater, Williams, Diaz and Jeffrey2005; Warne et al. Reference Warne, Webb and Larsen2005). In addition to loss of primary productivity, sedimentation and eutrophication can affect the structural composition of coral reefs (Loya Reference Loya1976; Kinzie et al. Reference Kinzie, Jokiel and York1984; Rogers Reference Rogers1990; Littler et al. Reference Littler, Littler and Brooks2006a, Reference Littler, Littler, Lapointe and Barileb) and associated fish communities (Bejarano Rodriguez Reference Bejarano Rodríguez2006). With maintaining water quality an important component of ecosystem management (Souter & Linden Reference Souter and Linden2000; Jameson et al. Reference Jameson, Tupper and Ridley2002), fisheries management will have to expand in significant ways to cover land-use practices affecting reef environments. As a consequence, management strategies will also have to expand to include indicators of water quality and ecological health (for example see Jameson et al. Reference Jameson, Erdmann, Karr and Potts2001; Bejarano Rodriguez Reference Bejarano Rodríguez2006).

Maintain ecosystem integrity

In terrestrial conservation biology, an important management goal is to maintain ecological integrity (Leopold Reference Leopold1966; Merriam & Wegner Reference Merriam, Wegner, Hansen and di Castri1992). Ecosystems are a composite of not only structural habitats, but also the species found within them, each interacting in innumerable ways. It is impossible to model and assess all species within an ecosystem. It is equally impossible to predict the long-term consequences of loss of biodiversity that may result from exploitation or management intervention. Risk aversion dictates that one goal should be to maintain all species within a system, subject to natural levels of variation, such as those resulting from island biogeographic (MacArthur & Wilson Reference MacArthur and Wilson1967) or successional (for example Duran & Castilla Reference Duran and Castilla1989) processes.

Maintain ecosystem function

A healthy self-sustaining ecosystem requires not only that the essential structural components (species, habitats) are intact, but also that the functional components are intact as well (Hughes et al. Reference Hughes, Bellwood, Folke, Steneck and Wilson2005). Key functional components would be primary production, herbivory, predation, trophic pathways, nurseries, migration, shelter and reproduction. Many of these are habitat based, such as the use of distinct nursery habitats by newly settled and small juvenile fishes (Eggleston Reference Eggleston1995; Appeldoorn et al. Reference Appeldoorn, Recksiek, Hill, Pagan, Dennis, Lessios and Macintyre1997; Beck et al. Reference Beck, Heck Jr, Able, Childers, Eggleston, Gillanders, Halpern, Hays, Hoshino, Minello, Orth, Sheridan and Weinstein2001; Lindeman et al. Reference Lindeman, Pugliese, Waugh and Ault2000; Nagelkerken et al. Reference Nagelkerken, Roberts, Van Der Velde, Dorenbosch, van Riel, Cocheret de la Morinière and Nienhuis2002) or site specific spawning aggregations for snappers and groupers (Shapiro et al. Reference Shapiro, Sadovy and McGehee1993; Sadovy et al. Reference Sadovy, Colin and Domeier1994; Lindeman et al. Reference Lindeman, Pugliese, Waugh and Ault2000; Claro & Lindeman Reference Claro and Lindeman2003; Ojeda-Serrano et al. Reference Ojeda-Serrano, Appeldoorn and Ruíz-Valentín2007). However, equally important is the flow of nutrients and species across habitats within the seascape. These include the movement of species across habitat boundaries during ontogeny, feeding or reproduction (Cocheret de la Morinière et al. Reference Cocheret de la Morinière, Pollux, Nagelkerken and van der Velde2002; Appeldoorn et al. Reference Appeldoorn, Friedlander, Sladek Nowlis, Ussegilo and Mitchell-Chui2003; Kendall et al. Reference Kendall, Christensen and Hillis-Starr2003; Nagelkerken & van der Velde Reference Nagelkerken and van der Velde2004) and associated transport of organic matter and nutrients (Meyer et al. Reference Meyer, Schultz and Helfman1983; Clark et al. Reference Clark, Monaco, Appeldoorn and Roque2005).

Maintain a series of control areas for monitoring

The ecosystem approach advocated here does not rely on exact prediction of the state of the ecosystem. History and simulation studies have demonstrated the inability to precisely and accurately determine both the current state of the fishery and its theoretical point of optimal production in the face of incomplete data, unknowns and annual variability of natural systems (Sissenwine et al. Reference Sissenwine, Brown, Palmer, Essig and Smith1982; Walters Reference Walters, Pitcher, Hart and Pauly1998). However, the approach does require monitoring, a process that is both much more tractable and useful for adjusting management tactics as needed. Monitoring can be used to determine directions and rates of change. Comparisons between fished and unfished areas would be critical in relating any observed changes to the impacts of fishing (Bohnsack Reference Bohnsack1998). Such an approach is not new; in the 1880s the Scottish Trawling Experiment compared open and closed areas to demonstrate the negative impact of trawling on plaice (Smith Reference Smith1994), while more recently, Polovina (Reference Polovina1994) compared fished and relatively unfished areas to suggest that the decline in the Hawaiian spiny lobster resources was due to climate, rather than fishing. New observations in unfished areas in the North-west Hawaiian and Northern Line Islands (Friedlander & DeMartini Reference Friedlander and DeMartini2002; Stevenson et al. Reference Stevenson, Katz, Micheli, Block, Heiman, Perle, Weng, Dunbar and Witting2007; Sandin et al. Reference Sandin, Smith, DeMartini, Dinsdale, Donner, Friedlander, Konotchick, Malay, Maragos, Obura, Pantos, Paulay, Richie, Rohwer, Schroeder, Walsh, Jackson, Knowlton and Sala2008) are redefining perceptions of the structure of unexploited reef ecosystems. These types of comparisons also help ground truth the validity of theoretically constructed reference points. Reference points are a fundamental component of current single-species fisheries management, and single-species reference points will continue to be valuable (Quinn & Collie Reference Quinn II and Collie2005). In the USA, key reference points are derived from an estimation of MSY and spawning potential ratio (SPR). Both of these are theoretical constructs and require detailed life-history data and/or long time-series of accurate data on catch and effort. The latter are generally not available for fisheries within the US Caribbean. Life-history and length-frequency data can only hope to be collected for the most important and abundant species. In addition, new reference points will need to be developed at the ecosystem level. Some of these may be specifically related to ecosystem health, such as indexes of biotic integrity (Jameson et al. Reference Jameson, Erdmann, Karr and Potts2001). Others (for example FAO [Food and Agriculture Organization of the United Nations] 1999; Busch et al. Reference Busch, Brown and Mayer2003) may be developed from theoretical and practical considerations.

Employ a precautionary approach at all times

The present degree of uncertainty in Caribbean fisheries is high due to the limited data relative to the large numbers of species. The precautionary approach to fisheries, embodied in the Code of Conduct for Responsible Fisheries (FAO 1999), was specifically designed to deal with the problems of uncertainty. It recognizes that all forms of fishing have negative impacts, and that management should be forward looking but must also operate with incomplete data. This requires a standard of proof to be used in decisions regarding authorization of fishing activities that is commensurate with the potential risk to the resource. Additionally, it states ‘that where the likely impact of resource use is uncertain, priority should be given to conserving the productive capacity of the resource’ (FAO 1999). In this view, adopting an ecosystem-based approach to fisheries management is itself a fundamental component of the precautionary approach.

Recognize there are limits to production; therefore control rates of extraction

All species have limits to their rate of production. As a first rule of limiting harvest to maintain production, juveniles should be allowed to mature and adults should be allowed to spawn. Typically, species productivity is related to M/K, the ratio of the instantaneous coefficients of natural mortality and growth (Beverton & Holt Reference Beverton, Holt, Wolstenholme and O'Connor1959; Pauly Reference Pauly1980), but more generally these can be related to longevity, and as a surrogate, maximum fish body size. Thus, productivity rates will vary among species and larger longer-lived species will have lower rates of production. This makes it impossible to maximize production across species in a multispecies fishery; either larger species will be overfished or smaller species will be underfished. Given the important role of larger predatory species in top-down control of ecological processes (Jackson et al. Reference Jackson, Kirby, Berger, Bjorndal, Botsford, Bourque, Bradbury, Cooke, Estes, Hughes, Kidwell, Lange, Lenihan, Pandolfi, Peterson, Steneck, Tegner and Warner2001), ecosystem considerations would place special emphasis on limiting fishing pressure on these species.

Sustainable Fisheries Act

The language and structure of the USA's Sustainable Fisheries Act (SFA) is not designed with the US Caribbean in mind. It serves to address the more typical industrial-scale fisheries that are characterized by a few high-value high-volume stocks where information can be channelled through a few centralized landing sites and which have a relatively large management and research infrastructure. Nevertheless, the SFA does have a degree of flexibility; most importantly, it does not strictly define what constitutes a ‘fishery’. It clearly does not link a fishery with either a species or species stock, which allows for using multispecies units or subunits when complying with the mandates of the SFA relative to each fishery. The SFA is rigorous in the demands placed on resource scientists and management. For every fishery, levels must be set for MSY and allowable catch, and the point when a fishery is considered overfished must be defined in specific terms. This approach, known as weak-stock management, tries to manage all fisheries such that none are overexploited. Where each fishery is a single species or species stock, predator-prey, competitive and technological (by-catch) interactions could lead to a situation where ‘almost all of the potential yield from the fishery will be forgone if the objective is to prevent overfishing for all stocks’ (Hilborn et al. Reference Hilborn, Punit and Orensanz2004). Viewing a fishery as multispecies could relax this constraint and allow a greater maximization of societal benefits. Using a multispecies approach, the SFA does not preclude using ecosystem metrics in the definition of overfishing, and MSY and allowable catch could be determined on a two-tiered level, one for the multispecies complex (satisfying the mandates of the SFA) and another for targeted sub-component species or species groups to control key ecological or economic components (reminiscent of, but not exactly like what was used in the fisheries off the north-east USA; Brown et al. Reference Brown, Brennen, Gosslein, Heyerdahl and Hennemuth1976). In some sense, the recent amendments of the Shallow Water Reef Fish Management Plan of the CFMC already represent a move in this direction, where multispecies subunits are defined (for example Grouper Unit 4 = misty, red, tiger, yellowedge and yellowfin groupers) and overfishing status is assessed for the group as a whole.

The importance of healthy habitats within a coral reef ecosystem framework cannot be overemphasized, and the SFA specifically mandates that important habitats be identified and cared for. However, the RFMCs are not given any authority to see that this is done. The main authority lies directly within the National Oceanic and Atmospheric Administration (NOAA) and its role is largely limited to issues that fall within NOAA's jurisdiction. The agency is at least required to communicate its concerns to other federal and local agencies, but it is limited in the capacity to which these concerns are addressed. Control of anthropogenic stresses in nearshore reef systems is critical, and authorities here lie with the local governments and various federal agencies (for example the US Army Corps of Engineers, the Department of Agriculture and the Environmental Protection Agency), regulated largely by the US National Environmental Policy Act (NEPA).

Lastly, the use of control areas for ecosystem assessment seems incompatible with the immediate mandates of the SFA. Closed areas will need considerable time, of the order of decades, before they might approach a condition that could be considered baseline. However, most marine reserves in reef ecosystems start to respond rapidly after implementation (Polunin & Roberts Reference Polunin and Roberts1993; Roberts Reference Roberts1995; Russ & Alcala Reference Russ and Alcala1996), thus allowing them to start being used to assess fishery impacts.

Federal versus local jurisdiction

All of the nearshore nursery habitats fall within local jurisdiction and much (almost all in the case of Puerto Rico) of the insular platform is also under local control. Any attempt at ecosystem-based management, especially using a network of no-take areas will require cooperation on a scale that is not evident at present. Both levels of government would have to share the same vision and have regulations that are compatible both with that vision and between governments. Data collection and analysis must also be shared, and enforcement must be modified and strengthened. Key to this endeavour is for the local governments to feel that they are full partners in this effort, so they take ownership in the process and responsibility for the outcome. It is unlikely that a top-down approach from NOAA or the CFMC would succeed.

The cultures of science and management

Implementation of an ecosystem-based approach requires changes in the way scientists and managers view their roles, the systems that they work with and the way in which they do business. Ecosystem-based management as outlined here constitutes a distinct paradigm shift. Managers will have to alter how they interact with fisherfolk and other stakeholders, with mechanisms devised to allow more direct involvement of these groups. Processes being developed for community participation in the design and implementation of marine protected areas (MPAs) may serve as a role model (see Delany Reference Delaney2003; Howard et al. Reference Howard, Connolley, Taylor and Mou2003). Scientists will have to break out of the narrow confines of current fisheries routines, think more creatively about the ways in which the status of a fishery can be assessed and incorporate concepts of habitat condition. This may explain why fish ecologists have been at the forefront of the push toward ecosystem-based management (see Jennings Reference Jennings2004; Kaufman et al. Reference Kaufman, Heneman, Barnes and Fujita2004; Hughes et al. Reference Hughes, Bellwood, Folke, Steneck and Wilson2005; Mangel & Levin Reference Mangel and Levin2005) while stock assessment scientists struggle with the question as to how to implement the process (see Busch et al. Reference Busch, Brown and Mayer2003; Babcock & Pikitch Reference Babcock and Pikitch2004).

While paradigm shifts inherently take time, there are indications that such transitions are already occurring (Berkes et al. Reference Berkes, , McConney, Pollnac and Pomeroy2001). A recent exercise in the single-species assessment of queen conch, a large gastropod, in the US Caribbean serves to show how open traditional assessment scientists are to new ideas. After reviewing existing conventional catch-effort and size-structure data, the review panel concluded ‘the available data on queen conch abundance and fishing pressure are inadequate or insufficiently informative to allow a stock assessment or associated projections and benchmark statistics to be presented’ (SEDAR [South-east Data Assessment and Review] 2007a, b). Yet when shown the relative density of conch in Puerto Rico (from survey data) compared to unfished and heavily fished areas around the Caribbean (Fig. 1), which clearly showed the Puerto Rico stock to be below a crude but theoretically defensible MSY, they also supported using such a collation of densities across fishing gradients and concluded ‘preliminary results of such an exercise . . . show considerable promise as an innovative approach to placing survey findings in the context of potential population benchmarks. The [Review Panel] considered that progress with this approach is a high, short-term priority, and recommended that further attention be paid to the influence of habitat type and stock structure (juveniles and adults) on the comparisons.’ (SEDAR 2007a, their emphasis).

Figure 1

Queen conch (Strombus gigas) densities for shelf wide surveys throughout the US Caribbean. Densities are unadjusted for any potential differences, such as habitat distributions or proportion of juveniles. Striped bars are values for western Puerto Rico. Solid line represents the density below which reduced reproductive output may be expected based on studies in the Bahamas. Dashed line represents the level at which density is half the maximum density observed in the deep stratum on Pedro Bank (PB95-30), assumed to reflect virgin stock density there. Arrow shows the general direction of density increase off western Puerto Rico over the past 10 years (from SEDAR 2007b).