The establishment of MPAs around the world has been extensive and is rapidly expanding, but the effectiveness of many MPAs is regularly questioned (Mora and Sale, 2011; Rife et al., 2013; Bergseth et al., in press). Much of the criticism of MPAs relates to allowed activities, poor enforcement, and alienation of stakeholders. However, there is a fundamental question that has seldom been asked (but see Spalding et al., 2013): How effectively are MPAs separating marine biodiversity from processes that threaten its persistence? This question relates to the basic purpose of any protected area. Too often, the establishment of protected areas is seen as equivalent to effective protection, and very often this conflation of ideas is mistaken. Protected areas fail in their basic purpose to the extent that they are residual to extractive uses. A strong focus on minimizing the opportunity costs of MPAs, combined with limited biological data and highly generalized conservation objectives, entails the considerable risk of pushing ‘protection’ into residual parts of the ocean (Figure 1).
Another look at the four questions related to minimizing opportunity costs of MPAs
At the outset four questions were posed (Figure 1) that decision makers should be able to answer about the implications of minimizing opportunity costs when designing MPAs and no-take zones. The evidence that minimizing the opportunity costs of MPAs and no-take zones has had perverse outcomes for marine biodiversity globally, in Australian waters, and in the Great Barrier Reef Marine Park is now discussed. This assessment involves a slight rephrasing of questions 2–4 by replacing ‘Should’ – for statements of intent in Figure 1 – with ‘Do’ – for assessment of outcomes here. Table 2 summarizes the assessments of this study, beginning with three alternative answers. ‘Yes’ indicates that the evidence for a positive answer outweighs negative evidence. ‘No’ indicates the opposite. ‘Uncertain’ indicates mixed or scarce evidence. The lens for these assessments is shaped by current scientific thinking about the location and configuration of MPAs. In the case of the Great Barrier Reef Marine Park, rezoned in 2004, using this lens therefore draws to some extent on hindsight. Hopefully, this assessment provides some lessons for the next rezoning of that region, whenever that occurs, and also for MPA planning in other regions that would seek to emulate what was, in 2004, world's best practice.
- Question 1:
Are MPAs/no-take zones intended to protect biodiversity?
Table 2. Analysis of patterns of marine protected areas globally, in the Australian marine jurisdiction, and in the Great Barrier Reef region, in relation to the four questions posed in Figure 1
At the global level, the intent to protect marine biodiversity using MPAs and no-take zones is stated explicitly in policy (e.g. WSSD, 2002 4; IUCN World Parks Congress, 2003 5; CBD COP, 2010 6) and in IUCN's definition of MPAs as primarily focused on conservation outcomes (Fitzsimons, 2011; Day et al., 2012). Nationally, Australia's in-principle commitment to the conservation of marine biodiversity through MPAs is clearly stated. Principles for expanding the Australian MPA system to represent and promote the persistence of biodiversity have long been established (ANZECC, 1998). Goals, apparently qualitative, underpinning the recent bioregional planning exercises in Australian waters also indicate a policy commitment to conserving marine biodiversity. For the Great Barrier Reef, both the enabling legislation (Commonwealth of Australia, 2011) and operating principles for the 2004 rezoning (Fernandes et al., 2005) are explicit about the primacy of biodiversity conservation in the region. In summary, it appears that governments and international NGOs have promoted, for the three case study contexts, the important role of MPAs and no-take zones in achieving biodiversity conservation. If the effectiveness of MPAs in protecting biodiversity is often questionable, one reason is the lack of resources for management and compliance (Mora and Sale, 2011; Rife et al., 2013). But another important factor determining the effectiveness of MPAs and no-take zones is their location relative to biodiversity features that need protection from threatening processes, addressed in the questions that follow.
- Question 2:
Do developing systems of MPAs/no-take zones give precedence to more threatened biodiversity features?
The rationale for giving precedence to biodiversity features that are most threatened (and with fewest spatial options for protection, Margules and Pressey, 2000) is that such a scheduling strategy will minimize the extent to which conservation objectives are compromised by threatening processes while systems of MPAs are being assembled. Globally, the emerging trend is toward very large MPAs in remote parts of the ocean with limited potential for extractive uses and distant from the most serious threats to marine biodiversity (and see Spalding et al., 2013). There is little evidence that large, remote MPAs are the best way of averting the decline in marine biodiversity. This approach appears to be shaped more by political pragmatism and by the explicit emphasis of some conservation NGOs than by insights into effective ways of maximizing the long-term persistence of biodiversity. Large and remote MPAs are in many cases the only way countries can meet, at minimal cost and political risk, their international conservation commitments. At the same time, despite possible benefits, the contribution of these very large MPAs to the most urgent conservation priorities in the world's oceans can be questioned (Agardy et al., 2003; Cressey, 2011; Anon., 2012; De Santo, 2013; Dulvy, 2013; Spalding et al., 2013).
For Australia there is no evidence that more threatened biodiversity features (e.g. coastal waters or Key Ecological Features) have been given precedence, and strong evidence for the opposite pattern. Australia's MPA system in Commonwealth waters is now so extensive (at 3.1 million km2) that residual patterns are clearly evident. Indirect evidence consists of very uneven representation of provincial bioregions and strong spatial biases, particularly of no-take MPAs, toward deeper waters distant from the mainland. More direct evidence relates to biases away from areas valuable for commercial fishing and extraction of oil and gas, particularly in the case of no-take MPAs. Both of these activities are known to have impacts on and to pose future risks to Australia's marine biodiversity (SoE, 2011), and some poorly protected species and ecosystems are in decline (SoE, 2011).
For the Great Barrier Reef, precedence of threatened features was, in one sense, not an issue. The whole zoning system, after protracted design and public consultation, was enacted simultaneously, so a sequence of protection was irrelevant (although whether the zoning system is ‘complete’ remains open to debate). The rezoning did, however, consider ecosystems and species known at the time to be threatened, including marine turtles and dugong (Fernandes et al., 2005, 2009). The assessment ‘Uncertain’ in Table 2, is influenced by two considerations. First, the rezoning had only a marginal effect on the extent of pre-existing trawling in soft-bottom bioregions (Figure 9, and see Grech and Coles, 2011). Second, there was a clear tendency for no-trawling zones to be configured around previous trawling activities, albeit with a minimum of 20% protection of all bioregions in no-take zones (Figure 9). This underlines the need for conservation objectives for bioregions and other features in the region to be scaled according to exposure to extractive uses. It is also acknowledged that the rezoning was preceded by a trawl management plan in 2001, although this plan was not focused on maintaining the region's biodiversity. Importantly, trawling is currently permitted over extensive parts of some bioregions (up to 78%, Figure 9).
In summary, there is an emerging residual pattern of MPAs globally, a strongly established residual pattern in Australian waters (and see Pressey, 2013), and some indications of residual protection in the Great Barrier Reef. At least for Australia and the Great Barrier Reef, these patterns have clearly been shaped by an emphasis on minimizing opportunity costs. Globally, the distribution of MPAs strongly suggests the influence of minimizing opportunity costs, perhaps by the political expedient of avoiding conflict with resource extractors in near-shore, heavily used waters. Moreover, the effectiveness of zoning and management of some very large, remote MPAs is questionable (Cressey, 2011; Dulvy, 2013).
- Question 3:
Do MPAs/no-take zones adequately represent all biodiversity features of interest?
The global MPA system covers a small percentage of the world's oceans, so representation, even at the coarse resolution of marine ecoregions and pelagic and benthic provinces, is inevitably poor (Spalding et al., 2013). This limitation is reinforced by the fact that MPAs are rarely no-take zones, spanning a broad range of protection types that do not necessarily avert threats to biodiversity. The commitment by governments and large NGOs to filling gaps in global representation remains uncertain despite explicit international policy statements about representation (e.g. Aichi Biodiversity Targets 7). While policies aimed specifically at establishing very large, remote MPAs help to increase the world's MPA coverage and protect large, relatively pristine areas, they accelerate the already strong trend towards large and remote MPAs (De Santo, 2013; Spalding et al., 2013) and uneven representation. Across a sequence of international conventions and conferences, there have been occasional proposals for quantitative objectives usefully framed in relation to marine ecoregions and habitats (De Santo, 2013), but there appears to be no international consensus on such objectives. Also, vagueness about the spatial context for objectives (such as Aichi's ‘10% of coastal and marine areas’) and objectives framed for national jurisdictions (CBD COP, 2010 8) could be counterproductive by encouraging politically expedient, highly biased protection (Agardy et al., 2003; Melick et al., 2012; De Santo, 2013).
Nationally, well considered principles for expanding the Australian Commonwealth MPA network (ANZECC, 1998) appear to have been discarded in designing the 2007 (Nevill and Ward, 2009) and 2012 MPAs. These very extensive MPAs were apparently not based on any quantitative objectives and, by any standards, failed to adequately represent many of the environmental features that had been mapped specifically for the bioregional planning process. Biases in representation were stronger for no-take MPAs than for all MPAs combined. Regionally, the 2004 rezoning of the Great Barrier Reef Marine Park remains one of the world's best examples of representing marine biodiversity, as well as attempting to promote the persistence of key processes (Fernandes et al., 2005). With hindsight, however, the 20% representation objective for all bioregions in no-take zones, although testing political will at the time, could now be improved. More sophisticated, variable objectives for individual ecosystems and species are needed to reflect factors such as spatial turnover of species within ecosystems, genetic heterogeneity within species, and exposure of features to threatening processes.
In summary, representation is very poor globally and hindered by the lack of explicit objectives. In Australia, the lack of objectives for recent bioregional planning was a retrograde step, with representation remaining uneven and, for some features, very poor. The approach to representation in the Great Barrier Reef, exemplary in 2004, would benefit from refinements. All three case studies indicate that uneven representation is related to minimizing opportunity costs in the form of short-term financial and political liabilities.
- Question 4:
Do MPAs/no-take zones adequately represent more threatened examples of features that are different from less threatened examples?
Globally, there is little information on variation within marine ecoregions or pelagic and benthic provinces; regardless, given the scale of intra-habitat variability in ecological features, it is likely that heterogeneity of both biological composition and extractive potential will be high in many such extensive features. Given the very poor and increasingly residual representation of many of these features, it is not possible for within-feature variation to be adequately addressed by MPAs in relation to threats. Among the principles for MPA expansion previously established in Australia (ANZECC, 1998) is one that concerns the representation of physical and biological variation within mapped features such as provincial bioregions. The available evidence indicates that within-feature representation is not high in the new Commonwealth MPAs, despite many of the mapped features, such as provincial bioregions, being very extensive and probably very heterogeneous. Williams et al. (2009a) demonstrated that geomorphic units used for planning MPAs in the South-east region were heterogeneous physically and biologically and that MPAs, and particularly no-take MPAs, covered a biased (least threatened) portion of this variation. The configuration of MPAs, especially no-take MPAs, around commercial uses in the Australian marine jurisdiction offers little promise that more threatened within-feature variation has been represented. For the Great Barrier Reef, it is clear that pre-existing trawling influenced the distribution of no-take and other protective zones designed to achieve objectives for marine bioregions. What remains unclear is whether this bias is associated with variation in biodiversity within bioregions. In summary: within-feature variation is poorly known globally and certainly not represented; in Australia, the few published analyses and residual biases at the feature level strongly suggest poor representation of threatened within-feature variation; and in the Great Barrier Reef it is unknown whether within-bioregion suitability for trawling is also associated with variation in biodiversity.
Overall, Table 2 indicates clearly that, globally and in Australia, a commitment to protecting marine biodiversity with MPAs has not been matched by action. Minimizing opportunity costs is leading to perverse outcomes for marine biodiversity. Protection is concentrated on ecosystems and associated species under least threat, while much biodiversity exposed to threats remains so, and is declining as a consequence. For the Great Barrier Reef, perverse outcomes of minimizing opportunity costs are possible. Objectives for the Reef's features need to be refined, partly in relation to exposure to threatening processes, and the implications for the Reef's biodiversity of minimizing costs to trawling are not understood.
Challenges for science and policy
Thirty years of systematic conservation planning have contributed greatly to designing effective systems of MPAs, and influenced policy and practice. Principles developed in Australia for establishment of MPAs, shaped strongly by systematic planning, include comprehensiveness, adequacy and representativeness (ANZECC, 1998). But the recent exercise in designing very extensive MPAs in Australian waters demonstrates that principles once endorsed by government can be abandoned when they would lead to politically unacceptable conflicts with resource users. Put another way, it seems that the opportunity costs of a comprehensive, adequate and representative system of MPAs in Australia were too high for the Australian Government to pay. Despite the small impact on existing extractive activities of the MPAs announced in 2012, the Australian government elected in 2013 started a review of the MPAs, suspending their management plans. This problem is similar to that faced by many conservation initiatives around the world that have to confront political and economic realities. The solution is often to aim for the ‘low hanging fruit’ in an attempt to demonstrate willingness to establish protected areas, even if the long-term costs – to society in the form of lost biological heritage as well as biodiversity – of continuing declines of ecosystems and species under extractive pressures are not assessed.
More extensive application of the principles of systematic conservation planning would help reverse the weaknesses of MPA systems described in Table 2. However, despite well known success stories in the application of systematic planning (Fernandes et al., 2009; Gleason et al., 2010), more effective systems of MPAs would also benefit from further development of systematic methods in at least three areas, each of which will require translation into policy.
First, representation of ecosystems and species – a foundation of systematic planning (Margules and Pressey, 2000) – is necessary but not sufficient. Reviewing systems of MPAs only in terms of representation (Barr and Possingham, 2013) ignores the relative urgency for protection of species and ecosystems. Chronological analyses of the development of terrestrial reserve systems have shown that progressive increases in representation reflect ‘protection’ of less threatened features while more threatened features remain exposed to further attrition (Pressey and Taffs, 2001b; Pressey et al., 2002). Measures of representation per se therefore need to be refined to reflect urgency for protection, and preferably complemented with estimation of conservation impact (Ferraro and Pattanayak, 2006). Importantly, impact refers to outcomes for conservation attributable to protected areas relative to the counterfactual of no intervention, considering the potential decline of unprotected features. Measuring impact therefore addresses the ultimate goal of conservation, whereas representation (and especially extent of protected areas) falls short in this respect. Retrospective analyses of the impact of protected areas on land are now operational (Andam et al., 2008; Nolte et al., 2013), providing lessons for the future. Predictive analyses to maximize future impact are also being developed on land (Withey et al., 2012). Both approaches need adaptation and application in the marine realm.
Second, a growing literature on using spatially variable costs in systematic conservation planning (Ban and Klein, 2009) has not come to terms with the risks entailed in minimizing the costs of achieving conservation objectives. The risks to biodiversity of minimum-cost conservation solutions are strongly related to the spatial resolution and heterogeneity of features identified for representation. But the risks to biodiversity might persist even when using relatively finely defined features such as the 70 marine bioregions in the Great Barrier Reef Marine Park. Among the unanswered questions related to costs are: 1. To what extent do apparent ‘win-win’ solutions that achieve objectives at minimum cost disadvantage biodiversity features most exposed to threats posed by extractive activities? 2. How do perverse outcomes from minimizing costs relate to the resolution at which conservation features are defined, relative to the resolution of cost data? 3. What measures best promote the persistence of features that are most costly to protect? 4. To what extent must society forgo economic gain or incur economic losses if the commitment to conservation is real?
Third, the respective benefits and risks of large MPAs in remote areas with little current threat and smaller MPAs in imminently threatened and heavily used waters remain poorly understood (but see Spring et al., 2007 for a terrestrial analysis). While this situation prevails, the debate will be shaped more by belief systems than real understanding, and the prospects will remain poor for designing balanced portfolios of MPAs that maximize long-term outcomes for marine biodiversity. Claims that remote, residual MPAs are good investments for the future are valid only if it can be demonstrated that this strategy gives better long-term outcomes for biodiversity than an alternative strategy based on addressing urgent priorities in relation to threat. There is a pressing need for the assumptions involved in advocacy for both strategies to be laid out and examined, and the implications of these assumptions being wrong to be identified. With this foundation, it will be possible to better understand the long-term conservation impact of shifting the balance of protection between remote and inshore marine areas. Assessments of the long-term outcomes of short-term decisions also need the context of climate change. Latitudinal shifts in species are already observed in many terrestrial (Chen et al., 2011) and marine (Cheung et al., 2009) ecosystems, and changes in ocean temperature and acidity will affect species and ecosystems in both coastal and remote oceanic environments.
This study has explored the residual nature of MPAs at different geographic scales, a consequence of the complex trade-offs between ecological, socio-economic and political considerations. The study demonstrates that reaching targets defined by the extent of MPAs, or even targets related to representation of marine features, can give governments, NGOs and the public a false sense of achievement for conservation, with potentially perverse outcomes for marine biodiversity. To expose and help reduce such residual tendencies of MPAs, we proposed a series of four questions or retrospective assessments that funders and planners should be able to address in the interests of accountability. Navigating these questions and assessments requires scientists and practitioners to develop a more explicit view of the consequences of minimizing the opportunity costs of marine conservation.