Coral reefs across the globe are in decline. Approximately 19% of reefs are considered to be lost (degraded to a point that they are non-functional or out of existence), 15% are in critical condition (likely to be lost in the next 10-20 years), 20% threatened (may be lost in the next 20-40 years), and 46% at low risk (are stable or are recovering quickly) (Wilkinson 2008). The threats causing this decline include global warming, unsustainable fishing practices, land based pollution, disease, and invasive species (Brown 1997).
These threats drive changes to reef processes such as reduced grazing (Hughes et al. 2007), increased algal growth (Fabricius 2005), reduced coral recruitment (Hughes and Tanner 2000), increased coral mortality (McCook 2001), increased bioerosion (Glynn 1997) and reduced accretion (Rooney et al. 2004), which may cause phase shifts from coral- to macroaglal dominated reef communities (Gardner et al 2003, Hughes et al. 2003, Pandolfi et al. 2003). The concern is that these phase shifts may not be readily reversed (Scheffer and Carpenter 2003).
A management tool that is increasingly advocated to mitigate coral reef decline is the marine protected area (MPA). There is extensive evidence that MP As enhance targeted fish populations (Palumbi 2001, Halpern 2003), and a number of studies showing cascading effects of protected fish populations on predation processes and the distribution of benthic organisms (Newman et al. 2006). An example of demonstrated cascading effects of enhancement offish populations within MP As includes increased parrotfish grazing levels, which results in reduced macroalgae on the reef (Mumby et al. 2006).
Cropped algal biomass may in turn create the opportunity for coral to proliferate, by not interacting with coral (Nugues and Bak 2006, Box and Mumby 2007). Enhanced populations of triggerfishes, puffer and wrasses have also been shown to increase predation levels on urchins, resulting in decreased urchin abundance (McClanahan et al. 1994), and reduced levels of bioerosion (Carreiro-Silva and McClanahan 2001). Despite a wealth of information on MPA effects, the direct influence of marine protection on specifically coral communities has yet to be fully established.
As the extraction of fishes by humans is not the only factor influencing coral reef communities, the ability of MPAs to enhance corals and promote coral reef resilience, may be somewhat limited. The other factors which may separately influence coral reef communities include: 1) larval dispersal to the reef (Connell et al. 2004); 2) abiotic conditions on the reef such as temperature, salinity, currents and wave energy (Storlazzi et al. 2005); 3) resources available to the reef organisms such as space, light, habitat and food (Friedlander et al. 2003); 4) biological interactions occurring on the reef (Hay et al. 2004); and 5) disturbance regimes that influence the various biological interactions (Sousa2001).
Efficacy of MPA
The popularity of MPAs has exceeded our knowledge of how best to design them and what we can expect from them. While MPAs are widely promoted, there is no guarantee that they will work as intended. Results from a recent meta-analysis showed that one third of the marine populations studied had either unchanged or decreased density due to MPA establishment (Halpern 2003). Because the implementation of MPAs incurs socioeconomic costs in terms of lost or displaced fishing opportunities, the answer to the question of design cannot be simply be 'bigger is better'. Due to differing processes on land and water, much of terrestrial reserve theory cannot be directly applied to the marine realm (Carr et al. 2003). The result is that the theory behind MPA design is newer and much less developed than terrestrial reserve theory (Gerber et al. 2003). In recent years, MPA theory has grown from rules of thumb based mainly on intuition (Lauck et al. 1998), to more specific guidelines learned from strategic modeling (Gerber et al. 2003), to tactical modeling approaches for implementing MPAs in specific places (Airame et al. 2003, Kaplan et al. 2009).
Previous modeling studies have found that the efficacy of MPAs in terms of population persistence and yield depends on several factors including, exploitation rate, larval dispersal and adult movement, and life history characteristics of the population (Gerber et al. 2003). The optimal design of an MPA network also depends on whether its goal is to maintain high fisheries yield or to conserve biodiversity (Hastings & Botsford 2003). Studies of marine populations with dispersing larvae have shown that populations with short larval dispersal tend to persist in all MPAs, whereas populations with long larval dispersal require larger MPAs or a network of MPAs in which a certain critical fraction of the coastline is protected (Botsford et al. 2001, Kaplan et al. 2009). A smaller number of modeling studies have also included the effects of adult movement (Polacheck 1990, DeMartini 1993, Walters et al. 2007). One commonality of these models is adult movement is represented by diffusion, which involves a constant flux of individuals away from a source. They found that large adult movement rates limit MPA efficacy by limiting gains in spawning stock biomass within MPA boundaries. Altogether, these studies have shown that species with smaller spatial scales of larval and adult movement are typically better protected within MPAs than widely-dispersing species.
Persistence and yield resulting from management with MPAs depend strongly on the level of exploitation outside the MPAs (Gerber et al. 2002, Botsford et al.
2003, Kaplan et al 2006). MPAs improve yield for populations which are recruitment over fished (when the exploitation rate is greater than the rate at which maximum sustainable yield occurs) in the absence of MPAs (Holland and Brazee 1996, Sladek Nowlis and Roberts 1999). Hastings and Botsford (1999) found that the management of fisheries with MPAs can lead to equivalent yields as compared to conventional management.
MPA Design
Marine protected areas (MPAs) are locations where certain anthropogenic disturbances (primarily fishing) are prohibited with the goal of conserving biodiversity and/or improving fisheries management. Effective MPA design is complicated by the movement of individuals across MPA boundaries. Marine organisms vary greatly in their movement ability, both in the planktonic larval stage (Kinlan & Gaines 2003) and as adults (Lowe & Bray 2006). Species with smaller spatial scales of larval and adult movement are typically better protected within individual MPAs than widely-dispersing species (Kaplan et al. 2009; Moffitt et al. 2009) yet in order to conserve biodiversity MPA designs must protect species with a range of movement characteristics.
Strategies for designing MPA networks have generally consisted of two approaches. The first approach relies on simulated annealing algorithms to optimize the combination of MPAs that meet habitat representation goals, with the assumption that such habitat will support persistent populations of a range of species. This approach has been successfully applied to MPA design problems in Australia (Fernandes et al. 2005), the California Channel Islands (Airame et al. 2003), and Mexico (Sala et al. 2002), thus, will be applicable in Scottland. A second approach to MPA design uses numerical simulations to determine whether an MPA network will support persistent populations of particular species (Kaplan et al. 2006). Population persistence can then be used explicitly as a criterion for comparing alternative MPA proposals. Population persistence requires that each adult replace itself within its lifetime.
Studies of marine populations with dispersing larvae have revealed two ways in which populations can persist in a system of MPAs: (1) self persistence and (2) network persistence. In the self-persistent case, enough locally produced larvae return to the same MPA to maintain replacement, regardless of contributions from other locations. Replacement in network persistence occurs through multiple dispersal paths connecting MPAs over several generations (Botsford et al. 2001, Hastings and Botsford 2006).
Species with short larval dispersal distances will generally be able to maintain self-persistent populations within MPAs, whereas species with long larval dispersal distances will typically exhibit network persistence. In general, MPA network design affects population persistence through two key variables: 1) the size of individual MPAs, which affects the protection of mobile adult individuals (DeMartini 1993; Moffitt et al. 2009; Polacheck 1990) and whether an MPA receives a sufficient fraction of locally produced larvae to be self-persistent (Botsford et al. 2009); and 2) the total fraction of a coastal region protected by MPAs, which determines network persistence (Botsford et al. 2001; Hastings & Botsford 2006).
The size of individual MPAs and their total area indirectly specify the spacing between MPAs in a network. As such, several researchers have suggested that population persistence could be ensured by tuning the size and spacing of MPAs to the movement scales of the multiple species the MPAs are intended to protect (Halpern et al. 2006; Palumbi 2004). Size and spacing guidelines are intended to inform the initial design of proposed MPA networks, and those same networks then may be compared by how well they meet the guidelines. Size and spacing guidelines are intended to account for the importance of larval connectivity and adult movement to MPA performance, but there has been no evaluation of the effectiveness of these simple rules in terms of the range of species that would have persistent populations if the guidelines were followed. That need will be addressed by using spatially explicit population models to evaluate whether MPAs with different size and spacing configurations actually support persistent populations. Increasing the minimum size of MPAs generally has a higher benefit than decreasing MPA spacing, and it is preferable to compare proposed MPA networks using spatially explicit population dynamics models that evaluate population persistence directly.
MPAs for Conservation of Species
Marine protected areas (MPAs) are portions of the ocean where at least some extractive activities are prohibited. MPAs are being established around the world for the conservation of species or entire ecosystems and/or as a tool in fisheries management (Fernandes et al. 2005, Claudet et al. 2006, Kaplan et al. 2009). While there is no guarantee that MPAs will work as intended, they do have socioeconomic costs in terms of lost or displaced fishing opportunities. Biological monitoring programs may be employed to determine whether they are meeting their particular goals. It is often suggested that results from monitoring may be used to assess the effectiveness of MPAs as early as five years after establishment (Ojeda-Martinez et al. 2007, CDFG 2009).
MPAs are implemented to meet goals such as protecting the abundance or diversity of marine species, conserving the structure and function of ecosystems, rebuilding depleted populations, protecting marine habitat, and improving study opportunities of unfished populations or ecosystems. Because it an be an unwieldy task to measure these goals directly, the effects of MPAs on marine populations are determined from several measurements: community variables (diversity, species richness); population variables (density, abundance, biomass, size structure, maximum or mean organism size); and fisheries variables (catch, catch per unit effort) (Halpern and Warner 2002, Goni et al. 2008). MPA effects are evaluated in several ways: measurements may be compared inside versus outside an MPA at the same point in time; measurements may be compared within the MPA some time after establishment compared to before; or measurements at more than one location relative to the MPA may be plotted against time to determine if the trajectories diverge (Halpern and Warner 2002, Claudet et al. 2008). If the measured variable (fish abundance, for instance) is found to be greater inside the MPA than outside, or within the MPA some time after establishment as compared to before, the effect is positive and the MPA is deemed effective.
Detecting the Effects of MPA
The question of the detection of MPA effects is multipart. First, is there an effect of the MPA; and second, can we detect it? The first question has been considered in deterministic modeling studies (Gerber et al. 2003); the second requires incorporation of statistical detectability and measurement error. Empirical studies of MPA effectiveness do not always detect positive MPA effects; one third of the organisms in Halpern's (2003) meta-analysis had either unchanged or decreased density due to MPA establishment. Comparing the results of empirical monitoring to the expectations of theoretical modeling should gain a better understanding of why positive MPA effects do or do not occur in particular systems, but exercises such as this have been rare.
The size and age of the MPA, and the movement and exploitation level of the study organism all have potentially large effects on whether a positive MPA effect is detected. There is theoretical and empirical consensus that exploitation level is a strong determinant of MPA effectiveness, with fished populations showing stronger responses to protection by MPAs than unfished species (Botsford et al. 2003, Micheli et al. 2004, Kaplan et al. 2006, Goni et al. 2008). Movement of individuals in a population relative to MPA size has been found by theoretical studies to be an important determinant of MPA effectiveness, both in the larval stage and as adults (Gerber et al. 2003, Kaplan et al. 2009, Moffitt et al. 2009).
''Spillover'' from the MPA, due to both larval export and adult movement, can homogenize the spatial distribution of measured values by leading to increases in measured values outside the MPA, while also reducing gains within the MPA (Halpern et al. 2004, Kellner et al. 2007, Kaplan et al. 2009, Moffitt et al. 2009). Empirical studies of the effects of movement have found mixed results; some studies have found that increased MPA size (Edgar and Barrett 1999, Friedlander et al. 2007, Claudet et al.
2008), and decreased fish mobility (Fisher and Frank 2002, Goni et al. 2008) are correlated with stronger MPA effects, others have not (Micheli et al. 2004, Guidetti and Sala 2007). MPAs increase the survival of individuals in a fished population by protecting them from fishing mortality, leading to more individuals in large size and age classes with time. These larger sizes subsequently can increase the population's larval production and lead to increases in biomass and abundance. Therefore, we would expect MPA effects to increase with age, but the importance of MPA age on effectiveness as determined from empirical studies is unclear.
Some empirical studies show a positive correlation between MPA effects and age (Micheli et al. 2004, Claudet et al. 2008), while others do not (Halpern and Warner 2002, Halpern 2003). Results from theoretical modeling studies may be different from those of empirical monitoring studies because most modeling efforts are deterministic and evaluated at equilibrium (Gerber et al. 2003), while empirical studies measure transient effects (effects at certain points in time which are not at equilibrium) and cannot avoid the inclusion of measurement error.
When positive MPA effects are not detected by monitoring studies, MPAs may be deemed ineffective. However, there has been no direct formal evaluation of how likely the detection of specific effects will be if the MPA is performing in an effective manner.
We explored the question of under what time frame, exploitation level, and organism mobility relative to MPA size should we expect to detect MPA effects using a spatial, size, and age-structured population dynamic model. We modeled the population dynamically in order to capture transient effects and the importance of time. We did not include measurement error directly, instead focusing on an initial effort of a high-dimensional problem. We looked at three measurements of the effects of an MPA on spatial density and found that detectability of MPA effects increased with exploitation rate and MPA age, and decreased with population mobility (particularly adult movement)
Fisheries Management and Policy
There exists a small but growing cluster of geographic work on marine capture fisheries (Young 2001; Mansfield 2004). Most work on fisheries, especially that informing fisheries management decision making, is not generated by geographers, however. Reviews of fisheries biology, resource economics and other literatures directly informing Scotland fisheries management agencies are available elsewhere. It is also important to note that an increasingly broad and prominent literature critiques mainstream fisheries management, including anthropologists, biologists, and a few economists (Jackson et al. 2001; Pikitch et al. 2004; Walters et at. 2005). Market based fisheries management alternatives gaining increasing traction in Scotland are also heavily critiqued, especially by empirically conscientious social scientists, but also by some environmental groups and public servants supporting small business (Ecotrust Canada 2004; GAO 2004; McCain 2004).
Alternative proposals for fisheries management include co-management, in which decision authority is shared between fishing dependent stakeholders and government, and community based management, in which most or all decision authority is ceded to fishing dependent groups (Acheson et at. 2000; Berkes et at. 2001; Wilson et al. 2003). Although these are often presented as radical departures from conventional management approaches, and differentiated from one another, most developed world regulatory processes provide for some level of fishing industry stakeholder participation in rulemaking and policy development. While they may be more symbolic than substantive, and burdened with gross inequities, they are institutional realities. Further, true community based management is largely confined to very isolated settings, since modern states and global markets now permeate most comers of the developed and undeveloped world. Even in cases where resource users and other interested groups have relative autonomy from government oversight, scientists wishing to support these processes often find it difficult to integrate their standardized tools and models with local knowledge and beliefs (Degnbol 2005).
Subsequent chapters of the present dissertation review the management histories of groundfish, lobster and several other species, based on direct observation by the present author, primary source materials and documents, and secondary literatures. Discrepancies among existing scholarly and policy interpretations of these management histories help to direct the analysis of the present case study data. Some of the most scientifically relevant debates surround the following questions: 1) Should strict single species population targets and regulatory timelines be the principal vehicles for recovery of depleted fish stocks? 2) How should ecosystem and habitat variables be considered in relation to any such thresholds and timelines? 3) What kinds of public participation can best ensure accountability in resource management decision making? 4) What kinds of institutional relationships can facilitate the development and use of high quality scientific information about resource management alternatives? 5) What kinds of resource access arrangements arc socio-ecologically appropriate and sustainable? 6) What can be gained by closing areas to fishing? 7) Are lawsuits helpful in ensuring fisheries sustainability?
In addressing these questions, most major EU environmental organizations have argued that strict single species population rebuilding timelines are necessary, often including regulatory caps, or quotas, on fish harvests (Ocean Conservancy 2004; Oceana 2004). Many of these same groups have argued that ecosystem and habitat variables are important, but have chosen to address them separately from stock rebuilding targets (Oceana et al. 2002; Ocean Conservancy 2004). Many have also argued that the inclusion of fishermen in management decision making is problematic, that greater ENGO participation is helpful in those venues, and that natural scientists should be responsible for setting mandatory regulatory parameters is isolation from social variables (Shelley 2001; Occana et al. 2002; Ocean Conservancy 2004). Some have opposed marine resource privatization while others have supported it or remained neutral on the issue (Emerson 2002; Oceana et al. 2002; Environmental Defense 2004). Few have questioned the scales and jurisdictions of management and information development. Many call for closing more and larger areas to fishing (Environmental Defense 2004; Ocean Conservancy 2004). Many originate, or actively support, lawsuits as a primary strategy for enforcing fisheries laws and attracting public support (Shelley 2001).
These groups have found effective allies among marine biologists and other scholars (Cloutier 1996; Brailovskaya 1998; Safina et al. 2005). Nonetheless, a growing number of natural and social scientists, and a few place-based NGOs interested in the sustainability of fishery dependent communities, question these prevailing ERGO policy assumptions in certain respects, though few directly challenge ENGO positions. Most of these authors assert the need for fishery management to acknowledge more complex and mutually determining relationships across social and ecological variables, or across single species and ecological indicators (Cury and Christensen 2005; Walters et al. 2005). Increasingly, many of these scholars call for the development of ecosystem based fisheries management. The concept of ecosystem based management is widely recognized, but definitions, much less implementation guidelines for the marine fisheries context, are few and far between (Brodziak and Link 2002; Alverson 2004).
These have advanced more quickly in the management of other resources, such as forests, rivers and rangelands (Ecosystem Principles Advisory Panel 1999). Most systematic discussions of ecosystem based fisheries management identify the need to supplement single species management performance measures with a more holistic set of criteria, with multiple reference points. These often include system biodiversity, structure, function, productivity, boundaries, spatial dynamics, and consumptive and non-consumptive values (Cury and Christensen 2005; Link 2005; Murawski 2005). Ideally, new ecosystem approaches would encourage and utilize higher quality information about complex human-environment systems, and provide more transparency and accountability in making inevitable management tradeoffs across socio-ecological considerations. Many advocates of ecosystem based resource management, in marine and other realms, recognize the need for adaptive management. This is an iterative process of social learning, one that recognizes the limitations of predictive modeling (Lee 1993; Johnson 1999). It would institutionalize information feedback loops to incorporate new knowledge and changes in dynamic systems over time, so that management strategies can be revised frequently.
Decision-Making Arrangements
This definition distinguishes collaborative management from centralized government management and community based resource management. Centralized government management is an arrangement where responsibility for resource management is held solely by the state government (Sen and Nielsen 1996). In contrast, community based resource management is an arrangement where responsibility for resource management is held solely by the local community (Sen and Nielsen 1996).
Sen and Nielsen (1996) note that collaborative management covers a broad spectrum of possible collaborative decision making arrangements between the government and user groups encompassing: "(1) the roles of government and user groups in decision making; (2) the types of management tasks that can and want to be co-managed by user groups and government; and (3) the stage in the management process when co-management is introduced (planning, implementation, evaluation). In this investigation, mechanisms for collaborative management decision making arrangements were introduced during the planning stage of the protected area management process and continued into the implementation stage for both case studies. During the initial planning stage, government agents worked with community members to determine the form and function of the protected area network. Furthermore, the roles that government and user groups play in the decision making arrangements are 'consultative' in both case studies.
Conclusion
It is important that winners and losers are accounted for in order to increase the legitimacy of the coastal governance regime. Developing policy recommendations that address issue linkages that are important to the diverse interests of the stakeholders in the region is a step towards greater legitimacy of the regime. Furthermore, it is important to take into account the role of equity in order to understand the effects of regime change. It is also important to account for equity in order to work towards solutions that may resolve the value conflicts that developed as a result of that regime change. This value conflict is related to the problem of scaling institutions from international programs into local projects.
As the problems of global environmental change continue to affect ecosystems, approaches such as marine spatial planning, ecosystem based management, integrated coastal zone management, and marine protected area networks offer solutions.
The community of scientists interested in MPA's has touted the ecological benefits by asserting that no-take MPA's boost abundance, size, and reproduction within MPA boundaries and provide spillover and export functions outside of the boundaries (PISCO 2002). Spillover occurs when animals from no-take MPA's may swim or crawl outside to supplement surrounding populations (PISCO 2002). Export occurs when larvae and plant propagules that disperse out of reserves may seed and boost populations in surrounding waters (PISCO 2002). As a result of spillover and export, fishing boats have often been observed congregating along the borders of no-take MPA's, because that is where catches are reported to be highest (PISCO 2002). The ecological goals of the zonal system were designed so that more productive and diverse habitats would have greater protection, and degraded habitats with little diversity would have less. This means that the most productive areas were closed to artisanal fishing, although, they promised to provide the greatest ecological and fisheries enhancement benefits. These results do not undermine the theory of ecological spillover or export in any way. These results do suggest that the human response to the implementation of such regulations is more complex than previously assumed.
Furthermore, the human response does not provide equitable benefits for all resource users. Older and middle aged people with local knowledge of the area paid the greatest social cost and provided the greatest ecological benefit. The younger people with no prior experience in the area received the greatest social benefit and provided the least ecological benefit. These results suggest that tradeoffs between conservation and development are inevitable and that inequity may result if attention is not paid to local resource users.
