Are marine protected areas effective tools for sustainable fisheries management? I. Biodiversity impacts of marine reserves in the temperate zone (systematic review)

Background

Marine Protected Areas (MPAs) have been proposed and designated as a mechanism to address the problem of achieving sustainable fisheries, while simultaneously preserving biodiversity. However, empirical studies and research syntheses indicate that there is considerable variation in the ecological effects of MPAs and there is considerable debate over whether they can produce fisheries benefits. Of the various types of MPAs, marine reserves (no-take areas) have been perhaps the best studied. Furthermore, by focusing on marine reserves, it is possible to eliminate protection level as a co-variate. Consequently, as a first review in a proposed series on the effectiveness of MPAs for sustainable fisheries management, we objectively collated data to ascertain the impacts of temperate zone no take areas on the density, biomass and species richness of marine biota within reserve borders (as reflected in the subtitle).

Objectives

  • To determine the impacts of establishing temperate zone no take areas on density, biomass and species richness of marine biota within reserve boundaries.
  • To ascertain whether the impacts of no take areas vary in relation to taxon, reserve parameters or organism parameters.

Search strategy

Multiple electronic sources were searched using a range of keywords. Bibliographies, expert contacts and website searches were performed to access grey literature. Information from soft sediment systems was explicitly sought as well as the more common reef-based assessments. Foreign language searches were not performed.

Selection criteria

Relevant subject(s): All marine biota • Types of intervention: Temperate Marine Protected Areas (Marine Reserves) defined as geographically defined areas subject to no fishing activity (no take areas). • Types of comparator: No restriction on fishing activity. • Types of outcome: Density (other abundance measures), biomass, species richness. • Types of study: Any primary study providing measures before and after implementation of a reserve, or comparing reserves to adjacent geographical areas without no-take protection.

Data collection and analysis

Data extraction was undertaken using a review-specific data extraction form. Random effects meta-analysis was used to examine species level data, reserve level data, and species-reserve interactions. Subgroup analyses and meta-regression were used to explore variation in effectiveness in relation to species and reserve level co-variates.

Main results

Density of marine biota is 23% to 196% higher within marine reserve no take zones than outside. Gains in biomass within no take are 20% to 422%, but there is uncertainty surrounding these results due to lower sample sizes. Species richness within marine reserves is 10% to 130% higher within no take areas than outside reserves, although this too is based on a small sample. Fish species density is 35% to 81% higher in marine reserves than in adjacent areas and variation in fish density response is not strongly linked to any species or reserve level parameters. Furthermore, it is not clear if these differences are due to marine reserve effects or other differences between the marine reserve and comparator, such as habitat variation. This lack of distinction is the primary reason for contention regarding marine reserve effectiveness and can only be resolved by improved monitoring.

Meta-analyses indicate that differences in the response of specific species, genera or functional groups are often small with effect sizes of <0.2. Power analysis indicates that detection of statistically significant differences for these small effects requires co- ordinated monitoring across a large number of marine reserves.

Conclusions

Implications for management / policy / conservation: The available evidence suggests that temperate MPAs that impose no-take zones may achieve higher densities, biomass and species richness of marine biota within the boundaries of the no-take zone than outside. The evidence for higher densities and biomass of fish is notable in this context. However, considerable uncertainty and low predictive power arises from small sample sizes and possible confounding factors such as habitat quality.

Implications for research: Lack of data is a hindrance in the development of an evidence-base regarding the effectiveness of MPAs. Small sample sizes result in high uncertainty about the impacts on specific taxa. In particular, algae and invertebrates are understudied; insufficient data are available regarding biomass, species richness, deep water no take areas, pelagic fish species and soft sediment systems. It is also very difficult to distinguish habitat effects from reserve effects without integrated experimental monitoring programs that present data regarding the baseline condition. Before After Control Impact studies are required to allow comparison of rates of change from the same baseline condition. Large numbers of marine reserves require monitoring if small effects are to be detected. Data regarding the intensity of resource use are required alongside biological metrics for monitoring to be effective.

Background

Marine protected areas (MPAs, marine reserves) are becoming widely established across the globe (Halpern 2003) in response to two related policy drivers. Firstly, traditional forms of fisheries stock management are unsustainable as evidenced by the collapse of many fisheries (FAO 1994). Secondly, traditional management such as not exceeding maximum sustainable yield estimates, do not address the multiple anthropogenic impacts on marine biota. MPAs have been proposed as a mechanism to address both these problems by maintaining sustainable fisheries, whilst simultaneously preserving biodiversity (Plan Development Team [PDT] 1990, Ballantine 1992, Dugan & Davis 1993, Bohnsack 1996, Nowlis & Roberts 1997, Allison et al. 1998, Lauck et al. 1998).

MPAs are located in coastal (Micheli et al. 2004) and offshore habitat (Halpern 2003), in temperate and tropical biomes, and are designed to protect a wide range of taxa from plants and invertebrates (Edgar & Barrett 1999) to whales (Gerber et al. 2005), as well as the target species of commercial fishing efforts (Mosquera et al. 2000). Their success (or otherwise) has been monitored using a range of outcome measures. Recent quantitative reviews have synthesised parameters such as biomass, species richness, density (and other abundance measures), and organism size (Mosquera et al. 2000, Halpern 2003, Micheli et al. 2004, Kaiser et al 2006, Halpern et al. in prep). Theoretical models have also been employed to predict impacts (Gerber et al. 2005, others reviewed by Halpern 2003). The general consensus to emerge from this work is that MPAs are effective (COMPASS and NCEAS 2001, Halpern 2003). MPAs can rebuild stocks through enhanced recruitment and spill-over effects, maintain biodiversity, buffer marine systems from human disturbances, and maintain the ecosystems that fisheries rely on (MEA). However, some studies have found that MPAs have not effectively maintained biodiversity (Hilborn et al. 2004; Edgar and Barrett 1999; Willis et al. 2003). In many cases failure was due to either not including MPAs as part of a broader coastal management system or a lack of management funding, or enforcement (MEA). In addition MPAs can simply displace fishing effort to other areas and increase the vulnerability of other stocks and endangered species (Coleman et al. 2004).

Many other questions regarding effectiveness remain unresolved, particularly with respect to the impact of ecological and methodological co-variables or effect modifiers (Mosquera et al. 2000, Halpern 2003, Micheli et al. 2004). Variation in the effectiveness of MPAs has been explored in relation to a wide range of co-variables (effect modifiers, reasons for heterogeneity).

Taxa (family or functional groups: inveterbrates, herbivorous fishes, planktivorus/invertivorus fish, carnivorus fishes) (Halpern 2003), (trophic groups: herbivores, detrivores, omnivores, invertivores, planktivores, piscivores (Froese & Pauly 2006, Micheli et al. 2004), and genera (Mosquera et al. 2000) have all explained variation in species response to MPAs. Different groupings of species produce different results because weak responses of individual species may result in greater effect sizes when species are pooled in functional groups, and conversely strong individual responses may be obscured (Micheli et al. 2004); (this is a compelling reason to define subgroups prior to analysis). Phylogenetic relationships also pose problems of independence (Mosqueira et al. 2000) that are hard to address.

Reserve size (Halpern 2003), adult mobility (range) (Kramer & Chapman 1999), community composition (similarity index, Bray &Curtis 1957) (Micheli et al. 2004) organism size (Mosqueira et al. 2000), fishing intensity (Mosqueira et al. 2000) and habitat (Nilsson 1998, Kaiser 2006) have also been proposed as variables influencing the effectiveness of MPAs. Interestingly, there is consensus that size of reserves has little or no proportional impact, although there are lots of arguments (with theoretical support) for the creation of larger rather than smaller reserves.

As with taxa, problems with independence and standardisation hinder interpretation of other relationships. In particular, measures of fishing intensity around reserves (Nowliss & Roberts 1997) are difficult to assess because they change over time and it is difficult to compare fishing intensities from different parts of the world (Halpern 2003). The policy community is also interested in the impacts of connectivity (JNCC pers comm.). This is seen as a key element in terrestrial reserve design (Williams et al. 2004), but despite a large literature on issues of marine connectivity there are a paucity of data in a marine context.

Here we attempt to build on the existing syntheses using systematic review methodology (Pullin & Stewart 2006) to address questions regarding the effectiveness of marine protected areas. The use of a protocol will provide clear evidence of a priori reasoning and will allow the marine and meta-analytical communities to comment on methodology. Broadening the scope of the quantitative analyses will provide a larger sample size providing more statistical power to address the questions regarding variation in effectiveness, whilst simultaneously reducing the probability of confounding. The use of a transparent methodology will minimise bias and allow updating of the analyses when more data and more sophisticated meta-analytical techniques become available. The review will also allow discrepancies in the magnitude of effect between the existing syntheses to be explored. Cote et al. (2001) found a non-significant 25% overall increase in fish density in marine reserves whereas Mosquera et al. (2000) indicated a statistically significant 3.7-fold overall increase and Halpern (2003) an approximate doubling. Confounded baselines have been suggested as a primary reason for this variation (Edgar et al. 2004), which could result in widely different policy implementation regarding marine reserves. A further problem with existing syntheses is the over-representation of MPAs in tropical reef systems. This reflects paucity of knowledge regarding the impact of MPAs in temperate non-reef systems, particularly on soft sediments, and offshore (MEA, Kaiser et al. 2006.). Here we will examine any variation in effectiveness across habitats using marine classification systems to define subgroups, and will explicitly search for data from soft sediment systems.