NETWORK DESIGN: ECOLOGICAL CONCEPTSOverview................................................................................1

MPA size.................................................................................3

Representivity........................................................................6

Replication............................................................................10

Connectivity.........................................................................12

Case Studies.........................................................................15

References...........................................................................18

OVERVIEW

The information presented here is a summary of ecological concepts relating to Marine Protected Area (MPA) network design theory and practice. It is intended as a guide for the Forum as to international science-based information on creating viable MPA networks in the face of limited data. The summary provides information that is specific to meeting the Policy objective. While other aspects of marine management such as fishery sustainability or land use do not form part of the terms of reference for the Forum, it needs to be acknowledged that there are existing management regimes in place that are important considerations in developing networks. MPAs are not tools to be considered in isolation of other management practices, but rather should be considered one tool that contributes to overall management in the marine space. The role of the Forum is to make recommendations on MPAs to provide effective protection of biodiversity, in particular considering representation.

The objective of the Marine Protected Areas Policy and Implementation Plan (the MPA Policy)[1] is to:

Protect marine biodiversity by establishing a network of MPAs that is comprehensive and representative of New Zealand’s marine habitats and ecosystems.

The MPA Policy includes a number of design principles to guide development of the marine protected area network in meeting the Policy objective. The Policy also lists some relevant literature that was available during the writing of the Policy to provide general guidance on developing a network1. Since 2007 a substantial amount of additional relevant literature has been published, and is

1 MPA Guidelines pg.47 “Network Design Principles”.

included in this summary where relevant. International examples of processes that have utilised ecological concepts are also given for further information. The guidance that these processes developed were based on scientific concepts, research, and best practice, as well as social and political factors relating to their jurisdiction. As such, the examples should not be applied uncritically to New Zealand but are included to demonstrate how ecological concepts can be applied to the real world.

In developing its recommendations, the Forum must determine what sites, taken together, could form part of an ecologically viable MPA network for the biogeographic region, and thus contribute to the national MPA network.

MPA SIZE

DEFINITION

MPA size is an important consideration relating to an individual MPA being able to maintain the integrity of its features, and be self-sustaining throughout natural cycles. Size is an important component of viability and representation.

RATIONALE

MPA POLICY

The MPA Policy and the Guidelines do not prescribe minimum dimensions for individual MPAs. But, the Guidelines2 state MPAs “should be of sufficient size to provide for the maintenance of populations of plants and animals.” The Guidelines also state, “it is desirable to protect fewer, larger areas rather than numerous smaller areas” which reduces potential edge effects.

In order to represent longitudinal (cross-shelf) differences, the MPA Policy suggests, “It may be convenient to extend protected areas from the intertidal zone to deep waters offshore.” This will also help to ensure linkages between habitats are protected and could result in fewer, larger MPAs rather than many, smaller MPAs.

WHY DOES SIZE MATTER?

The size of an individual MPA is a core element of determining whether it is able to maintain the integrity of its natural and physical features and be self-sustaining throughout natural cycles.

VULNERABILITY TO FISHING

Aspects of vulnerability to fishing mostly relate to Type-1 marine reserves, where the emphasis is on eliminating fishing pressure altogether. In contrast, in Type-2 MPA the emphasis tends to be on habitat protection and large scale processes, so exploited species may still have some vulnerability to fishing, regardless of size.

In establishing a marine reserve, the level of protection afforded to a species is dependent on the size of its home range and the details of management outside the reserve. That is, the choice of reserve size will determine what subset of species present will potentially benefit from the protection measures. For example, a modelling study of fish home range in respect to protection[2] concluded that, to reduce fishing mortality to 2% of that that occurs outside the reserve, a reserve should be at least 12.5 times larger than the home range of the species. Figure 1 represents this graphically for a species with a large home 2 Page 21.

range (red arrow) where, assuming fishing occurs close to the boundary, the potential benefit from scenario A is greatly diminished due to the daily movements making them available to fishing for at least part of their activity. In a larger reserve as with scenario B, exposure to fishing may still occur creating a gradient in abundance across the reserve boundary (edge effect), but there remains a core area where the species has low vulnerability to being caught.

Figure 1: Adult exposure to fishing from reserves of different sizes (as indicated by the box with the solid border). In scenario A, the small reserve size compared to fish home range results in high vulnerability to fishing, even within the reserve. Scenario B has a core protection area (white square) where vulnerability to fishing is very low, with a gradient of increasing fishing vulnerability towards and across the reserve boundary.

An additional consideration relating size to vulnerability to fishing is protecting whole habitats. In some cases natural barriers to species’ movements occur, such as a reef system surrounded by sand. Some reef fish do not move across bare sand and are therefore largely restricted to their home reef. In cases such as these, a smaller reserve may provide better protection than a reserve where the reef is crossed by the reserve boundary. In Figure 2, scenario A has a larger core protection area than in scenario B, even though the reserve is the same size (for a species where sand is a barrier).

Figure 2: The effect of different habitat structures on reserve effectiveness. Scenario A has a natural buffer of sand around a rocky reef within a reserve. Scenario B has a reef that crosses the reserve boundary.

GUIDANCE ON SIZE

Generally, MPAs will be most effective for biodiversity protection if they are larger than the distance the majority of individual adult and juvenile fish and invertebrates move within and across habitats. Unfortunately there is limited information on the home-range of many species within New Zealand biogeographic regions. Most invertebrates move relatively small distances and may potentially benefit from smaller reserves, for example paua. However, to maintain, or restore, ecosystem functioning it is important to protect species that exert substantial influence on the ecology of the area through biogenic or trophic effects; such species are sometimes called ecosystem engineers. For example, evidence from Leigh marine reserve suggests that rock lobster has the potential to drive tropic changes in the ecosystem, at least under certain conditions[3][4][5]. Many other studies discuss the optimal size of reserves with reference to species and larval movement [6][7][8][9][10].

Because of the limited species information from New Zealand, we need to rely on generalities and evidence from elsewhere that can be applied in New Zealand. Creating a network of variable sized and connected MPA is most likely to ensure protection is afforded to the widest range of species, while minimising impacts on existing users[11][12].

REPRESENTIVITY

DEFINITION

Representation refers to the inclusion of each habitat type within a marine reserve. To be included as ‘representative’, the habitat must be of sufficient extent and quality to enable the maintenance and/or recovery of biological diversity at the habitat and ecosystem level in a healthy functioning state.

RATIONALE

MPA POLICY

In terms of the MPA Policy, in order to protect the full range of marine biodiversity, 'representation' requires each of the 44 classified habitat types that occur within a bioregion to be represented in a no-take MPA. The MPA Policy also requires “outstanding, rare, distinctive or internationally or nationally important marine habitats and ecosystems”3 be included in the MPA network. Within a region, the Guidelines state, “care should be taken to identify potential protected areas sites that include differences in habitats and ecosystems that cover both latitudinal and longitudinal or cross-shelf ranges.” Latitudinal in this case is to capture variations in biodiversity at a local scale that are not reflected at the biogeographic region scale. Longitudinal and cross-shelf reflects the variation from the coast to the deeper offshore waters.

WHY IS REPRESENTATION IMPORTANT?

Ensuring representation in MPA networks is recognised as important internationally. Virtually all modern MPA processes refer to representation, and it is a core principle included within the Convention on Biological Diversity (CBD) Aichi Target 114. It means not just picking the ‘hotspots’, or the areas that have least conflict[13], but to ensure all habitats are included. This recognises areas that may have lower species richness are important habitats to include within a network.

Ensuring representation of all habitat types maximises the overall biodiversity gains, while minimising the total area that needs to be put under protection to achieve the same level of biodiversity protection. As such, representation has important consequences for the efficiency of the planning outcomes. For example, Figure 3 shows seven different MPA options (A–G). At first glance, and driven by the desire to encompass the greatest degree of biodiversity, it is

3 See paragraph 16, p10 of the MPA Policy.

4 https://www.cbd.int/doc/strategic-plan/targets/T11-quick-guide-en.pdf

intuitive to select the most diverse (species rich) sites first, those labelled A and B in Figure 3.

This illustrates a common ad hoc approach of selecting ‘hotspots’ of biodiversity at the expense of other habitats. However, this is often not the most efficient approach; in this case three MPAs are required to represent the full range of biodiversity because site C is also required to represent all species and habitats. If a systematic approach is taken, the same biodiversity could be included in MPAs by selecting only sites C and E (this is often referred to as complementarity, i.e. the 2 MPA complement each other in achieving the goal).

Figure 3: Representivity of species/habitats in marine protected areas. A-G show different potential MPA sites, with the symbols representing different species or habitat types (Redrawn and adapted from Marxan “How to choose marine reserves” video5).

HABITATS AS SURROGATES

Representation would ideally be based on actual biodiversity information, such as the distribution of species and their habitat requirements. Unfortunately this information is not generally available so we need to rely on what information we do have to approximate biodiversity. The MPA Policy approach is essentially a systematic planning process. At its core is a classification system to systematically define different habitat types that are to be represented, as a surrogate for the actual biodiversity. Achieving adequate representation across the range of surrogate habitats is then assumed to provide adequate

5 https://www.youtube.com/watch?v=1IDeKJJO7s8

representation of actual biodiversity. Including all habitats in a MPA network increases the range of biodiversity protected and increases the likelihood of preserving ecosystem process that operate across different habitats.

ADEQUACY

'Adequacy' describes the concept of ensuring that the individual components of a protected area network are of sufficient size, shape and appropriate spatial distribution to ensure the ecological viability and integrity of populations and species. MPA networks should be self-sustaining or viable in the sense that they must be able to maintain the persistence of populations and ecosystems through natural cycles of variation[14].

The Policy does not refer to ‘adequacy’, but states the "network should be viable"6. For a network to be viable it must include enough of each habitat to sustain the ecological objectives (references [15]–[17]) under the MPA definition of "the maintenance and/or recovery of biological diversity at the habitat and ecosystem level in a healthy functioning state"[18].

In terms of proportion of habitat protected and its implication for overall biodiversity protection, the number of species (species richness) increases with increased area of habitat, which can be modeled/measured using species accumulation curves (for example Figure 4 below). The concepts around species/area relationships are well understood and the general curve of the graph below is consistent across habitat types and geographical area. However, it is important to note that while the general shape of the curve is consistent, the slope of the curve can be quite different across habitats and between types of organism. This means that, for different habitat types or different species, a greater or lesser amount of area is required to represent biodiversity. This becomes an important consideration when investigating cost/benefits and tradeoffs, as the steepness of the curve determines how quickly benefits are realized with minimal area protected. Unfortunately there is limited information that directly looks at habitat area and species diversity within NZ7, however overseas examples [15] [19]can provide some guidance as to how different habitat types may vary within NZ.

6 MPA Policy 2005, Network Design Principle 3.

7 Note that there is a substantial literature from within NZ on species/area curves for many habitats at a finer scale that lend support to the relevance of overseas research, though they do not specifically address habitat extent.

Figure 4: Example of the species-area relationship for intertidal sediments in the United Kingdom[15]

Internationally, there is little consensus on what percentage should be applied, however many organisations and researchers have proposed targets. The CBD target of 10% of coastal and marine areas to be conserved was proposed to ensure the survival of 50–70% of the species within the area. Other researchers have proposed targets of between 20 and 50%, dependant on habitat type, to protect 70–80% of the species within the area [19], and the World Parks Congress proposed preserving 30% of each habitat type in 2014. The MPA Policy does not specify a percentage target and care needs to be taken when interpreting targets for protection in a NZ Policy context. Even the CBD target that New Zealand has signed up to does not directly apply to the Policy objective and requirements.

REPLICATION

DEFINITION

Replication is the protection of the same habitat type across two or more sites within a network.

RATIONALE

MPA POLICY

Network Design Principle 3 states, "The number of replicate MPAs included in the network will usually be two. However, in circumstances where a habitat or ecosystem is particularly vulnerable to irreversible change, more replicates may be established as a national priority." The MPA Policy also states, “A marine reserve will be established to protect at least one sample of each habitat or ecosystem type in the network.”8

WHY IS REPLICATION IMPORTANT?Replication of representative habitats within a network of MPA is suggested as an insurance policy for biodiversity in that it promotes ecological resilience. Ecological resilience is the ability of a population or community to return to some “natural” state after a disturbance. Multiple replicates of each habitat reduce the risk of losing biodiversity due to a catastrophic natural or anthropogenic disturbance [20].

That is, if one example is destroyed through natural or human derived events (as shown in Figure 5) there remains healthy habitat that can resupply the impacted area once conditions improve. The distance between replicate habitats to enable effective recolonisation could be in the order of metres or kilometres, and is dependent on the species present and their movement and dispersal methods.

8 See paragraph 93, page 19 of the MPA Policy.

Figure 5: Schematic of mosaic of biogenic habitat patches undergoing disturbance and recovery (Source: [21]). If a patch is disturbed, its recovery is determined by presence and proximity of habitats that can recolonise the disturbed patch.

Connectivity (defined below) is also important in relation to replication and recovery dynamics. If replicate MPA are too far away from the impacted area, the benefits from having a replicate are diminished in terms of their ability to assist the recovery of the impacted site. Re-colonisation may, of course, occur via biodiversity in the areas between MPAs, as long as the impacted MPA is not “unique”.

CONNECTIVITY

DEFINITION

Connectivity is the extent to which populations in different parts of a species’ range are linked by the movement of eggs, larvae or other propagules, juveniles or adults.

RATIONALE

MPA POLICY

The MPA Policy notes connectivity as an aspect of the viability of the network.9 The Guidelines state, “the design of the protected area network should seek to maximise and enhance the linkages among individual protected areas, groups of protected areas within a given biogeographic region, and across biogeographic regions.”

WHY IS CONNECTIVITY IMPORTANT?

SOURCE AND SINKS

Some species’ larvae (particularly demonstrated for fish larvae) can swim long distances, including both aligning themselves with oceanographic features and going against dominant currents[28] [29]. This allows for a greater level of self-recruitment than the previous assumption of larvae as passively transported particles, and has implications for size and spacing of reserves[24].

Maintaining a supply of recruits is important for the resilience of a habitat, whether that source is from another MPA or habitats outside the MPA network. Ensuring that MPAs are connected maximizes any benefits protection may have on levels of recruitment. In some circumstances, a reserve may occupy an area that has no self-recruitment. In this case the location is a ‘sink’ where it is entirely dependent on an external source population, and does not contribute to larvae supply (Figure 6). A reserve at this location (on its own) has lower resilience as it is dependent on other areas. Protecting an area that operates as a sink for a species may still have ecological value, but long term it is most viable if source habitats are also protected. This protection may be through existing management outside the MPA network (even though it may not meet the protection standard of the Policy). This is in part what a network is designed to do, maintain and restore areas through adequate larval recruitment, working alongside other management tools.

9 See paragraph 69, page 16 of the MPA Policy.

Figure 6: In this case habitat A is a sink where it relies on external recruitment (for a particular species), and B is the source of recruitment.

RECOVERY DYNAMICS

The ability for an area that has been disturbed to recover, either by natural or human derived means, is dependent on the availability of recruitment source. The length of time for an area to recover is related to the connectivity to its regional species pool[25]. As illustrated in Figure 5, recovery generally progresses through multiple stages, with each stage dependant on access to recruitment (e.g. initial recovery by intermediate / opportunistic species, followed by recolonisation of the original habitat species). If habitats that provide the recruitment source (both for the intermediate and final stages of recovery) don’t exist or are too far away, recovery will take a long time, if it happens at all. If the disturbance regime is greater than the rate of recovery then a permanent shift in community structure and function may occur.

GUIDANCE ON CONNECTIVITY

Ideally, distances between habitats for the purposes of 'connectivity' would be determined based on information on species dispersal, both biological (e.g. time in the water column, habitat requirements), physical (e.g. ocean currents) and on knowledge of key source and sink populations.10 Although it is known that connectivity is an important factor in MPA network design, the information to assess this is very limited. So, rules of thumb are generally used.

Ideally a reserve would be large enough that it retains a proportion of the larval recruits (intra-reserve recruitment) and within reach of another reserve (inter-reserve recruitment). That is, MPAs within a network should be close enough together that sufficient larvae can disperse from one to the next.

However, the problem when designing MPA networks is the substantial variability across species in terms of their potential for larval transport. Figure 7 demonstrates the variability in the number of days that larvae are in the water

10 Sink populations are those that rely on immigration to survive and are not self-sustaining.

Key habitat provisioning species

column[26]. Note that the actual distance the larvae will move are dependent on multiple factors including currents and larval behaviour, but generally it was proposed that algae would disperse from 1m to 5km, invertebrates 10m to 1000km, and fish 1km to 1000km. It was also noted that key habitat-forming species tended to have lower dispersal potential, which is a key consideration for MPA design when the objective may be to maintain biogenic habitats (for example bryozoan beds).

Figure 7: Number of days larvae are in the water column for different functional groups. Key habitat forming species usually have low dispersal potential[26]

Another study of the dispersal potential of paua on the east coast of the North Island modelled the dispersal of paua from an existing marine reserve site[27]. Paua larvae reached settlement maturity 4–up to 80km from the marine reserve depending on local climate and currents, with a ‘typical’ larval transport of 10–30km from the marine reserve based on a 9 day planktonic phase before being able to settle. It was proposed that spacings of 10–30km between undisturbed adult populations would be optimal for maintaining larval supply for species like paua.

If the reserves within a network of MPA are too small to promote self-recruitment, and spaced too far apart to promote inter-reserve recruitment, then the ability for the network to be self-sustaining and meet network objectives are likely to be compromised[28]. What works for one species with a particular

dispersal regime, may not work for species that have shorter or longer dispersal distance. Therefore, no single size and spacing guidance is likely to benefit all species. As such, recommendations are usually given as a range of distances between reserves, to incorporate as much variability as possible (and hence benefit to as many species as possible).

CASE STUDIES

HOW ECOLOGICAL CONCEPTS HAVE BEEN APPLIED INTERNATIONALLY

There have been numerous conventions, agreements and best practice documents in relation to ecological guidelines on Marine Protected Area design[9], [15], [17], [29]–[42]. The New Zealand MPA Policy guidelines stem from New Zealand’s international commitment to the Convention on Biological Diversity (CDB), in particular Aichi target 1111. The table below provides international examples of processes aimed at creating MPA networks based on multiple principles that are generally consistent with the NZ MPA Policy, namely representativeness, Replication and Connectivity. The guidance that these processes developed were based on scientific concepts, research, and best practice, as well as social and political factors relating to their jurisdiction. As such, the examples should not be applied directly to the New Zealand context but are included to demonstrate how ecological concepts can be applied to the real world.

11 Aichi Target 11 Quick Guide https://www.cbd.int/doc/strategic-plan/targets/T11-quick-guide-en.pdf

EXAMPLES

A ‘Y’ indicates the principle is considered but no specific measure is provided.

MPA Size Representation Replication MPA Connectivity Additional Information

Californian Marine Life Protection Act (MLPA)12

5-10km Coastline length, preferably 10-20km

All key habitats should be protected (13 identified)

3-5 replicates per biogeographic region (4 bioregions)

Separated by 50-100km

Stakeholder process supported by Science Advisory Team, implemented as 4 regions 2004-201213

Marine Conservation Zones project (MCZ)14

5km minimum dimension, average of 10-20km

Include all broad-scale habitats (23 identified); Include examples of rare and threatened habitats

2 replicates for broad-scale habitats; 3-5 for key features (4 regions)

Similar habitat separated by no more than 40-80km

Stakeholder process implemented as 4 regions 2009-2015; Series of reports on design guidance commissioned and provided

Australian Great Barrier Reef Marine Park15

>10km coastal; >20km offshore

Represent 70 bioregions; 20% of each habitat type

3-4 70-100km Operating principles developed by Independent Scientific Steering Committee

12 California Marine Life Protection Act Master Plan for Marine Protected Areas (guidance) [39]

13 California MLPA analysis of outcomes [45], [46]

14 Marine Conservation Zone Project [19]

15 http://www.gbrmpa.gov.au/__data/assets/pdf_file/0011/6212/tech_sheet_06.pdf

Channel Islands[43]

Must include reserves of various sizes; Dependant on goal of reserve

30-50% of each representative habitat (3 Biogeographic regions, 15 habitat types = 43)

1-4 reserves) in each biogeographic region

Via replication of species/habitats in several different sites

Specific advice on reserve design was by Science Advisory Panel

OSPAR North-East Atlantic[38]

Dependant on aim (including integrity and management)

Y Y Y Multi-national regional process

HELCOM (Baltic Marine Environment Protection Commission) MPAs16

Preferably 30km2 minimum

Y Multi-national regional process

South Australia[41]

Represent samples of each major habitat type in each of 8 bioregions; Represent biogeographic significant areas

Y Scientific Advisory Group recommended representative areas

MPA Networks in Canada[44]

Average of 10-20km minimum dimension

Every broad-scale habitat must be represented; 30% of each habitat area should be in no take

At least 2 replicates for broad-scale habitats; 3-5 for rare or special sites

Generally should be within 20-200km from nearest MPA in network

Scientific panel guidelines produced, process not yet implemented

16 http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:1992:206:0007:0050:EN:PDF

reserve

REFERENCES[1] Government, “Marine Protected Areas Policy and Implementation Plan,” 2005.[2] D. L. Kramer and M. R. Chapman, “Implications of fish home range size and

relocation for marine reserve function,” Environ. Biol. Fishes, vol. 55, pp. 65–79, 1999.

[3] N. Shears and R. Babcock, “Marine reserves demonstrate top-down control of community structure on temperate reefs,” Oecologia, vol. 132, no. 1, pp. 131–142, Jun. 2002.

[4] R. C. Babcock, N. T. Shears, a C. Alcala, N. S. Barrett, G. J. Edgar, K. D. Lafferty, T. R. McClanahan, and G. R. Russ, “Decadal trends in marine reserves reveal differential rates of change in direct and indirect effects.,” Proc. Natl. Acad. Sci. U. S. A., vol. 107, no. 43, pp. 18256–61, Oct. 2010.

[5] N. T. Shears and R. C. Babcock, “Indirect effects of marine reserve protection on New Zealand ’ s rocky coastal marine communities,” 2004.

[6] B. S. Halpern and R. R. Warner, “Matching marine reserve design to reserve objectives.,” Proc. Biol. Sci., vol. 270, no. 1527, pp. 1871–8, Sep. 2003.

[7] J. B. Kellner, I. Tetreault, S. D. Gaines, and R. M. Nisbet, “Fishing the line near marine reserves in single and multispecies fisheries.,” Ecol. Appl., vol. 17, no. 4, pp. 1039–54, Jun. 2007.

[8] A. Grüss, D. M. Kaplan, S. Guénette, C. M. Roberts, and L. W. Botsford, “Consequences of adult and juvenile movement for marine protected areas,” Biol. Conserv., vol. 144, pp. 692–702, 2011.

[9] C. M. Roberts, S. Andelman, G. Branch, B. S. Halpern, and H. P. Possingham, “Ecological criteria for evaluating candidate sites for marine reserves,” vol. 13, no. 1, pp. 199–214, 2003.

[10] D. Abecasis, B. Horta, P. Afonso, E. J. Gonçalves, and K. Erzini, “Early reserve effects linked to small home ranges of a commercial fish , Diplodus sargus , Sparidae,” Mar. Ecol. Prog. Ser., vol. 518, pp. 255–266, 2015.

[11] B. S. Halpern and R. R. Warner, “Matching marine reserve design to reserve objectives.,” Proc. Biol. Sci., vol. 270, no. 1527, pp. 1871–8, Sep. 2003.

[12] S. D. Gaines, C. White, M. H. Carr, and S. R. Palumbi, “Designing marine reserve networks for both conservation and fisheries management.,” Proc. Natl. Acad. Sci. U. S. A., vol. 107, no. 43, pp. 18286–93, Oct. 2010.

[13] G. J. Edgar, N. S. Barrett, and R. D. Stuart-smith, “Exploited reefs protected from fishing transform over decades into conservation features otherwise absent from seascapes,” Ecol. Appl., vol. 19, no. 8, pp. 1967–1974, 2009.

[14] UNEP-WCMC, “National and Regional Networks of MPA: A review of progress,” 2008.

[15] C. Rondinini, “Meeting the MPA network design principles of representivity and adequacy: Developing species-area curves for habitats,” 2011.

[16] C. Rondinini, “A review of methodologies that could be used to formulate ecologically meaningful targets for marine habitat coverage within the UK MPA network,” JNCC Rep., no. 438, 2011.

[17] C. Rondinini and F. Chiozza, “Quantitative methods for defining percentage area targets for habitat types in conservation planning,” Biol. Conserv., vol. 143, no. 7, pp. 1646–1653, Jul. 2010.

[18] Ministry of Fisheries and Department of Conservation, Marine Protected Areas Policy and Implementation Plan. 2005, pp. 1–25.

[19] Natural England and Joint Nature Conservation Committee, “Marine Conservation Zone Project: Ecological Network Guidance,” 2010.

[20] H. Thomas and N. Shears, “Marine Protected Area Networks : Process design and ecosystem- based approaches,” 2013.

[21] C. Lundquist, S. Thrush, G. Coco, and J. Hewitt, “Interactions between disturbance and dispersal reduce persistence thresholds in a benthic community,” Mar. Ecol. Prog. Ser., vol. 413, pp. 217–228, Aug. 2010.

[22] D. L. Dixson, G. P. Jones, P. L. Munday, S. Planes, M. S. Pratchett, M. Srinivasan, C. Syms, and S. R. Thorrold, “Coral reef fish smell leaves to find island homes,” Proc. R. Soc. B Biol. Sci., vol. 275, no. 1653, pp. 2831–2839, 2008.

[23] I. C. Stobutzki and D. R. Bellwood, “Sustained swimming abilities of the late pelagic stages of coral reef fishes Ilona,” Mar. Ecol. Prog. Ser., vol. 149, pp. 35–41, 1997.

[24] S. Planes, “Connectivity - Spacing a network of marine protected areas based on connectivity data,” in Marine Protected Arteas: A multidisiiplinary approach, J. Claudet, Ed. Cambridge University Press, 2011, pp. 322–333.

[25] S. F. Thrush, J. E. Hewitt, A. M. Lohrer, and L. D. Chiaroni, “When small changes matter: the role of cross-scale interactions between habitat and ecological connectivity in recovery,” Ecol. Appl., vol. 23, no. 1, pp. 226–238, 2013.

[26] T. J. Anderson and C. J. Lundquist, “Pelagic larval dispersal: a review of New Zealand marine species.,” Unpubl. Manuscr.

[27] S. a. Stephens, N. Broekhuizen, a. B. Macdiarmid, C. J. Lundquist, L. McLeod, and R. Haskew, “Modelling transport of larval New Zealand abalone ( Haliotis iris ) along an open coast,” Mar. Freshw. Res., vol. 57, no. 5, p. 519, 2006.

[28] B. J. Puckett, D. B. Eggleston, P. C. Kerr, and R. a. Luettich, “Larval dispersal and population connectivity among a network of marine reserves,” Fish. Oceanogr., vol. 23, no. 4, pp. 342–361, 2014.

[29] S. S. & S. W. Day J., Dudley N., Hockings M., Holmes G., Laffoley D., Guidelines for Applying the IUCN Protected Area Management Categories to Marine Protected Areas, no. 19. 2012.

[30] IUCN, “IUCN2008_Laffoley_ Establishing MPA networks.pdf.” 2008.[31] IUCN World Commission on Protected Areas (IUCN-WCPA), Establishing Resilient

Marine Protected Area Networks — Making It Happen. 2008.[32] N. Dudley, “Guidelines for Applying Protected Area Management Categories,”

2008.[33] G. Kelleher, “Guidelines for Marine Protected Areas,” 1999.[34] Government, “The NZ Biodiversity Strategy,” no. February, 2000.[35] the U. of Q. The Ecology Centre, “Scientific Principles for Design of Marine

Protected Areas in Australia : A Guidance Statement,” 2009.[36] Secretariat of the Convention on Biological Diversity, Technical advice on the

establishment and management of a national system of marine and coastal protected areas, vol. 13. 2004.

[37] C. M. Roberts, J. P. Hawkins, J. Fletcher, S. Hands, K. Raab, and S. Ward, “Guidance on the size and spacing of Marine Protected Areas in England,” 2010.

[38] OSPAR Commission, “GUIDANCE ON DEVELOPING AN ECOLOGICALLY COHERENT NETWORK OF OSPAR MARINE PROTECTED AREAS,” 2006.

[39] California department of Fish and Game, “CALIFORNIA MARINE LIFE PROTECTION ACT MASTER PLAN for Marine Protected Areas,” 2008.

[40] J. L. Baker, “Guide to Marine Protected Areas,” 2000.[41] J. L. Baker, “Towards a system of ecologically representative marine protected

areas in South Australian marine bioregions,” 2004.[42] P. Desmet and R. Cowling, “Using the Species – Area Relationship to Set Baseline

Targets for Conservation,” Ecol. Soc., vol. 9, no. 2, 2004.

[43] S. Airame, J. E. Dugan, K. D. Lafferty, H. M. Leslie, D. McArdle, and R. Warner, “APPLYING ECOLOGICAL CRITERIA TO MARINE RESERVE DESIGN : A CASE STUDY FROM THE CALIFORNIA CHANNEL ISLANDS,” Ecol. Apllications, vol. 13, no. 1, pp. 170–184, 2003.

[44] S. Jessen, K. Chan, I. M. Côté, P. Dearden, E. De Santo, M. Fortin, F. Guichard, W. Haider, G. Jamieson, D. Kramer, A. McCrea-Strub, M. Mulrennan, W. a Montevecchi, J. Roff, A. Salomon, J. Gardner, L. Honka, R. Menafra, and A. Woodley, Science-based Guidelines for Marine Protected Areas and MPA Networks in Canada. 2011.

[45] E. T. Saarman and M. H. Carr, “The California Marine Life Protection Act: A balance of top down and bottom up governance in MPA planning,” Mar. Policy, vol. 41, pp. 41–49, 2013.

[46] E. Saarman, M. Gleason, J. Ugoretz, S. Airamé, M. Carr, E. Fox, A. Frimodig, T. Mason, and J. Vasques, “The role of science in supporting marine protected area network planning and design in California,” Ocean Coast. Manag., vol. 74, pp. 45–56, 2013.