The road towards effective governance and management of marine protected areas in South Africa: evolving policies, paradigms and processes

The efficacy of marine protected areas (MPAs) depends on their governance and management. We review their history in South Africa and recognise four periods. Period 1 (1964–1994) provided initial protection but was based on exclusionary, preservationist policies, was ad hoc in the absence of a national plan, and neglected social considerations. Period 2 (1994–2010) began introducing people-oriented policies, focused on ecosystems rather than species, and was strengthened by the formation of a national coordinating body. Period 3 (2010–2019) heralded improvements in design, ecosystem representation and stakeholder engagement, yet fractured governance hindered coordination and management. Period 4 (Since 2019) added challenges in managing new offshore MPAs. Progress in achieving effective MPA governance and management was assessed over these periods for 17 components of governance and management, representing key issues for which changes could be identified throughout. Fifteen components indicated overall improvements—most notably legislation and policies, MPA establishment, planning and design, and staff training and skills—whereas progress for most of the other components was weaker. Zero net gains were recorded for enforcement and compliance, and for the staff complement. Our recommendations flow principally from components assessed as faring poorly. We conclude with eight critical needs: (1) specify detailed objectives for every MPA; (2) fast-track management plans for new MPAs; (3) improve law enforcement and compliance; (4) enhance participation of adjacent communities and other stakeholders; (5) address MPA-related social impacts and injustices, and improve benefit sharing; (6) ensure financial sustainability; (7) strengthen evaluations of management effectiveness; and (8) improve cooperation between government authorities responsible for MPAs and fisheries.

these objectives by limiting, controlling or disallowing human activity in particular areas (Kelleher 1999). Over the past decade, maritime countries worldwide have been increasing MPA coverage in their territorial waters in line with targets set by the Convention on Biological Diversity (CBD n.d., 2010) and the United Nations Sustainable Development Goals (UN General Assembly 2015). For example, in South Africa, where MPAs have been used as tools for coastal and marine area-based management since 1964 when the Tsitsikamma MPA was declared (Lombard et al. 2020), successive MPAs have been proclaimed and, in 2019, 20 new or expanded areas were declared under marine protection (RSA 2019). This increased the number of MPAs in the waters of the exclusive economic zone (EEZ) for mainland South Africa (excluding the Prince Edward Islands) from 25 to 41 (Appendix 1), and the percentage coverage under MPAs from <0.5% to 5.4% (Sink et al. 2019).
In the process of attempting to meet ambitious targets for protection, nations and management authorities risk losing sight of the objectives of MPAs and their becoming mere 'paper parks' (Edgar et al. 2014;Wilhelm et al. 2014;Schéré et al. 2021). It is well known that for protected areas to be effective, appropriate governance systems and management processes must be in place (e.g. Pomeroy et al. 2004;Lockwood 2010;Borrini-Feyerabend et al. 2013; Bennett and Dearden 2014;de Koning et al. 2016;Bennett et al. 2019; Jones and Long 2021;Fidler et al. 2022). Governance relates to the interaction of policies, institutions and processes that determine who participates in decisions, how decisions are made, and who is responsible for implementation (Lockwood 2010;Bennett and Satterfield 2018), and thus includes both institutional and legislative frameworks within which management is performed. Management relates to the resources, plans and actions that result from applying governance systems (Lockwood 2010; Bennett and Satterfield 2018), including planning, management implementation, law enforcement, staff training and MPA staff relationships with local communities and the private sector. Governance and management indicators are frequently considered in the evaluation of protected area effectiveness, along with ecological (or biophysical) and socio-economic indicators (Pomeroy et al. 2004;Arkema et al. 2006).
Recent works by Kirkman et al. (2021) and Mann-Lang et al. (2021) have considered the effectiveness of South Africa's MPAs from ecological and socio-economic perspectives, respectively. This article presents the third component of the evaluation, which focuses on governance and management effectiveness, thus completing the 'BEG' trilogy (biological-ecological, economic-social, and governance-management) of assessments required to track the overall progress of MPAs (Thorpe et al. 2011b). There have been noteworthy shifts in the legislative environment affecting the governance and management of MPAs in South Africa since the declaration of its first MPA (and also in perceptions regarding their role), corresponding with momentous changes in the country's political history. Progress has also been made in evaluating the governance and management effectiveness of South Africa's MPAs, including the introduction of the Management Effectiveness Tracking Tool (METT) in 2009 (Tunley 2009) and its further development in subsequent iterations (Chadwick et al. 2014; Adams and Kowalski 2021). Here, we review various sources of information relating to past and present governance and management of MPAs in South Africa, and assess progress in achieving the objectives of MPAs through effective governance and management. Finally, we also identify gaps and recommendations for improvement.

Materials and methods
To track the history of MPA governance and management in South Africa, we consulted experts, books, scientific papers, published and unpublished reports, government legislation and national guidelines, MPA management evaluations, and other sources. Details of the relevant legislation and institutional developments for MPAs that were consulted, as well as ancillary legislation and guidelines that indirectly guided MPA establishment, are provided in Appendix 2. Key reports and publications that were considered include: the report of the Marine Reserve Committee established in 1976 by the Minister of Economic Affairs to advise the Minister of Sea Fisheries on the establishment of marine reserves through the Fishery Advisory Board (Komitee Insake Seereservate 1977); the report of the Marine Reserve Task Group (MRTG) commissioned in 1996 to review the administration, management, design and representativeness of South Africa's MPAs (MRTG 1997); several reports and papers on the state of MPAs or MPA management in South Africa, such as Hockey and Buxton (1989), Attwood et al. (1997), Lemm and Attwood (2003), and the three METT reports completed to date for South Africa's MPAs (referred to as METT-SA reports: METT-SA1 [Tunley 2009]; METT-SA2 [Chadwick et al. 2014]; METT-SA3 [ Adams and Kowalski 2021]). A list of abbreviations and acronyms appears in Appendix 3.
Based on the literature, we identified broad temporal periods between which there were shifts in key governance and management inputs. This timeline was used to systematically evaluate trends and changing paradigms in the governance and management of MPAs through the comparison of categories-representing various key components of governance and management for which changes could be readily identified throughout-across and between the different periods. The selection of components was guided by categories used in the state of management effectiveness evaluation reports ( Lemm and Attwood 2003; METT-SA1-3), issues that were identified in various South African MPA reports and reviews (e.g. Komitee Insake Seereservate 1977; Attwood et al. 1997;MRTG 1997) or in the analysis of MPA governance by the United Nations Environment Programme McCay and Jones 2011), and elements of the CBD's Programme of Work on Protected Areas (PoWPA) (CBD n.d.).
Progress (or lack thereof) for each component between the periods was scored by the authors of this study (who include a mix of scientists, managers and policy experts with a wealth of experience in MPA research, monitoring, management and governance in South Africa) and informed by the reviewed literature. Scores of 2, −2, and 0 represent improvement, decline, and no change, respectively. Where clear consensus could not be reached between the authors on progress between periods (always between 2 and 0 or between 0 and −2), intermediate scores of 1 or −1 were given. For applicable components, scoring that was conducted within completed METT assessments was considered. However, since METT scores are largely products of self-assessments by management authorities and thus have the potential for bias, these were not always followed precisely-especially if the authors agreed that information from the METT did not provide an accurate reflection.
Finally, based on our assessment of progress and gaps, future priorities were identified for each of the assessed components of governance and management.

Time periods and key governance and management components
Four broad temporal periods were identified (Figure 1), namely 1964-1994 (Period 1), 1994-2010 (Period 2), 2010-2019 (Period 3) and 2019 to the present (Period 4). It was not the intent to assign strict divisions between the periods, since there were transitions between the periods corresponding either to shifts in South African society, or political and organisational shifts during which new institutional arrangements, structures and procedures replaced previous regimes. The following transitions were adopted: between Periods 1 and 2, the onset of democracy in South Africa in 1994; between Periods 2 and 3, when MPA functions and fisheries functions were split between two government departments in 2010, antedating a split in legislation between these functions; between Periods 3 and 4, when the re-unification of MPA and fisheries functions under the same ministry occurred, and 20 new MPAs were declared in 2019 following the Operation Phakisa initiative (Sink et al. 2019).
The 17 key components of governance and management that were selected are shown in Figure 2, which summarises the scores for the changes between time periods for each of the 17 components of governance and management. The basis for Figure 2 is Supplementary Table S1, which summarises the review of literature for each component and uses directional arrows based on the scoring to present trends across the time periods, allowing progress in implementation to be contextualised over time. As such, Supplementary Table S1 is central to the results that are presented below under the subheadings of different time periods.

Period 1 (1964-1994)
Declarations of MPAs during the first period were largely ad hoc and were often declared adjacent to existing terrestrial coastal protected areas (Hockey and Buxton 1989), with a terrestrial perspective being applied to their management despite important differences between the two systems (Hockey and Branch 1994). At the institutional level, the responsibility of establishing and managing the MPAs that were declared during this period ( Figure 1) fell mostly on provincial or national conservation agencies (Appendix 1).   Table S1). These included inadequate legislation (Glavovic 2006), which, for example, did not control for non-fishing related activities in MPAs (Attwood et al. 1997), fragmented and inconsistent policies, and the absence of a national MPA coordinating body responsible for administrative and management functions (Hockey and Buxton 1989;Attwood et al. 1997;MRTG 1997). At the time, the focus of MPAs was on preservation, which disregarded zonation for multiple sustainable uses, and the objectives for establishing the MPAs were unspecific and both poorly articulated and publicised (Attwood et al. 1997;MRTG 1997).
Stakeholder engagement was generally limited during this period and typified by a lack of participatory processes that involved consultation with all users and adjacent communities ( Lemm and Attwood 2003;Sunde and Isaacs 2008). Any stakeholder engagement largely excluded people of colour given the racial segregation at the time. Indeed, during this period, protected areas in South Africa (including those along the coast) were often declared as part of the segregational conservation and land development policies of the government at the time (Carruthers 1989;Thondhlana et al. 2016;Ramutsindela and Shabangu 2018). Some of these policies led to the dispossession of land and relocation of marginalised groups which limited their access to coastal and marine environments, with reported impacts in several of South Africa's MPAs (Attwood et al. 1997; Kepe 2004; Sunde and Isaacs 2008;Sowman et al. 2011;Sunde 2013).
Although the management of MPAs was top-down by the state with no co-management systems in place (Sunde 2014), there was a trend towards increasing community involvement in some MPAs, such as iSimangaliso (Mann 1995;Kyle et al. 1997). Other major shortcomings in MPA management were also identified: lack of management plans that were required in terms of the Sea Fishery Act (RSA 1988); jurisdiction and compliance issues related to the enforcement of MPA regulations (Hockey and Buxton 1989;Attwood et al. 1997;MRTG 1997); failure to fulfil MPA awareness and educational functions that contributed to inadequate public awareness about MPAs ( Hockey and Buxton 1989;Attwood et al. 1997;MRTG 1997); a shortage of funds (Attwood et al. 1997;MRTG 1997); insufficient numbers of marine-trained staff (Hockey and Buxton 1989;Attwood et al. 1997); inadequate equipment (e.g. vessels for patrolling) (Attwood et al. 1997); and inadequate monitoring programmes (Attwood et al. 1997;MRTG 1997).
A Marine Reserve Committee that was established in 1976 to investigate and recommend guidelines to advise on the management of marine reserves in terms of the Sea Fisheries Act (Komitee Insake Seereservate 1977) presented several criteria, guidelines and recommendations. These were, in the words of Attwood et al. (1997, p 343), "sound advice even by today's reasoning", yet the Committee was "unfortunately disbanded after submitting the report" and "its recommendations were not widely implemented." Their recommendations included the following: to publicise MPAs and their roles; to consult with, inform and educate the public; to demarcate MPAs properly; to enforce regulations  and counter illegal activities. Therefore, some of the shortcomings discussed above may have been averted had the recommendations been implemented.

Period 2 (1994-2010)
The second time period was heralded by the establishment of South Africa's new democracy, with associated changes in policy, legislation and government institutions (Figure 1). A key development was the Constitution of the Republic of South Africa (RSA 1996) (Appendix 2) articulating the Bill of Rights, including the right to an environment that is not harmful to health or wellbeing, and to an environment that is protected for the benefit of present and future generations. Simultaneously, the Constitution makes provisions for sustainable development and the use of natural resources while promoting justifiable economic and social development, equality, human dignity, sufficient food, and the right to practice culture and religion. This change in focus from simple preservation of the environment to a people-centred approach that prioritises human dimensions prompted a re-evaluation of the concept of MPAs in South Africa over the long term.
Calls for an overarching legislated policy to guide the governance and management of MPAs were partly taken up in the development of new fisheries legislation, namely the Marine Living Resources Act (MLRA) (Appendix 2). Among other objectives, the MLRA addressed some of the past fishery-related racial injustices ( Cochrane and Payne 1998). While the MLRA focused predominantly on the management and transformation of commercially valuable fisheries, Section 43 of the Act made specific reference to the establishment and management of MPAs in South Africa, specifying that MPAs could be declared to address one or more of three functions: (i) fisheries management; (ii) the protection of flora and fauna and their associated habitats; and (iii) the reduction of competition and/or conflict between users.
At the institutional level, in 1998, the former Department of Sea Fisheries was incorporated into Marine and Coastal Management (MCM), the country's first national coordinating body for MPAs, under the Department of Environmental Affairs and Tourism (DEAT). MCM would be largely responsible for the establishment of MPAs in South Africa, and a Marine Reserves Task Group was established to develop a policy on MPAs and review progress and challenges (Attwood et al. 1997;Beaumont 1997;MRTG 1997). In 2000, 19 of South Africa's existing MPAs were re-proclaimed in terms of the MLRA (RSA 2000), and several new MPAs were proclaimed from 2000 to 2008 ( Figure 1; Appendix 1). Many of these MPAs filled important gaps in marine biodiversity representation and protection that were identified for the South African coast (Emanuel et al. 1992;Turpie et al. 2000;Awad et al. 2002). The development of the COMPARE system (Hockey and Branch 1997) provided a more robust, evidence-based approach to MPA planning and design, and systematic conservation planning for MPA expansion came into effect during this period with the first National Spatial Biodiversity Assessment (NSBA) (Lombard et al. 2004). The NSBA systematically assessed the representation of MPAs and found that offshore ecosystems in particular were poorly protected (Lombard et al. 2004). This in turn led to the initiation of the Offshore Marine Protected Areas project (Sink et al. 2011), which systematically prioritised offshore areas for protection, and influenced the development of MPAs during the next period (2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018)(2019). The NSBA also informed the marine component of South Africa's first National Protected Area Expansion Strategy in (DEA 2010, setting medium-term targets for additional MPAs. South Africa's environmental policy landscape shifted during this period with a focus on redressing past injustices, as reflected in the emergence of people-centred language in several environmental and conservation policies, all seeking to balance environmental and socio-economic rights (Appendix 2). Furthermore, the establishment of the People and Parks Programme in 2004 signalled recognition of the importance of public participation in protected area management (DFFE n.d.) (Appendix 2). However, despite the establishment of stakeholder forums or committees for some MPAs and the development of a protocol for co-management-initially for subsistence fishers, but applicable to MPAs (Harris et al. 2002)-communities living adjacent to MPAs and dependent on MPA resources were generally under-represented (Tunley 2009). As a result, the devolution of decision-making powers needed for effective co-management was lacking, as was the actual implementation of co-management arrangements (Sunde and Isaacs 2008;Tunley 2009). This underlines the issue that while the process of establishing MPAs during this time period had greatly improved in terms of the MLRA, and some communities gained greater access rights to MPAs (Harris et al. 2003;Kyle 2003;Mann 2003), inclusivity was largely lacking and stakeholder engagement processes remained generally inadequate throughout most of this period (Beaumont 1997;Sunde 2014). However, some improvements were brought about by the initiation of the Offshore MPA project in 2006, which included some of the first engagements with offshore industry stakeholders regarding the potential establishment of offshore MPAs, and the establishment of the national MPA Forum in 2007, which brought together key stakeholders and decisionmakers on issues of marine protection ( Sink and Attwood 2008;Sink et al. 2011).
Two major assessments of the state of management of South Africa's MPAs occurred during this time period, which were more exhaustive than previous assessments since they involved site visits and interviews with those directly involved in MPA management. Lemm and Attwood (2003) indicated that both the then-new national legislation (i.e. the MLRA, which governed both fisheries and MPAs) and the assignment of a responsible national coordinating body (i.e. MCM at that stage) had brought about considerable improvement in MPA management. However, those authors reported that most of the management weaknesses identified in previous assessments (e.g. Attwood et al. 1997) persisted. The subsequent evaluation by Tunley (2009) was METT-SA1, in which the following notable progress was reported: the coordination and formalisation of MPA management through the signing of management agreements to delegate responsibility from MCM to provincial or local conservation authorities; activities such as stakeholder engagement, monitoring, and education being recognised as important MPA management activities; the drafting of management plans for MPAs, albeit on an ad hoc basis; the development of formalised national and regional MPA training courses; and the allocation of budgets and acquisition of equipment for MPA management. The following specific weaknesses were identified: continued inadequate stakeholder engagement; lack of specific legislated objectives for MPAs; and lack of comprehensive monitoring programmes of MPA attributes to facilitate adaptive management. Although substantial improvements in enforcement procedures were reported in several MPAs (attributable to the development of management plans, improved staff training and acquisition of equipment), enforcement remained hampered by staff shortages, lack of skilled staff with marine experience, skipper availability and other related factors.
No existing legal framework provides for specialised environmental courts in South Africa ( Kotzé and du Plessis 2011). Despite this, two specialised criminal environmental (or 'green') courts were developed and operated successfully in Hermanus and Port Elizabeth (now named Gqeberha) from to 2006to 2009, respectively (Pring and Pring 2016. These courts largely dealt with fisheries cases and achieved prosecution rates of over 80%-compared with about 10% prior to their inception (Craigie et al. 2009;Branch and Branch 2018). Since their dissolution "due to the unwillingness of the Department of Justice and Constitutional Development to continue to provide extra personnel and facilities for a specialised court that was not mandated by specific legislation" (Craigie et al. 2009, p 98), MPA managers have pleaded for such environmental courts to be re-established (Tunley 2009).

Period 3 (2010-2019)
Whereas the MLRA had been administered under a single national department (DEAT), in 2010, the coordinating body (MCM) was divided into the Branch: Oceans and Coasts (O&C) within the Department of Environmental Affairs (DEA) and the Branch: Fisheries Management within the Department of Agriculture, Forestry and Fisheries (DAFF) (Appendix 2). Under this division, the responsibility for managing MPAs fell to the Branch: O&C, while all fisheries management responsibilities were allocated to the DAFF, thus dividing functions and responsibilities between two separate government departments and hampering communications and mutual agreement. To improve the alignment of MPAs and terrestrial protected areas within the DEA and to eliminate management-related confusion between the Branch: O&C and the Branch: Biodiversity and Conservation (responsible for terrestrial protected areas), all South African MPAs were transferred in 2014 from the MLRA to the National Environmental Management: Protected Areas Act (NEM: PAA) (RSA 2003) through an amendment of the latter (RSA 2014a) (Appendix 2). Section 14 of the amendment refers to the 'continued existence of MPAs', in that all MPAs that were declared under the MLRA must be considered as MPAs declared in terms of the NEM: PAA. The latter Act sought to provide coherence in governance across all South African protected areas, both terrestrial and marine. Of the different functions for which an MPA may be declared according to this Act, most are related to protection or are conservation-oriented, with only one function making provision for the sustainable use of natural and biological resources (Appendix 2). Remaining functions that address socio-economic considerations are limited in their extent and only broadly framed to contribute to human, social, cultural, spiritual and economic development (Mann-Lang et al. 2021).
Notably, Section 39(2) of the NEM: PAA requires that each protected area has a management plan to guide the development and management of the area to meet the purpose for which it was declared. Whereas MPA management plans were historically developed on a relatively ad hoc basis because of a lack of clear guidance from the MLRA, all MPA management plans now had to align with a management planning framework provided by the NEM: PAA. During this period, national policies also sought to mainstream small-scale fisheries into Period 3 also saw the increasing use of systematic conservation planning (SCP) in MPA planning (Figure 1), enabled through advancements in technology and data assimilation (Kirkman et al. 2019). The ability to conduct SCP exercises facilitated the identification of biodiversity priority areas by considering principles such as ecological representivity, persistence and connectivity, while minimising the opportunity cost to resource-use sectors (Margules and Pressey 2000). This data-driven method allowed for a more systematic approach to MPA identification, with a simultaneous shift towards planning for MPAs in deeper waters (Sink et al. 2011;Majiedt et al. 2013). The SCP products produced during this period (e.g. Sink et al. 2011;Chalmers 2012;Harris et al. 2012;Majiedt et al. 2013) were accompanied by some improvement in stakeholder engagement with coastal resource users and offshore industries (Chadwick et al. 2014). These efforts became focused through the Operation Phakisa initiative (Sink 2016), which aimed for representative protection of 5% of South Africa's marine domain (excluding the marine area of the Prince Edward Islands MPA). This represented a 10-fold increase in the existing mainland MPA coverage. The efforts of scientists and managers involved in this initiative culminated in the declaration of 20 new or expanded MPAs under the NEM: PAA and the promulgation of their associated regulations (RSA 2019).
During Period 3, METT-SA2 was completed (Chadwick et al. 2014), reporting that many of the challenges identified during previous evaluations persisted, thus indicating the need to better understand these challenges and the barriers to addressing them effectively. Insufficient budgets were reported as one of the key limiting factors, underpinning the fact that management authorities were unable to secure skilled staff and adequate equipment to effectively implement MPA management plans. Another major finding was that MPA management effectiveness had been severely impacted by the 2010 division of MCM into the DAFF and DEA. In particular, compliance suffered from reduced support by Fishery Control Officers (FCOs) in enforcing the provisions of the MLRA in MPAs (Kramer et al. 2017). This was especially challenging for MPAs where poaching and non-compliance were problems, because reduced support by FCOs resulted in management staff having to focus on enforcement at the expense of other activities, such as education and monitoring. Nevertheless, while METT-SA2 reported that education and public-awareness programmes had not improved at most MPAs, there were improvements in ecological monitoring, with many MPAs having initiated relevant monitoring or at least prioritised monitoring and research needs.
Guidelines for integrating human dimensions into MPA planning and management were produced (Sowman et al. 2014a), and stakeholder engagement was reported to have improved across most MPAs with the effective use of the national MPA Forum, MPA stakeholder committees and institutional coastal committees created in terms of the Integrated Coastal Management Act (RSA 2008), yet ongoing conflicts between local communities and MPA authorities remained a major barrier to effective management at several MPAs (Chadwick et al. 2014). It was also reported that very few people living adjacent to MPAs received tangible financial benefits from them, and where there were benefits, they rarely exceeded foregone opportunity costs (Cundill et al. 2013). Pre-existing communal property rights in the coastal environment also remained unrecognised, with the impacts of dispossession during the apartheid era persisting (Sunde 2014;IWPA 2020). However, in a landmark case with considerable significance for the governance and management of coastal MPAs, customary fishing rights in the Dwesa-Cwebe MPA were first recognised in the Supreme Court of Appeal ( Monyamane and Bapela 2019).

Period 4 (Since 2019)
The new or extended MPAs declared since 2019 support multiple objectives for biodiversity conservation in alignment with ocean economy goals (provided in detail in the respective declaration notices). In determining the objectives, Operation Phakisa's MPA technical team drew from lessons that emerged from work refining MPA objectives in a project led by DEA and the World Wide Fund for Nature -South Africa (WWF-SA), initiated in 2017. The objectives included several specific to particular species and ecosystem types, but the scopes of objectives were broadened to support the cultural, heritage and educational benefits of MPAs. Systematic planning was taken further during this period with the development of the National Coastal and Marine Systematic Biodiversity Plan (Harris et al. 2022a), which includes Critical Biodiversity Areas and sea-use guidelines to inform area-based conservation and biodiversity management in South Africa's ocean space. This Plan provides a basis for the Biodiversity Sector Plan, which remains under development at the time of writing and is intended to inform the drafting and negotiation of marine spatial planning (MSP) from the perspective of biodiversity priorities. A regional strategy for MSP has been developed for the western Indian Ocean, including South Africa (Nairobi Convention Secretariat et al. 2021), and a robust procedure for the input of biodiversity considerations into MSP has been formulated (Harris et al. 2022b).
The re-unification of MPA and fisheries functions into one government department also occurred in 2019(DFFE 2022, with the DAFF and DEA once again combined under the new Department of Forestry, Fisheries and the Environment (DFFE). However, despite this re-unification, alignment between different pieces of legislation and policy remains elusive (e.g. between the NEM: PAA, MLRA, and the Small-Scale Fisheries Policy [SSFP]). With MPAs and fisheries still managed under separate branches of DFFE (Branch: O&C, and Branch: Fisheries Management, respectively), greater coordination and cooperation are required, and this need likely contributed to various management challenges identified in previous METT assessments, which remained unresolved in the latest assessment, METT-SA3 ( Adams and Kowalski 2021).
Co-management agreements are currently in place at three MPAs (iSimangaliso, Dwesa-Cwebe and Pondoland), with only one MPA specifically mentioning the development of effective co-management as an MPA objective (Sixteen Mile Beach) (Fielding 2021). Despite some initial steps being made, co-management of MPAs in South Africa remains in its infancy.
Refinement and revision of indicators and scoring templates between the latest and previous METT-SA assessments limited the tracking of trends among assessments, with only a small subset of indicators being directly comparable between assessments. However, the latest assessment ( Adams and Kowalski 2021) was useful for informing management-oriented categories (Supplementary Table S1). This showed that the majority of MPAs continue to lack adequate funding, staffing and resources, effective law enforcement, adequate public awareness programmes, and extensive monitoring to inform adaptive MPA management. The consideration given to cultural heritage was found to be particularly inadequate. Both enforcement and operational budgets, which were two of the indicators that were directly comparable with previous METT-SA assessments, declined since the previous assessments. Management components that improved (with over 50% of MPAs showing optimal management) included improved community and stakeholder engagement and the development of management plans and annual plans of operation.

Discussion
The overview below synthesises changes in the implementation of the various components of governance and management across the time periods, and discusses progress in an international context where applicable. A natural step from evaluating progress was to identify future priorities for implementation, especially to address weaknesses and shortcomings. These priorities are discussed against each component listed in Supplementary  Table S1, and many are integrated here with the overview of progress, to help map the way forward. One of the components is discussed in a unique section, 'Management effectiveness evaluation', and the section 'Further components of governance and management' discusses important issues of governance and management that were outside the components considered here. The Conclusions section includes recommendations that flow principally from components faring poorly in the assessment.

Overview of progress in the implementation of governance and management components
The management component that showed the clearest positive progress was MPA establishment, planning and design (Figure 2), with the introduction and development of an SCP approach for prioritisation of areas, and the declaration of 20 offshore MPAs under Operation Phakisa. Reviewing MPA progress in the western Indian Ocean, Andriamahefazafy et al. (2022) rated South Africa as having achieved 'good progress' in establishing MPAs. Despite this, the network that has been created is still insufficient to ensure adequate connectivity among the MPAs, while gaps remain in terms of representivity, with unprotected or poorly protected ecosystems including threatened ecosystems and areas of biodiversity importance (Kirkman et al. 2021). Addressing these shortcomings should be prioritised in the future expansion of conservation areas in line with national targets (RSA 2018) or international commitments (e.g. CBD 2022).
In general, stakeholder engagement and participation concerning MPAs has shown modest but consistent improvement over time (Figure 2), with the initiation of MPA forums, committees, workshops and other participatory processes involving stakeholders and local communities. This corresponds with an increasing global awareness of the importance of involving local communities in the affairs of protected areas, as reflected in the CBD and other international forums (Sunde 2014). Despite this, conflicts between local communities and MPA authorities persist in several South African MPAs and there are continued complaints from communities about inadequate consultation (Muhl and Sowman 2020;DK and SD pers. obs.). Required improvements include ensuring that each MPA has its own forum or committee with full stakeholder representation. There is also a need to facilitate greater stakeholder engagement within governing institutions and to tailor engagements to meet stakeholder needs, which is an adaptive process. Importantly, for any new MPAs, stakeholder engagement should be initiated in the planning phase using a well-designed and coordinated process.
The co-management of MPAs is a component where progress has not lived up to expectations. Despite the development of a co-management protocol (Harris et al. 2002), recognition of MPAs in the People and Parks Programme, and research undertaken on ways to improve co-management (Napier et al. 2005;Francolini et al. 2023), examples of effective and lasting MPA co-management structures and implementation remain elusive. This may partly stem from a lack of coherence between legislation and policy (e.g. the NEM: PAA and SSFP) in defining co-management and determining methods for effective community-based natural resource management (Sunde 2014), but there are also capacity shortcomings (Chadwick et al. 2014). Thus, there is a need to define and develop a common understanding and a practical framework for co-management. In some MPAs, such as iSimangaliso, there has been increasing recognition of the need for community involvement in determining access rights to marine resources for adjacent local communities, and several projects have been initiated. Of particular relevance is a community project spearheaded by WWF-SA, which piloted a system of co-management for mussel harvesting. Important successes were: the establishment of a local co-management committee of fishers, managers and scientists; experiments jointly undertaken by fishers and scientists; the promotion of fishers' rights to harvest under controlled conditions; and monitoring by members of the community (Harris et al. 2003(Harris et al. , 2007. However, while the programme was highly successful and set guiding principles for its expansion to other communities (Napier et al. 2005), it lapsed when government funding was withdrawn.
Local communities, including small-scale fishing communities living in and around MPAs, have historically carried disproportionate costs and losses associated with the establishment of MPAs without enjoying their benefits ( Sunde and Isaacs 2008;Sowman et al. 2014b). This is apparent from the progress in rights allocation, access, equity and benefit-sharing within MPAs (Figure 2), which has largely fallen short of expectations. An example is the continued lack of recognition for communal property rights in the coastal environment pre-dating MPA declarations, with the associated impacts of dispossession persisting (Sowman and Sunde 2018). This came to a head when customary fishing rights in MPAs (specifically in the Dwesa-Cwebe MPA) were first recognised in the Supreme Court of Appeal ( Monyamane and Bapela 2019). The recognition of customary fishing rights has profound implications for both access rights and potentially for the effectiveness of MPAs in achieving ecological objectives that require the removal or control of human threats such as fishing. The need for a greater understanding of customary marine-tenure systems in South Africa is highlighted, as well as the need for these to be explicitly addressed in settlement agreements. While Sunde (2013) considers this to be especially relevant for the eastern seaboard of the country, where customary governance systems for resource access and use are prevalent, this issue likely holds national relevance. This is also in line with the SSFP in South Africa (RSA 2012), which prescribes that the relevant organs of state must resolve issues where tenure to coastal land involving communities affects the implementation of the Policy. In the global context, recent studies have demonstrated the importance of community involvement and equitable governance for protected-area effectiveness (e.g. de Koning et al. 2016;Fidler et al. 2022), but these are elements where progress in protected-area governance and management has been lagging, as revealed in an evaluation of the CBD's PoWPA (IUCN-TILCEPA 2010;Borrini-Feyerabend et al. 2013).
After improvements initially brought about by the MLRA and the creation of MCM, the efficacy of enforcement and compliance began to decline (Figure 2). This was linked to the division of fisheries and MPA functions in 2010 and related legislative changes, and was affected by other components showing little improvement, such as the operational budget, staff complement, and equipment and infrastructure of MPAs (Chadwick et al. 2014). Enhancing the capacity to enforce regulations in MPAs would involve, among other things, overcoming barriers brought about by different mandates for conservation and fisheriesfor example, MPA staff being unable to enforce fisheries regulations. In general, closer cooperation is needed between MPA management staff and FCOs. This needs to originate from higher levels of government to ensure improvements in communication and coordination across branches (Branches: Fisheries Management and O&C) within the DFFE. Learning from the negative effects that the division of functions at the political level had on MPA governance and management at the beginning of Period 3, a permanent, dedicated structure should ideally be created within the DFFE to support the coordinated management of MPAs.
A lack of adequate resources, including insufficient funding, staffing and operational equipment, is the norm for protected areas globally (e.g. Coad et al. 2019), and is also the case for South Africa's MPAs, where the operational budget and staff complement are additional areas of unsatisfactory progress ( Figure 2). Furthermore, the lack of suitably trained and skilled staff has consistently affected the management of MPAs in South Africa, including aspects such as enforcement and monitoring (Attwood et al. 1997;Chadwick et al. 2014). However, there has been recent good progress in MPA staff training, with the development of national and regional MPA management training courses (Tunley 2009;Chadwick et al. 2014), and certification programmes such as the Western Indian Ocean Certification of Marine Protected Area Professionals (WIO-COMPAS) (Squillante et al. 2010). The need for training to extend beyond short courses to provide peer-learning, broader experience and mentorship is highlighted, as well as the usefulness of local and regional networking and exchange programmes. Challenges regarding operational equipment and infrastructure and their maintenance are related to both inadequate or insecure funding and restrictive supply-chain management systems (Chadwick et al. 2014;Adams and Kowalski 2021). Government disbursement processes must be revised and improved to ensure greater flexibility and the timeous allocation of funds. There is also an urgent need to investigate alternative funding approaches, including partnerships with the private sector, nongovernmental organisations (NGOs) and international funders, to ensure adequate financing.
Educational and awareness programmes and activities are well advanced in some MPAs (e.g. West Coast National Park) but remain limited at many MPAs because of insufficient financial resources, facilities and suitably trained staff. This is linked to budget constraints causing staff to focus on other essential components, especially enforcement and compliance (Chadwick et al. 2014). Clearly, improving this component will depend on addressing these other components. Development of national coordinated awareness and educational programmes and resources across MPAs would assist those MPAs needing to advance programmes, although each MPA should be free to develop its own site-specific content.
Similar constraints have affected progress in the monitoring of ecological, socio-economic, cultural and other attributes in MPAs. While monitoring has improved, METT-SA3 reported that only one-third of the evaluated MPAs have an established monitoring programme that is fully implemented and influences MPA management, and most monitoring programmes are focused on ecological attributes, at the expense of socio-economic ones (Adams and Kowalski 2021). Appropriate monitoring is essential to determining outcomes against different MPA objectives (Kirkman et al. 2021), but suitable indicators of progress for assessments of MPA effectiveness are largely lacking, in common with most protected areas worldwide (UNEP- WCMC and IUCN 2021). The focus of such monitoring should be linked to the goals and objectives of MPAs and how these are taken up in management plans. In terms of integrating the ecological, social and economic objectives of MPAs, there has been progress in developing overarching goals, values and objectives of MPAs in legislation and policy (Mann-Lang et al. 2021). Compared with previous MPAs, the 20 new or extended MPAs declared in 2019 have broader scopes in their declaration notices, supporting multiple objectives for biodiversity protection that take ecological, social and economic considerations into account (RSA 2019). These include conservation objectives for particular species and ecosystems while also providing for economic, cultural, heritage and educational benefits. However, the specific objectives of many MPAs declared before 2019 are often unclear, partly because their objectives are often undocumented and cannot be ascertained (Mann-Lang et al. 2021), but also because dramatic changes in legislation have affected both how they should be managed, and the perceptions about their roles. Although all MPAs for which declaration notices could be traced include at least one social or economic objective, these are largely underrepresented when compared with ecological objectives (Mann-Lang et al. 2021). For those MPAs with unclear objectives, the social context of the MPA should be adequately considered, and relevant stakeholders need to be consulted.
Management plans are intended to guide the management of a protected area to meet the objectives for which it was declared and to benefit the area and those who rely upon or benefit from its good management. Consensus-based, MPA-specific objectives included in management plans can supplement legislated objectives for which MPAs were declared, including locally relevant social and economic objectives. However, just over 40% of MPAs evaluated by METT-SA3 (excluding the new MPAs declared in 2019) still lacked a management plan, or the management plan needed updating. At the time of writing, management plans are still being developed for the new or extended MPAs declared in 2019 (Appendix 1). Thus, there is a need to fast-track management plans in line with the NEM: PAA requirement that each protected area must produce a management plan for approval by the Minister or relevant member of the executive council within 12 months of the assignment of an MPA (RSA 2003). Where MPAs are incorporated within management plans for adjacent terrestrial reserves, adequate weight must be afforded to the marine areas, or an MPA-specific management plan needs to be developed to prevent the marginalisation of MPAs in terms of the resources and effort, as has previously been reported.
Successful conservation often depends on adaptive management, strong societal engagement, effective and equitable benefit-sharing, sustained funding, and the effective monitoring and enforcement of rules (Brondízio et al. 2019)-all of which are among the governance and management components considered here. Since the national government plays a central role, we highlight the need for a national framework to guide MPA planning and implementation for MPAs in South Africa to achieve their desired outcomes. The groundwork for such a policy was conducted during Period 2 but never came to fruition (MRTG 1997). Legislative changes have improved the alignment and integration of legislation and policy for marine and terrestrial protected areas under a single management statute, but the multifarious and unique issues affecting coastal and offshore MPAs in South Africa warrant specific policy guidance. Such a policy should address all the components assessed in this study, as all of them are relevant to the CBD's PoWPA. The policy should certainly include the following main elements: MPA planning and establishment, governance, law enforcement, participation, equity and benefit-sharing, enabling activities, and the monitoring and evaluation of management effectiveness.

Management effectiveness evaluation
The need to improve evaluations of management effectiveness has been highlighted as one of the key challenges for protected areas globally. By 2019, only approximately 14% of the global MPA area had undergone such evaluation (CBD 2020). Evaluations of MPA management effectiveness have shown progress in South Africa, from early assessments of the state of management that identified gaps and inadequacies ( Lemm and Attwood 2003), to the progressive adaptation of the METT for MPAs (Tunley 2009;Chadwick et al. 2014;Adams and Kowalski 2021).
The METT was originally developed by the World Wide Fund for Nature (WWF) and the World Bank in 2001 as a relatively simple and rapid questionnaire-based tool to assess protected area management effectiveness (Stolton et al. 2019). It covers the various elements of management but places an emphasis on the context, planning, inputs and processes of protected areas to provide guidance to protected area managers, allowing them to systematically identify needs, constraints and priority actions. Simultaneously, it provides a mechanism for monitoring progress. For example, it has been used by the CBD to report on worldwide progress towards the effectiveness of protected-area management (CBD 2020). The subjectivity of indicators and the potential for biased scoring-which are essentially due to self-scoring by management authorities themselves-are common criticisms of METT assessments (Sunde 2014;Coad et al. 2015;UNEP-WCMC and IUCN 2021). Hockings et al. (2018) also identified that a focus on the METT scores instead of on the steps required to improve effectiveness inhibits the tool from being used optimally. This is especially true if the scores are linked to the performance incentives of MPA managers and staff-a trend that has been observed in the South African context (RA pers. obs.). To address this, the METT-SA3 deliberately shifted away from presenting a numerical scoring system and focused on recommendations to improve management effectiveness using a 'traffic light' approach ( Adams and Kowalski 2021). The METT-SA3 is also more concise and user-friendly, has refined the indicators (Supplementary Table S2) and has developed a web-based tool for online completion. Such improvements will assist management authorities in their self-evaluations and in understanding the gaps and deficiencies within their MPAs. However, the inclusion of stakeholders in the evaluations (Sunde 2014)-especially for indicators addressing stakeholder and community communication, participation and benefits-would achieve more objective evaluations. An unfortunate consequence of the refinement and revision of indicators and scoring templates between successive METT-SA assessments is that it has made it difficult to track trends. Only a small subset of indicators has remained comparable between all three evaluations, reducing the extent to which METT-SA can monitor progress. Thus, it is recommended that if further updates to the METT-SA are contemplated, existing indicators should remain unchanged to allow assessment of long-term trends-even if more indicators need to be added.

Further components of governance and management
The governance and management components considered in this comparison were selected because they represent key issues for which changes could be readily identified over the duration of this study. However, they are not the only important components of MPA governance and management effectiveness. Other important factors include quality of governance, respect for stakeholders and their rights, fair distribution of costs and benefits, and fairness and legitimacy of decision-making processes-including transparency, accountability and dispute resolution (Borrini-Feyerabend et al. 2013; Bennett and Dearden 2014;Schéré et al. 2021;UNEP-WCMC and IUCN 2021). Concerning the legitimacy, accountability, and transparency of governance and management, rights-holders and stakeholders in South Africa have little input into the appointment of MPA management agencies and individual managers. However, some MPAs have prioritised the appointment of MPA managers from the region, which can increase legitimacy and/or acceptability (McNeill et al. 2018;Bennett et al. 2019). This component, as well as the transparency of management decisionmaking, necessitates legal requirements for public access to information, appeals, public hearings and judicial reviews (Borrini-Feyerabend et al. 2013; Bennett and Dearden 2014;Jones and Long 2021). In South Africa, the need for community engagement, local stewardship, and recognition of traditional ownership and knowledge in all MPA processes (policy, implementation, compliance, monitoring, research and spatial planning) has been highlighted as essential to achieving greater acceptance and success of MPAs (Sowman and Sunde 2018;Peer et al. 2022). In cases where plural governance arrangements lead to contradictory views on land-use rights, their hierarchy needs to be clarified to reduce conflict and confusion (Mbatha 2022). The perception that recent MPA expansions have continued to use a top-down approach remains prevalent-despite concerted efforts to be inclusive and consultative-and highlights the need to redress past injustices in any future expansions of the MPA network.
Although not the sole responsibility of MPA authorities, capacity-building for local people, and a long-term commitment to developing supplementary livelihoods and the infrastructure to support these (Borrini-Feyerabend et al. 2013; Bennett and Dearden 2014), are important in an integrated approach to managing MPAs. Realistic supplementary livelihoods for communities impacted by MPAs can reduce pressure on marine resources (Tunley 2009). The absence of supplementary livelihood options has seriously affected compliance in some MPAs (Isaacs and Witbooi 2019). Local communities have, however, benefited from infrastructure and supplementary livelihood options at several South African MPAs, which has often occurred through government initiatives or partnerships with NGOs or the private sector. Several examples of initiatives that have promoted local social and economic benefits are provided in Supplementary Table S3. However, perceptions differ as to whether these benefits are adequate-especially considering the cost carried by those whose livelihoods or food security may be impacted by the existence of MPAs (Sowman et al. 2011;Mbatha and Wynberg 2014). Supplementary livelihood opportunities need to be explored by MPA authorities, who can learn from existing successful initiatives. At the strategic or planning level, the type and format of MPAs must fit their ecological and social contexts and purposes (Borrini-Feyerabend et al. 2013; Bennett and Dearden 2014). There is consensus that no-take areas provide the best protection ( Sala and Giakoumi 2018;Jacquemont et al. 2022), but that multiple-use MPAs with no-take areas and co-management arrangements that balance ecological and social demands are ideal. Although the ecological context has been central to designing and managing MPAs in South Africa (Kirkman et al. 2021), there has been progress with the integration of social and economic contexts into the goals and objectives of MPAs (RSA 2019), with diverse (extractive and non-extractive) uses being considered through appropriate zonation in multipurpose MPAs. However, given the current rigidity of governance systems in South Africa and the type of spatial management tool used (currently confined to MPAs), insufficient attention is given to existing customary governance systems and the role they could play in resource governance and management. This is especially relevant for indigenous fishing communities where alternative models for fisheries management could be explored, such as locally managed marine areas (LMMAs), in which areas are "largely or wholly managed at a local level by the coastal communities, land-owning groups, partner organisations, and/or collaborative government representatives who reside or are based in the immediate area" (Govan et al. 2009, p 28). The LMMA approach has been enthusiastically introduced in Madagascar, Kenya and Tanzania (Rocliffe et al. 2014), but has yet to be trialled in South Africa. Whether this or other models that may meet the criteria for 'other effective area-based conservation measures' (OECMs) (CBD 2018), will be able to better balance South Africa's social and biodiversity objectives (Gould et al. 2021;Gurney et al. 2021) and secure long-term marine conservation, remains to be assessed. The CBD accepts these as alternatives to formally protected conservation areas when assessing conservation gains (CBD 2022) despite the fact that their efficacy needs to be evaluated.

Conclusions
This article has reviewed the past and present governance and management of MPAs in South Africa to assess progress in achieving effective MPA governance and management. Progress (or lack thereof) for various components of governance and management, as reviewed from multiple sources, was assessed across four periods (spanning 1964 to the present day) using a simple scoring system.
Key requirements and opportunities for future improvements in each component were also identified, and include the need for clear MPA objectives to specify the purposes of each MPA, greater consideration of the social contexts of MPAs, and securing stakeholder cooperation. Similar needs were highlighted in a review of the social and economic effectiveness of South Africa's MPAs (Mann-Lang et al. 2021), and studies on the importance of social acceptance of MPAs for them to be effective are well-documented globally (e.g. Jones et al. 2011;McCay and Jones 2011;McNeill et al. 2018;Bennett et al. 2019;UNEP-WCMC and IUCN 2021;Horta e Costa et al. 2022). In addition, the importance of well-managed no-take MPAs (or MPA no-take zones) for effective biodiversity conservation has also been highlighted worldwide (e.g. Edgar et al. 2014) and in South Africa (Kirkman et al. 2021). This objective requires effective enforcement and compliance, which our study has shown is largely inadequate among most South African MPAs. Balancing socio-economic and biodiversity objectives in the context of the historical exclusion and dispossession of coastal communities, as well as ambitious area-based conservation targets in response to escalating biodiversity losses (Brondízio et al. 2019), arguably presents the greatest challenge to current MPA governance and management in South Africa. This includes the contemporary challenge of extending governance and management to the numerous and mostly remote offshore MPAs that were recently declared.
Notably, many of the challenges and deficiencies faced by MPA managers in South Africa stem from their limited operational budgets-an issue that has trickle-down effects on nearly every aspect of protected area management (e.g. staffing, equipment, infrastructure, enforcement, monitoring, stakeholder engagement, education and awareness). Without appropriate, timely, sustained and equitable disbursement of funding to MPAs, many of the future priorities and recommendations covered here will be difficult to achieve. Finally, this article highlights the connection between good governance and effective management and suggests that improving the alignment between the two will require strong coordination and clear communication with stakeholders and between all levels of government. This is particularly true in terms of integrating MPA and fisheries management to address enforcement and compliance challenges. Eight key recommendations are proposed to improve MPA governance and management in South Africa: 1) Ensure that all MPAs have clear and specific objectives that encompass both biodiversity conservation and socio-economic considerations, which will require retrospective reviews for MPAs with unclear purposes. 2) Every MPA (including all newly declared MPAs) must have an up-to-date management plan that is aligned with its objectives. 3) All MPAs (including offshore MPAs) must have dedicated teams of well-trained, suitably equipped and legally empowered staff to support effective enforcement of and compliance with MPA and fisheries regulations, and environmental 'green' courts should be re-established. 4) Develop the means and capacity to enhance the participation of adjacent communities and other stakeholders in MPA design, decision-making and management-including co-management. 5) Acknowledge and address current social impacts and injustices associated with MPAs and catalyse initiatives to implement equitable benefit-sharing. (6) Ensure the financial sustainability of MPAs, including the use of innovative financing models. 7) Strengthen the evaluation of MPA management effectiveness while maintaining consistency in indicators to track progress, complemented by monitoring the ecological, socio-economic and governance attributes linked to MPA objectives. 8) Strengthen relations between the government authorities responsible for MPAs and fisheries to ensure cooperation and coordinated decision-making and management. Motivation for action stems from two sources. The first is a reminder that many of these recommendations reiterate those previously highlighted by publications dating as far back as the report of the Ministerial Committee on Marine Reserves in 1977, through the review of the state of MPAs in South Africa by Attwood et al. (1997), to the three METT-SA assessment reports published between 2009 and 2021. We cannot continue to stall at the stage of providing recommendations: implementation is imperative. The second (and more positive) motivation for action is recognition of the fact that a great deal has been accomplished in the South African MPA sector. A network of representative MPAs has been brought into existence after extensive consultation, greatly expanding the MPA coverage of all ecoregions and habitats to span 5.4% of South Africa's mainland EEZ, which is a massive accomplishment. In line with the aforementioned recommendations, it is necessary to build on these accomplishments to ensure efficient and effective governance and management. environments. The strategy also recommended several steps that could be followed to help coastal communities identify appropriate interventions to achieve sustainable coastal livelihoods. The strategy has the potential to empower coastal fishing communities in South Africa, especially those that need alternative livelihoods as a result of the overexploitation of marine resources. 2021 -Initiation of the process to explore 'other effective area-based conservation measures' (OECMs) in South Africa's coastal and marine space under the DFFE National Marine Biodiversity Scientific Working Group. 2022 -The Kunming-Montreal Global Biodiversity Framework (GBF) of the CBD was adopted during the 15th meeting of the Conference of the Parties (COP 15). The GBF builds on the Convention's previous Strategic Plan and sets out an ambitious pathway to reach the global vision of a world living in harmony with nature by 2050. Target 3 of the GBF is similar to Aichi Target 11, but with increased ambition, calling for 30% each of terrestrial, inland water, and coastal and marine areas to be conserved through protected areas or OECMs. As for the Aichi target, qualitative aspects are emphasised, including a focus on areas of particular importance for biodiversity, ecological representivity and connectivity, as well as equity and effectiveness of governance and management. The target also calls for recognition of respect of indigenous peoples and local communities (IPLCs), their rights and lands. The target acknowledges that sustainable use may be appropriate in conservation areas, but that it must be fully consistent with conservation outcomes.