A broader view of stewardship to achieve conservation and sustainability goals in South Africa

AFFILIATIONS: 1Sustainability Research Unit, Nelson Mandela Metropolitan University, George, South Africa 2South African National Parks – Scientific Services, George, South Africa 3Marine Stewardship Council – Science and Standards Team, London, United Kingdom 4School of Natural Resource Management, Nelson Mandela Metropolitan University, George, South Africa 5World Wide Fund for Nature – Land Programme, Cape Town, South Africa


Introduction
Over the past decade, 'stewardship' has become one of the dominant terms used to describe goals, principles and actions that aim to achieve sustainability in natural resource management, contribute to conservation priorities, and curb environmental degradation that threatens societal well-being. [1][2][3] Stewardship is not a new term 4 , nor is it unique to a conservation perspective, e.g. in corporate management 5 . Even within an environmental context, its definition and interpretation varies greatly in its scale and application. At planetary scale, the terms 'ecosystem' and 'earth' stewardship are sometimes used interchangeably (e.g. by Chapin et al. 2,6 ) to describe an overarching framework for dealing with social-ecological vulnerability and promoting general actions and systems that would enhance resilience in the light of global environmental change. [6][7][8] On the other side of the spectrum, the stewardship tag is also applied in a much more focused manner, e.g. to describe market-linked incentives such as certification schemes for specific commodities such as the Marine Stewardship Council (MSC) for fisheries 9 and Forest Stewardship Council (FSC) for timber 10 . In some countries, stewardship is mostly associated with sustainability in agri-environmental systems (e.g. in the United Kingdom 11 ) or with the adoption of better land and catchment management (e.g. 'Landcare' in Australia 12 ), while elsewhere it may designate the management of formally protected or wilderness areas 13 . In South Africa today, 'stewardship' in a literal sense is understood to refer mainly to protecting biodiversity on privately owned land, under the banner of so-called Biodiversity Stewardship Programmes (BSPs). 14,15 Although such initiatives were identified over a decade earlier as a strategy to incentivise 'off-reserve' conservation 16 , and well before adopting the term stewardship, it is now rarely used in any other context. The broadscale, global interpretation of stewardship is based primarily on a developed country perspective, in which it is viewed by some as a possible 'successor' to earlier resource management regimes (namely steadystate and ecosystem management approaches -see Chapin et al. 2 ). This view is embodied by a set of nine 'stewardship goals' 8 widely accepted as the guidelines for promoting earth stewardship 17 . These goals, grounded in the theory of social-ecological sustainability, include predominantly social aspects, e.g. equitable access to basic needs and opportunities, and sustaining ecosystem services. 3 The goals also include a number of other crosscutting characteristics: (1) voluntary (as opposed to mandatory) participation 18 ; (2) altruistic and moral-ethical connotations, sometimes associated with religion 19,20 that engender a sense of care 21 and shared responsibility, with consideration for the interests of human society, other species, and the natural world 22 ; (3) an emphasis on inter-generational rather than short-term benefits 8 ; (4) applicability across different spatial scales, i.e. from 'backyard to planet' 23 ; and (5) the need for multiple partnerships, collaborations and linkages. Social networks and the stakeholder relationships that they represent are increasingly recognised as important features of natural resource management and conservation approaches. 24,25 Stewardship actions are often visualised as networks of actors with linkages within specific contexts, e.g. in urban ecosystems 26 , at multiple scales 27 , or across institutional and other divides 28 .
Achieving these stewardship goals will require the implementation of practical mechanisms that could be viewed as the 'building blocks' of earth stewardship, preferably with metrics to indicate progress. 8 Such mechanisms could encompass decisions and actions at multiple scales (local, regional and global) based on the familiar principles of 'reduce, reuse and recycle' 29 . Other practical contributions may be the practice of 'civic' or 'urban' ecology: something as simple as planting a tree in the neighbourhood 28 or as complex as incorporating ecological principles into urban designs 30 . In practice, it is sometimes difficult to judge how stewardship differs from other environmental governance or natural resource management systems, or similar concepts like custodianship or trusteeship. 31 For example, co-management 32 , like stewardship, is not necessarily driven purely by conservation objectives 33 but also by the need for benefit sharing 34 . In developing countries, stewardship principles are inherent to many community-based management systems 35 based on traditional and indigenous cultural values and beliefs (but for an opposing view see Fennell 36 ).
One of the most compelling notions to emerge from proponents of the Earth Stewardship Initiative of the Ecological Society of America is the opportunity for less developed countries to 'leap frog' steps (e.g. steady-state resource management) on a typical Western resource management continuum directly to stewardship (see Figure 1 in Chapin et al. 2 ), presumably avoiding the unsustainable practices of the past. How does this perspective relate to advancements in a developing or middle-income country context in which socio-economic disparities (e.g. developmental and income gaps) are far more pronounced and capacity to implement stewardship may be reduced? To assess this question we: (1) identified stewardship or stewardship-like mechanisms and their proponents or implementers in terrestrial, freshwater, and marine social-ecological systems in South Africa; (2) examined the relative influence and relationships between identified organisations or actors as a network; and (3) present the stewardship schemes as a simple typology, based on their objectives and operational scales. We discuss our findings relative to South African conservation and sustainability priorities, and in the wider context of Earth Stewardship Goals. 8 Our results not only provide a broader overview than the more traditional interpretation of stewardship in South Africa, but also allow us to reflect on whether stewardship has indeed emerged as a possible holistic or 'fast-track' option toward achieving conservation and sustainability goals.

Scoping and information retrieval
We anticipated considerable variation in the literal use of the term 'stewardship' and whether or how it is applied to initiatives that may be considered as stewardship activities. Therefore, during our review process, we did not take a purely systematic approach and adopted the following broad definition of stewardship: Any initiative, activity or voluntary involvement by an individual or organisation in the private, nongovernmental or governmental sectors (including parastatal agencies), which seeks to contribute to, or promote, natural resource conservation or sustainability goals in social-ecological systems, both terrestrial and aquatic.
Further selection criteria were: (1) voluntary participation, i.e. not legislated (although a legal framework might apply); (2) non-commercial motivation, while acknowledging some operational costs (e.g. auditing costs for an eco-label); and (3) a natural resource or ecosystem management focus (as opposed to industrial processes).
We identified stewardship-related activities or initiatives ('schemes') and the most prominent organisations, individuals and other stakeholders involved in promoting and implementing these -collectively referred to as 'actors'. Importantly, compliance with our definition rather than explicit association with the term 'stewardship' was the main criterion for inclusion into our database, which was populated using both systematic and non-systematic search methods over a period of about 16 months (October 2012 -January 2014). We only considered schemes active within, but not necessarily restricted to, the Republic of South Africa. Actors could be based anywhere.
We did initial scoping through keyword searches on the Internet and in primary scientific indexing services using the terms 'stewardship' and 'Africa'. Next, we expanded our list of actors, schemes and associated terminology by a process of chain-referral (cf. snowball sampling 37 ). We contacted or met with the most prominent actors, and asked about their own involvement in stewardship and for referrals to others, allowing us to identify more cryptic actors or schemes. Some referrals included suggestions to attend specific local and international meetings, including the Fynbos Forum (Cape St Francis, South Africa, July 2012), and the Symposium on Science & Stewardship to Protect & Sustain Wilderness Values at the 10th World Wilderness Congress (Salamanca, Spain, October 2013). Finally we conducted a more exhaustive round of searches based on two models of information retrieval: the 'berrypicking' model of Bates 38 , and the 'Web moves' behavioural model of Choo 39 . The first is an 'evolving' search approach in which the cognitive response to results by the researcher may lead to on-the-fly modifications to the search process, e.g. by adding additional terms such as 'custodianship' or 'conservancies', or doing searches on a specific organisation, to broaden the sample. The latter model describes a progression of 'moves' whereby the researcher, starting on one website, follows links to other sites with relevant content ('chaining'), scans browsing results for most prominent returns, and differentiates between various results while bookmarking or capturing useful information. It includes an element of 'monitoring' whereby the sites are checked for updates and changes, and 'extraction' whereby a site is systematically searched for pertinent information (including type of scheme and its objectives, scale and mechanism of action). 40 These approaches enabled us to satisfy the objective of capturing the most readily available information on representative examples of stewardship within the region, including websites and primary scientific, academic and grey literature. As a minimum, for an initiative to be included, it had to comply with our definition and criteria, and we recorded additional information needed to identify the type of actor and scheme (detailed below).

Social network visualisation
We classified actors into five broad categories: (1) non-governmental organisations (NGOs), including registered charities or not-forprofit organisations; (2) funds -organisations that provide financing for schemes but do not normally undertake implementation; (3) governmental entities, including national and provincial ministries, departments or agencies; (4) private entities, including profit-driven companies and industry associations; and (5) partnerships -other groupings that do not fit the statutory entities described by (1)-(4), including collaborative networks, associations or programmes. All these actors represented the 'nodes' in our network. We then identified direct linkages ('edges') between pairs of nodes from stated collaborations on web pages and other documents, or implied through co-branding or logos on stewardship schemes, with each link assigned an arbitrary weight of one (i.e. multiple collaborations between the same actors would result in a weight greater than one). It is important to note that deriving linkages in this way did not allow us to assign directionality. We visualised relationships between actors as a social network using the software Gephi 0.8.2 beta 41 . We used the betweenness centrality -the number of times a node rests between two others which themselves are not linked -as a measure of relative prominence in stewardship (calculated with the algorithm of Brandes 42 ). Actors with high betweenness centrality are considered important for long-term resource management planning, bringing together disconnected segments of a network. 43

Typology
We identified broad types of stewardship schemes compliant with our definition, and based on information available about their objectives (e.g. focus on biodiversity or ecosystem services), mechanism of action (e.g. conservation on private land or market-based incentives), operational scale or footprint (global, national, sub-national or local). Despite some overlap between schemes and a lack of quantitative measures, we

Stewardship network and typology
Our final database included 38 NGOs, 14 governmental entities, 10 private entities, 5 funds and 27 partnerships (between any of the recorded actors). These 94 nodes and 180 edges between them formed the basis for the social network visualisation (Figure 1; see also Appendix 1 and the supplementary material). Among these there were seven global NGOs, and three global funds; most other NGOs were national (n=18) or sub-national (n=10), noting that international NGOs with South African branches were considered 'national', e.g. the World Wide Fund for Nature (WWF-SA) and Conservation South Africa. Partnerships were mostly national (n=7) or sub-national (n=16) with three examples of global and one local partnership. The network appears well-connected with many linkages but relatively few prominent actors, and some disconnected nodes. Most prominent with the highest betweenness centrality values were NGOs and partnerships that relate to biodiversity and landscape conservation initiatives, notably the Cape Action Plan for People and the Environment (CAPE) -a systematic conservation plan for the Cape Floristic Region, initiated in 1998 with funding from the Global Environmental Facility's Critical Ecosystem Partnership Fund (CEPF) and coordinated by WWF-SA 44 . National (e.g. South African National Biodiversity Institute) and sub-national governmental agencies (e.g. CapeNature) also feature as important bridging nodes.
Note: green = non-governmental oganisations; yellow = partnerships; red = government entities; blue = funds; lilac = private entities We distinguished six main types of stewardship schemes (defined in Table 1), some with sub-types or some degree of overlap. In terms of scale, 5 schemes were operational at a global scale, 2 at the African continent level, 13 at national, 38 at sub-national (i.e. coverage limited to one or more of South Africa's nine provinces), and 19 at local level (i.e. limited to smaller areas such as a city or catchment). Based on the above we positioned the main stewardship types or sub-types on a conceptual plane ( Figure 2) and describe and present examples of each type in more detail below (also see Table 1).

Conservancies
Conservancies represent the oldest form of voluntary conservation on private land in South Africa -the first conservancy was established by a group of farmers in 1978 in the Balgowan District of KwaZulu-Natal (KZN), through encouragement by the former Natal Parks Board (now Ezemvelo KZN Wildlife). Although conservancies are required to be registered with the regional conservation authority, there is no binding agreement between these parties. Conservancies are viewed as the entry-level to more formal stewardship agreements, but are not   Definitions of main types of conservation and sustainability stewardship schemes identified in South Africa

Conservancies
Registered voluntary associations, established between like-minded landowners, residents, communities and other users, in a specified area with the shared aim of co-operative management of its natural resources in an environmentally sustainable manner, without necessarily changing the land use on the properties.
Biodiversity stewardship programmes Mechanism to incentivise formal conservation on private lands with high biodiversity conservation value. Different participation levels are available but the ultimate aim is to proclaim such areas as formally protected by national laws.

Land-and seascape initiatives
Initiatives that focus at a land-or seascape level, often determined by unique or specific biophysical or other characteristics or features (e.g. geological or heritage), to promote resilience of protected areas through inclusion of buffer areas, or enhanced connectivity between formally protected areas through multiple mechanisms.

Market-linked schemes
Initiatives that focus on the production, management, or value chain of specific commodities or services and aim to promote sustainability by incentivising consumers to support such schemes, thus harnessing market forces to reward such producers.
Ecosystem services Initiatives that broadly address issues around maintenance or restoration of ecological infrastructure or ecosystem services through practical or policy interventions.

Education and awareness initiatives
Initiatives aimed at education or raising awareness in specific or multiple sectors of society (e.g. the youth, or consumers and retailers) about particular or broader issues relating to sustainability or conservation, thus encouraging the voluntary adoption of behaviours and attitudes that contribute to such causes.
generally recognised as components of BSPs (see below). Participation is often based on shared aims and a sense of identity (e.g. expressed through logos displayed in media forums and individual farm signage), which may enable members to access funds to implement conservation action, if they so choose.

Biodiversity stewardship
The conservation of biodiversity using so-called 'stewardship agreements' was conceived at national level by the South African National Biodiversity Institute (SANBI) but is implemented sub-nationally by provincial conservation agencies and NGOs. Biodiversity stewardship is most prominent in the Western Cape where it was piloted as the 'Conservation Stewardship Programme' in late 2002 by CapeNature and the Botanical Society 48 as part of the CAPE strategy [49][50][51] . The approach was later adopted in other provinces such as KZN. 52 The underlying objective of these BSPs is to improve protection of critical biodiversity and threatened ecosystems occurring on private and communal land as determined by national conservation plans and spatial assessments. 53,54 This protection is to be achieved by encouraging formal conservation agreements between the conservation agency and landowners through financial (e.g. tax relief 55,56 ) and in-kind (extension services -habitat and land management advisory) incentives. The programme recognises various levels of participation, namely biodiversity agreements, protected environments and contract nature reserves (as defined in the Protected Areas Act 57 ) that differ in degree of legal protection status, land-use restriction (on title deeds) and minimum duration of management tenure: 10 years for biodiversity agreements, 30 years for protected environments and 99 years for contract nature reserves 48,51  In other provinces (e.g. Gauteng and Eastern Cape), BSPs are more recent (post-2009) and outcomes are not readily available. As the programme has developed, more local authorities and NGOs have expressed interest in adopting this model.

Land-and seascape initiatives
These schemes share a broad focus at land-(or sea-) scape level, usually determined by unique or exceptional biodiversity, geographical features or other characteristics, sometimes in combination. They aim to improve the protection of an area by raising awareness about the unique features or conservation profiles through special listings or other means of recognition. They vary in spatial scale from sub-national to regional, although some are international initiatives e.g. the United Nations Educational, Scientific, and Cultural Organisation's (UNESCO) Man and the Biosphere (MAB) Programme 59 and World Heritage Sites 60 . Landscape initiatives often correspond closely with national (e.g. CAPE) or international (e.g. IUCN 'Key Biodiversity Areas') bioregional programmes.

Corridors
One of the most diverse sub-types of landscape initiatives involve the concept of corridors. These include corridors that link formally protected areas primarily for conservation of biodiversity and processes (e.g.

Market-based schemes
These schemes focused on environmental sustainability objectives (we did not consider primarily social ones such as Fair Trade) at the resource production or ecosystem level of a value chain, by trying to influence consumers to reward more sustainable supplies of a product (e.g. seafood) or service (e.g. tourism) through their choices. 69 There were two sub-types: eco-labels and business and biodiversity initiatives.

Eco-labels
Eco-labels rely on a certification standard for a specified commodity or service; its adoption entitles the producer/service provider to use the eco-label mark as a marketing tool. These included leading international third party eco-labels: one MSC certified fishery (South African demersal 'Cape' hake (Merluccius spp.) trawl fishery of ca 120 000 t per year, first certified in 2004); 20 forestry management areas certified by the FSC covering >1.48 million ha; and the Blue Flag tourism eco-label for 36 beaches, 4 marinas, and 3 whale-watching boats implemented by the Wildlife and Environment Society of South Africa (WESSA) 70 . There were several national eco-labels addressing specific issues, e.g. badger-friendly honey, predator-friendly meat 71 , sustainable golf courses (e.g. one in Audubon Cooperative Sanctuary Program) or tourism accommodation (e.g. Green Leaf). At a continental level, the African Eco-Labelling Mechanism 72 has developed draft standards for the agriculture, fishery and forestry sectors. It is important to note that national or regional eco-labels do not always make use of third-party verification or traceability mechanisms for certified products.

Business and biodiversity initiatives
Business and biodiversity initiatives (BBIs) focus on production systems for specific products by making a 'business case' for biodiversity conser-

Ecosystem services
These schemes focus on restoration of specific ecosystem services. They include primarily government-driven initiatives, but with more or less voluntary adoption (or in lieu of financial payments for ecosystem services 78  Other schemes more specific to agricultural production include state initiatives such as the National LandCare Programme 83 which addresses, inter alia, soil management and erosion control on farms. Others are driven by NGO and private/corporate partnerships, like the Sustainable Sugarcane Farm Management System (known as SUSFARMS). 84 Some water stewardship initiatives focus strongly on the link between the supply chain and catchment management, e.g. WWF Water Futures Partnership with SABMiller on hops production 85 , or the standards set by the Alliance for Water Stewardship that have now been adopted by South African producers of export stone fruits 86,87 .

Education and awareness
These schemes either focus on a specific cause, e.g. sustainable seafood, or incorporate information about multiple causes into a 'basket' of sustainable options aimed at the general public. They also promote more sustainable living among specific sectors, e.g. scholars, through actions such as saving water and energy or recycling. For example, the Southern African Sustainable Seafood Initiative (SASSI) which encourages seafood consumers to consult a 'traffic-light' species list of sustainable seafood choices when buying fish 88,89 ; through this market pressure its influence may extend into regulatory or policy areas 90 . Eco-Schools 91 is a sustainable schools programme from the international NGO Foundation for Environmental Education, but implemented in South Africa by WESSA with 1200 registered schools. Often, because education and awareness are ancillary functions to the main objectives of NGOs, such schemes were difficult to isolate, and tend to have a cross-cutting function (represented by the arrow in Figure 2) by linking multiple schemes, e.g. GreenChoice which markets a 'basket' of sustainable options from different schemes (including eco-labels and BBIs) to the general public 92 .

Discussion
Our broad overview of stewardship schemes in South Africa is, to our knowledge, the first such at a countrywide scale. Our findings represent a much wider perspective on stewardship than has ever been used in any developing country. We present our findings under broad themes that aim to capture the key features of stewardship in South Africa, while maintaining a global context.

Biodiversity focus and the role of NGOs and partnerships
The strong focus on biodiversity conservation on private land over the past decade is perhaps not surprising, given that much of South Africa's globally recognised biodiversity and threatened environments, especially in the Cape Floristic Region, is located outside of formally protected areas. 54 This focus not only explains the prominence of CAPE (Figure 1), but also why many aspects of BSPs, especially within the CAPE planning domain, have been examined more in-depth: policy and governance frameworks 93,94 ; perceptions and motivations for participation 95 , e.g. tax incentives 55,56 ; the relationship between biodiversity stewardship and social learning 96 ; and evaluating the contribution of BSPs to national conservation goals 49 . The CAPE partnership, together with major 'global' NGOs (e.g. WWF-SA), form dominant elements of the stewardship network, in effect combining as a 'bridging organisation' 97 . Such organisations, on the one hand, leverage external resources or 'bridging ties' like international funding (e.g. from the CEPF), while on the other hand, connect and enable diverse local actors to utilise new 'possibilities for action' 98 . Although our data set did not allow an in-depth analysis or understanding of the stewardship network, it suggests that more social network analysis could be valuable in gaining a better understanding of stewardship at specific spatial scales, within specific groups of actors, and the links between international and local conservation priorities and actions. 99

Non-biodiversity goals: Common pool, markets and ecosystem services
In contrast to the above, some stewardship schemes are not necessarily tied to biodiversity and bioregional focus. For example, the establishment of conservancies pre-dates spatial prioritisations. Conservancies are found in all provinces and motivations for their establishment are more diverse, sometimes tending toward self-interest (see below). Another exception to a singular biodiversity focus is stewardship schemes dealing with common pool resources, value chains and markets, or ecosystem services. Marine and coastal ecosystems present classical examples of common pool natural resources 103 held in 'public trust' by the state on behalf of its citizens 104 . Stewardship activities by citizens or interest groups in the marine environment thus present something of a conundrum: they are trying to be co-stewards of something already under government custodianship on their behalf (but see the concept of marine citizenship 105 ). In the South African context, stewardship schemes in the marine environment are predominantly market-based or educational (e.g. MSC, Blue Flag and SASSI) with seascape-level schemes such as International Ocean Hope Spots only a recent development -not unexpected when bioregional planning and prioritisation has lagged in the marine environment. Surprisingly, co-management, which is generally considered conducive to sustainable harvesting and resource stewardship 106,107 , has struggled to emerge within South Africa's current fisheries management regime 108 . Increasingly, the term 'stewardship' is adopted to describe collaborative governance approaches to manage global commons such as the deep ocean. 109  Stewardship based on value chains, markets or commodities can sometimes be at odds with biodiversity conservation; for example, the FSC, which in South Africa primarily certifies monoculture plantations of exotic (often invasive) tree species like pines and gums, located in biodiverse fynbos and grassland habitats. Ironically, the only exception to this contradiction is the FSC certificate held by South African National Parks for harvesting indigenous hardwoods in the Garden Route National Park. The fact that 'plantations are not forests' is strongly advocated by some lobby groups. 111 Similarly, many conservationists dispute that any bottom trawl fishery should be certified as sustainable. 112 Focus on specific ecosystem services or concepts like biodiversity offsets (or other mitigation measures) within the stewardship discourse is likely to remain uncomfortable, if not controversial, when there is evidence that nonbiodiversity objectives are not always compatible with biodiversity ones. 113

Motivations and mechanisms
Published sources suggest that intrinsic motivations to participate in stewardship include altruism and acting in societal interest. 114 Although environmental consciousness (cf. biophilia 115 ) is an assumed prerequisite for private landowners to create conservancies in South Africa, a range of reasons are reported, ranging from nature conservation (primary) and security for domestic and wild animals to securing recreational or tourism opportunities including hunting, or sometimes to oppose development. 116 Some of these reasons may be equally applicable when entering into more formal BSP arrangements, but often it is up to the proponent (e.g. provincial conservation agency and NGOs) to 'sell' the concept to the potential steward. Incentives may include financial ones 55 , but also 'extension services': specialist input and management assistance relating to land and biodiversity. The type of landowner (commercial versus lifestyle farmer), land size and opportunity costs can all impact on willingness to participate in conservation. 117 Recent work using the Biodiversity & Wine Initiative as example, suggests that both intrinsic and extrinsic factors are important for farmers to join this BBI, notably their own value systems. 118 The importance of issue 'champions' as a key driver for participation was also emphasised.
International eco-labels are seen to inadvertently encourage global steward ship by empowering mainly northern hemisphere consumers to take personal responsibility for the production of a commodity elsewhere, especially in the developing world. 119 Initially, adoption of both the FSC and MSC in South Africa was motivated by the demands of the export market, rather than local consumer choice. 120,121 Some argue that payment for ecosystem services is inherently easier to leverage from a business perspective than payment for biodiversity 122 , hence diversification of stewardship mechanisms to include ecosystem services 78 or value chains. Market-based interventions, together with consumer awareness schemes (e.g. GreenChoice), contribute to making a 'business case' for biodiversity conservation. 76 The notion of a business case often finds resonance and expression in corporate stewardship 'sustainability journeys' of retailers 123 , although there are possible weaknesses in using 'journey' as a metaphor for measuring progress in sustainability 124 .

Successes, benefits and shortcomings
Although we could not directly measure the efficacy and drawbacks of stewardship schemes from our data, some published results offer indicators of their success. These indicators include participation levels in stewardship schemes, hectares of land in BSPs, or more tangible conservation outcomes, for example, significant reduction in seabird mortality in the hake trawl as a result of MSC certification 125 . Conversely, the withdrawal of Blue Flag status at Margate in KZN because of poor water quality has been equated to a substantial revenue loss. 126 For conservancies, the growing number of voluntary participants, their presence in all provinces, and a national alliance that includes 'community level stewardship' in its vision are all positive trends. There are several recognised benefits to game ranching of consolidated estates, including more profitable (from an eco-tourism perspective) and viable wildlife populations, especially for larger species with bigger ranges. 127 Another apparent benefit of BSPs is achieving national conservation goals at much lower cost to the state (than land acquisition). While this may be so for provincial agencies, conservation on private land is sometimes viewed as an 'unfunded mandate' by national agencies (e.g. South African National Parks) -in other words, the budgetary and human resource requirements are not commensurate with the area to be managed. Thus, while participating in BSPs is considered 'voluntary', the underlying biodiversity objectives may confound the voluntary nature of participation, as land with 'low' conservation value is not wanted, given the financial and human capacity requirements for extension services and other management costs. There must remain serious concerns regarding the statutory security of conservancies and other forms of biodiversity stewardship. The dependence of conservancies on the personal values of the participant casts doubt on whether they should be included under national conservation targets. 116 For instance, an evaluation in 2010 of 280 BBI members indicated coverage of 250 153 ha of natural habitat; however, since 2006 there has been a loss of 2827 ha to habitat transformation and 892 ha to degradation, 128 bringing some doubt over the sensibility of 'banking' on a volunteer mechanism to achieve national mandates.
Inasmuch as governmental agencies tasked with biodiversity conservation have embraced these new governance arrangements 93 to achieve conservation targets, a single-minded focus on one mechanism may have additional drawbacks. For example, it may inadvertently cause neglect on other land with equally important biodiversity, such as 'escapee' pines invading state-controlled watersheds. 129 The position of the state may even appear 'schizophrenic', especially when the state defaults on its fiduciary duty as public biodiversity custodian. 130 At times, a government's action or inaction may pose a direct threat to biodiversity inside and outside protected areas, e.g. by assigning prospecting rights for shale gas across entire bioregions, or for benthic phosphate mining 131 ; by permitting coal mining adjacent to nature reserves (e.g. at Hluhluwe-Imfolozi 132 ); or by on-going political support to permit angling in Africa's oldest no-take Marine Protected Area, Tsitsikamma 133,134 (recently gazetted by the Department of Environmental Affairs 135 ). In such instances, civic or special interest groups or industries may adopt stewardship as an anti-measure to such threats 136  A contrasting scenario is presented by the primarily government driven schemes for ecosystem services restoration. For example, the cost-effectiveness of WfW has been assessed at local and national scales 81,144 and, although many regard it as overwhelmingly positive, there is a sense that its overall performance needs to be improved, inter alia, by better prioritisation of alien invasive species, more targeted actions, and less emphasis on social benefits as a measure of success. Even less successful has been the ability to stimulate stewardship actions among private landowners, i.e. by maintaining cleared areas to prevent re-invasion after initial WfW clearing.

Relationship to other management approaches
The stewardship approaches found in South Africa are similar to other conservation and natural resource management strategies elsewhere. For example, conservation easements in the United States of America 146,147 are very similar to biodiversity stewardship agreements in South Africa: the term 'cooperative environmental governance systems' has been used to describe such arrangements 148 . Some believe that stewardship differs from other management systems by its recognition of 'embedded values' and preoccupation with conservation and sustainability. 31 However, these traits are common to community-based natural resource management (CBNRM) -another decentralised approach to achieving environmental, social and economic goals by balancing the exploitation and conservation of valued ecosystem components. 149 In CBNRM, voluntary local civic institutional arrangements are formed to manage natural resources, suggesting that it may be viewed as a form of stewardship 150 or as a mechanism for achieving stewardship of watersheds 151 , and wildlife and forests 152,153 , and for sustainable rural agriculture 154 . Further, developing social capital, collaborative partnerships and networks have been highlighted as key principles of CBNRM 155 , which is echoed in the stewardship metaphor 156 . It is noteworthy that CBNRM predominantly focuses on common pool resources and is often underpinned by cultural and traditional values, e.g. the conservation of sacred landscapes 157 , totemic species or culturally important natural features 158 .
In practice, different conservation mechanisms are rarely applied in isolation and stewardship schemes may be seen as ancillary to tactics such as land acquisition (e.g. by the Nature Conservancy in the United States of America) that all form part of a modern strategic conservation approach 159 in areas of high biodiversity, as in South Africa 160 . For example, WWF-SA, in addition to facilitating private land stewardship, actively pursues the expansion of existing or establishment of new protected areas through land acquisitions (ca 400 000 ha or 5% of the national terrestrial protected area estate), predominantly financed through land trusts.

Earth stewardship in South Africa?
In retrospect, it is apparent that many of the described stewardship schemes could fit under the banner of 'earth stewardship': operational at multiple scales with diverse stakeholders, and recognising interconnectedness, ethics and indigenous knowledge (for a snapshot of examples see Sayre et al. 161 and papers in that volume). However, the dominance of contractual biodiversity conservation initiatives in the South African stewardship narrative has masked the emergence of a more holistic stewardship strategy as advancement on contemporary resource management approaches (as contemplated by Chapin et al. 2 ).
The narrow association of stewardship with systematic biodiversity conservation plans and associated spatial priorities is perhaps not surprising, given that most post-colonial countries still develop within the 'constraints' of governance or management systems inherited from the North -much as 'global assemblages' can impact on poorer nations 162 -in effect inhibiting the emergence of earth stewardship. Furthermore, weaker governance in some developing countries has driven 'decentralised' mechanisms of environmental decision-making and policy implementation at community level. 163 Our research approach was unlikely to provide adequate resolution to detect the emergence of local-level governance approaches, especially for tacit (e.g. CBNRM and co-management) and local grassroots level (e.g. urban greening initiatives) stewardship forms that were underrepresented in the information sources we consulted, and so more difficult to detect. In fact, the recent broadening of the stewardship narrative in South Africa -firstly by NGOs adopting the term 'earth stewardship' 164 and, secondly, through the fairly rapid diversification away from a strictly biodiversity focus to more holistic models -suggests that a shift is taking place. In countries with a legacy of post-colonial land ownership, this shift may reflect recognition of the need to acknowledge and incorporate local socio-political issues into any stewardship approaches. This new approach is typified by the Community Ecosystems Based Adaptation ('CEBA') sites of the Wildlands Conservation Trust which uses a 'basket of products' approach (including 'Green-preneurship' and restoration), with implementation strategies 165 that mirror many of the earth stewardship principles, while strongly emphasising involvement of local communities. There is thus a clear need to evolve Westernbased concepts of stewardship and conservation to include indigenous values 157 or more collaborative management approaches 166 .
Finally, while we believe that 'earth stewardship' may be an appropriate metaphoric term to describe the link between primarily conservationdriven schemes with more social and economic ones, it is unlikely to be an implementable 'catch-all' solution in countries with weak or ineffective governance systems (as suggested by Kinzig et al. 167 ). Overuse of the term in a global or philosophical sense may eventually dilute its value and practicality in an implementation context. We contend that, to achieve sustainability or conservation outcomes, reliance on a single mechanism -whether voluntary (i.e. 'stewardship') and thus dependent on the social norms, ethics, values or behaviours of individuals, or as determined by government policies (i.e. mandatory or legislated) -is risky. This contention is important, as formalising any voluntary participation into binding agreements may result in issues similar to those faced by extant formal management systems, e.g. lack of capacity or 'non-compliance' by participants, and corruption. In a developing world context the need for complementarity between different management approaches 168 is key to achieving the desired conservation and sustainability outcomes.