Introduction

Transfrontier conservation areas (TFCAs) play an important role in the long-term cooperative conservation of biodiversity, cultural values, and supporting ecosystem management. They cover vast landscapes spanning borders between two or more countries and include at least one protected area (Wolmer 2003; Lunstrum 2011). TFCAs (also referred to as ‘Peace Parks’) pursue peace, conservation, and socio-economic development across ‘man-made’ borders (van Aarde and Jackson 2007; Hanks and Myburgh 2015). More specific objectives include (i) long-term cooperative conservation of biodiversity and cultural values (ii) promoting landscape-level ecosystem management; (iii) trust and capacity building amongst actors; (iv) joint learning and knowledge co-creation; (v) sharing resource management skills, information, and experience and (vi) promoting access to natural resources and their equitable and sustainable use (Hanks 2003; McKeever 2008; Chitakira et al. 2022). The wide range of objectives that spread across different countries and administrative levels in TFCAs pose unique challenges in the governance of such transfrontier landscapes.

With the growing human-induced transformation of landscapes over the past decades, landscapes have become more fragmented and hence vulnerable to biodiversity loss (Elliot et al. 2014; Cushman et al. 2018). One of the fundamental objectives of the TFCA is biodiversity conservation which entails connecting and jointly managing fragmented habitats and natural systems (Elliot et al. 2014; Loveridge et al. 2022). Biodiversity conservation is considered a vital approach for protecting ecologically valuable landscapes and wildlife (Wolmer 2003; van Aarde and Jackson 2007; Hanks and Myburgh 2015). Various strategies which include habitat restoration, protected area management, community-based conservation programs, and the development of transboundary policy frameworks have been implemented to foster the conservation of biodiversity in TFCAs (Michel et al. 2006; Loarie et al. 2009; Kansky et al. 2021). Therefore, TFCAs serve as a supporting concept for the improved management of large-scale, naturally interconnected areas for biodiversity conservation. Successful biodiversity conservation also depends on socio-economic aspects (Selier et al. 2016; Ntuli et al. 2019). Supporting local livelihoods and pursuing community-based conservation and natural resource management is expected to serve, secure, and even improve conservation outcomes and the well-being of local communities (Selier et al 2016; Kansky et al. 2021). Balancing conservation with sustainable development is vital for the long-term viability of TFCAs. TFCAs pursue the management of vast and biodiversity-rich landscapes through fostering a sense of ownership and responsibility among local actors by acknowledging the significance of biodiversity and wildlife (Hanks 2003; Chitakira et al. 2012; Chiutsi and Saarinen 2017). This sense of ownership encourages a deeper connection and commitment to the landscape, resulting in increased engagement, collaboration, and efficient application of governance strategies (Chitakira et al. 2012; Chirozva et al. 2013). These outcomes also illustrate the principles of landscape governance, which calls for integrated management approaches to protect ecological connectivity, biodiversity hotspots, and promote sustainable development of transfrontier landscapes (Sayer et al. 2013; Ros-Tonen et al. 2014).

Landscape Governance (LG) refers to the collective, multi-level processes, institutions, and mechanisms through which decisions and actions regarding landscapes are made, implemented, and monitored (Lemos and Agrawal 2006; Penker 2009; Bennett and Satterfield 2018). It involves the coordination and management of various stakeholders (Görg 2007; Elbakidze et al. 2010; Njoroge et al. 2020), including governments, communities, civil society organizations, and businesses, to achieve sustainable and equitable landscape outcomes. By involving local communities and stakeholders in decision-making processes, LG can contribute to improved livelihoods, social equity, and well-being. It recognizes the importance of local knowledge (Olsson et al. 2004; Angelstam et al. 2013), traditional practices (Langston et al. 2017; Špaček et al. 2022), and cultural values (Olsson et al. 2004; Plieninger and Bieling 2012). Active involvement of different actors (multi-actor) in such governance processes, supports the activation and integration of different ‘bodies of knowledge’ (Lang et al. 2012), knowledge systems (Rathwell et al. 2015; Ingram 2018), and different types of expertise (e.g., know-how; know-what) (Bammer et al. 2010). Consequently, LG is also expected to strengthen the resilience of landscapes and communities by fostering adaptive management. It supports the integration of relevant issues like climate change adaptation, disaster risk reduction, land degradation, biodiversity loss, and food insecurity (Hoole and Berkes 2010; Ros-Tonen et al. 2014; Robinson and Kagombe 2018; van Oosten et al. 2021) and thus ensures that landscapes and communities are better prepared for future challenges. Thus, LG recognize landscapes as complex socio-ecological systems where nature, humanity, and culture interact and influence one another. It acknowledges that effective management and decision-making require a holistic and integrated approach that takes into consideration competing interests, values, perceptions, and objectives (Dressler and Büscher 2008; Reed et al. 2016; Levin et al. 2018). Governing transfrontier landscapes that span large geographical areas, encompassing different institutional systems (countries), and involving numerous stakeholders with competing interests create specific challenges. Effective communication, transnational and intra-country (horizontal and vertical) coordination, and collaboration among stakeholders can be demanding when dealing with diverse values, perspectives, conflicting priorities, power dynamics and legacies (Elbakidze et al. 2010; Sayer et al. 2013; Reed et al. 2016).

While LG can be a valuable approach to landscape management, it is important to be attentive to potential governance failures resulting from the multifaceted context. Governance failures can result in environmental crises such as the degradation of vital landscapes (Pahl-Wostl 2009; Plieninger et al. 2014; Foli et al. 2018). While governance failures are only modestly addressed in LG discourses, there is ample evidence in related scholarly debates covering forested landscapes (Neudert et al. 2017; Gellert 2022) and water resources (Walker 2014; Nicollier, et al. 2022). Governance failures cover different modes of governance, such as legal, market and network governance (Howlett and Ramesh 2014; Pahl-Wostl 2019). Hence governance failure can be attributed to a mismatch of the governance mode to problem context (governance design) or governance capacity (resources and skills) (Howlett and Ramesh 2014; Jessop 2023). Such governance failures can be attributed to different reasons, such as (i) outdated LG practices with shortcomings regarding local participation (Reed 2008; Milder et al. 2014), power struggles, and asymmetries (Dawson and Martin 2015; Ros-Tonen et al. 2018) or (ii) the strong reliance on single-objective based strategies which are only conservation oriented neglecting integration of socio-economic or other policy goals (Kostov and Lingard 2004; Sayer et al. 2013; Reed et al. 2016). Operating within existing institutional frameworks and governance structures that predominantly focus on specific sectors and remain fragmented can result in a lack of coordination and integration across sectors, leading to conflicting policies, inefficient resource allocation, and missed opportunities for holistic and sustainable LG. Overcoming sectoral silos requires improved cross-sectoral collaboration and the development of integrated governance mechanisms and approaches (Milder et al. 2014; Ros-Tonen et al. 2014; den Uyl and Driessen 2015). These approaches highlight integrated policy alignment, as well as horizontal and vertical coordination at different governmental levels (Sayer et al. 2013; Arts et al. 2017a).

Pursuing such integrated approaches on the landscape scale follows a set of principles which include: continual learning and adaptive management, shared values and objectives, trade-offs on differing landscape uses, awareness of various governance levels, recognition of all actors, transparency through mutual understanding amongst actors, clarity on rights and responsibilities of the actors, participatory monitoring whilst recognizing different knowledge systems, resilience and enhancing the capacity of actors to engage (Sayer et al. 2013; Ros-Tonen et al. 2014; Foli et al. 2018; Reed et al. 2019). Trade-offs emerge as a central facet for integrated governance mechanisms. With conflicting interests, goals, and priorities, trade-offs are inevitable outcomes of decision-making and mutual understanding among actors (Ros-Tonen et al. 2014; van Oosten et al. 2021a). LG principles have been adopted by international organizations (e.g. the International Union for the Conservation of Nature, World Wildlife Fund, the United Nations Environmental Programme, and the World Bank), government departments (e.g. conservation, agriculture, economic development), and transboundary conservation areas (e.g. Serengeti-Mara ecosystem shared between Tanzania and Kenya) (Freeman et al. 2015; Arts et al. 2017b; Veldhuis et al. 2019). Some of these principles such as continual learning and adaptive management, enhancing the capacity of actors to engage, and participatory monitoring, are closely related to Pahl-Wostl’s (2009) concept of ‘Learning for Governance’.

This article is driven by our interest in the governance of large-scale transfrontier landscapes, such as the Great Limpopo Transfrontier Park (GLTP) which spans South Africa, Zimbabwe, and Mozambique. The initial preparation for a research project and data collection pointed us towards a research gap. We noticed that despite the daily governance tasks in transfrontier landscapes, most of the local literature and references strongly focus on biodiversity conservation, while governance perspectives are only modestly represented. Given the TFCAs’ cross-border nature, multifaceted challenges are often experienced in issues of coordination and policy harmonization (Wolmer 2003; Bhatasara et al. 2013) to which LG principles are imperative. Likewise, LG inherently includes a wide spectrum of landscapes (intra and cross-border) encompassing natural border features such as mountain ranges, river basins, forests, wetlands, and coastal areas (Beunen and Opdam 2011; Arts et al. 2017). Valuable lessons can be drawn from TFCAs in aspects related to cross-border cooperation, collaborative wildlife conservation, human-wildlife conflict resolution, and community involvement (Wolmer 2003; Munthali 2007; Hanks and Myburgh 2015). This observation fuelled our interest with regards to the extent to which these two discourses and their related topics share commonalities. While previous studies have highlighted the principles of LG (Sayer et al. 2013; Oosten et al. 2018; Ros-Tonen et al. 2021) which also relate to the objectives of TFCA (Hanks 2003; Wolmer 2003; Kansky et al. 2021), a systemic analysis to comprehend the degree to which these academic discourses are linked or take each other into account, is still lacking. As a first step to bridge this gap, we investigate the following research questions: (i) To which degree do the two discourses share commonalities in their narratives and take each other into account? (ii) How have the discourses developed from 1998 to 2022? (iii) What is the geographical distribution of publications on the two discourses? To answer these research questions, we use a bibliometric study and a qualitative content analysis using PRISMA.

Materials and methods

This study covers the period from 1998 to 2022 and is based on the combination of a bibliometric analysis (Ellegaard and Wallin 2015; Donthu et al. 2021) and qualitative content analysis using PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analysis protocol) (Liberati et al. 2009; Moher et al. 2009; Page et al. 2021). 1998 was selected as the starting point since it marks the publication year of the first TFCA Web-of-Science (WoS) publication whereas the first LG publication dates to 2004. Whilst bibliometric studies are regularly used for the quantification of published research articles, they can also play a role in the qualitative exploration and construction of narratives (van Eck and Waltman 2010; Herrera-Franco et al. 2021). Bibliometric methods support broad, systematic, and cross-sectional analysis that illustrate the evolution of academic discourses over time and along with their conceptual backgrounds, research trends, and gaps employing both qualitative and quantitative methods (Manriquez et al. 2015; Retrouvey et al. 2020). Nonetheless, there are limitations associated with bibliometrics and network analysis methods for this purpose. The method tends to overemphasize the core themes of a cluster, while smaller niche topics only appear in the peripheries of the network analysis maps hence might not be well represented in the narratives. PRISMA on the other hand demonstrates that a systematic review is an explicit way to identify, select, and evaluate relevant research. PRISMA is a set of guidelines for systematic reviews which enable researchers to report reviews of the existing literature clearly, transparently, and with sufficient detail to enable reproducibility (Cruz-Garcia et al. 2017; Malapane et al. 2022). We selected PRISMA because previous research emphasises its comprehensiveness, improved validity, and strong reliability across reviews (Page et al. 2021). Both bibliometrics and PRISMA methods have proven useful in previous research for conducting robust and comprehensive analysis of existing literature (López-Rodríguez et al. 2022; Diwan and Amarayil Sreeraman 2023). The combination of these two methods offers a comprehensive overview of the existing literature while also providing insights into the scholarly landscape and academic networks within specific discourses. Previous researchers also highlighted the added value of combining qualitative and quantitative methods and stressed that quantitative methods alone should not be overestimated as they can sometimes overlook the ambiguity of scientific measurements (Jordan et al. 2017; Retrouvey et al. 2020).

Data collection and search criteria

For this research, WoS served as a database, based on two considerations: (i) focus on the academic discourse in high-quality publications (SCI, SSCI) (Li et al. 2018) while acknowledging the limitation that parts of the discourse might not be reflected; and (ii) WoS encompasses a wide range of scientific domains over an extensive time span (Falagas et al. 2008; Pranckutė, 2021). The search strategy involved separate searches for the two discourses using the following keywords: (i) “landscape governance” and (ii) “transfrontier conservation areas” OR “peace parks”. The term “peace parks” was also used in combination with TFCA, as the two refer to the same concept and are often used interchangeably (Aarde and Jackson 2007; Hanks and Myburgh 2015). The search strategy employed in this study was comprehensive, as the query was set to “All field”, enabling a full text search across entire articles. A filtering date range from 1998 to 2022 was also used to retrieve all article records published within this period. Furthermore, filtering was also done to select only “Articles” written in the “English” language. After the filtering process, the number of records on LG amounted to 112 records whereas the search on TFCA yielded 198 results. These records were exported as CSV (Comma-Separated Values) files which formed the basis of both bibliometric and PRISMA analysis. The bibliometric analysis was conducted in the VOSviewer software using the co-occurrence and the co-authorship technique. Since the results from the bibliometric analysis only provided limited information for constructing narratives, qualitative content analysis was conducted as a second phase to provide comprehensive results. Further analysis was done to link the publications and their respective geographic locations based on the countries where they were published. This was necessary to gain insight into the research focus and areas of expertise in each country regarding the two discourses.

Bibliometric analysis: phase one

The bibliometric analysis was conducted using the LG (112) and TFCA (198) record files which were downloaded as CSVs. For this phase, the VOSviewer software (van Eck and Waltman 2010) was used to perform the bibliometric analysis using the co-occurrence and co-authorship technique. Co-occurrence technique is an analysis method embedded within the VOSviewer software used to analyse and visualize the patterns of co-occurrence of keywords within a dataset. The method is used to identify the relationships amongst keywords, providing insights into the thematic structure and interconnections within a particular study field. This analysis was used to determine (a) associations between keywords within the two discourses, and (b) the emerging keywords to emphasize the recent developments and emerging topics within our research. The ‘minimum number of occurrences of a keyword’ was set to four for the analysis. This resulted in a co-occurrence network map (van Eck and Waltman 2010). The analysis for this map was based on network morphology, total link strength, node proximity, and cluster density (Table 1). The ranking number assigned to each cluster in the keyword co-occurrence map was based on the size of the cluster, which is determined by the number of keywords assigned to that particular cluster (Fig. 1).

Table 1 Definition of terms in Bibliometric analysis
Full size table
Fig. 1
figure 1

The flowchart of the bibliometric analysis which served as the first phase of the literature review. Phase one provided the central keywords for qualitative content analysis from the keyword co-occurrence results

Full size image

We also conducted an overlay visualisation on the co-occurrence network map to determine the emerging keywords. This was done to capture significant developments and shifts in the research focus within the LG and TFCA discourses. The second analysis within phase one was conducted using the co-authorship technique, which was also done in VOSviewer using the downloaded records for LG (112) and TFCA (198). This analysis identifies author connections based on their affiliations in publications. One of the features of this analysis is the ability to identify the countries associated with an article through examining the affiliations of the co-authors. In this analysis, the focus was to identify the geographic distribution of the two discourses. In this case, geographic distribution refers to the distribution of articles based on location of affiliation, production, or publication which provides insights into the geographic representation and contribution of different countries. A ‘minimum number of documents per country’ was set to one in order to retrieve all countries which have published on the topic. This analysis also complimented the co-occurrence analysis to reveal locations in which the discourses are being frequently investigated.

PRISMA: phase two

In the second phase of the analysis process, WoS was also used to identify the article records associated with the two discourses LG and TFCA. The initial results amounted to 122 records for LG and 214 for TFCA. From the identified records, we filtered the records based on the “English”, and “Articles” query and removed the duplicates. This resulted in 112 records for LG and 198 records for TFCA. For this phase, the records were downloaded in Excel format and combined into one Excel sheet for ease of manipulation, resulting in a total of 310 records. The 310 records were the same article records used in the bibliometric analysis which generated the co-occurrence results from which we selected the top five keywords based on the total link strength (TLS) and node size (Fig. 2).

Fig. 2
figure 2

Qualitative content analysis flowchart based on the PRISMA (Moher et al. 2009). Keywords from the bibliometric analysis were used for filtering articles for each cluster

Full size image

The top five keywords per cluster were then used as filtering keywords to retrieve articles for qualitative content analysis. These keywords were incorporated simultaneously using the Boolean operator “OR” to retrieve the articles from the pool of 310 records. Further filtering was done to refine the dataset to a viable number for content analysis. For each cluster, the following scheme was used to identify the articles for content analysis: (i) three most highly cited articles, (ii) three most recently published articles (iii) three oldest articles, and (iv) three of the recent and highly cited articles in the last five years (2017–2022). This process resulted in a total of 72 articles that were included for qualitative content analysis. A qualitative content analysis which informed the narrative of the clusters was then carried out as part of the qualitative synthesis process based on Kuckartz (2019). The titles for each cluster reflect the core concepts of each cluster resulting from the content analysis.

Results of the bibliometric and qualitative content analysis

Since 1998, we can see a continuous increase in publications in the discourses of LG and TFCA (Fig. 3). Whilst LG emerged in publications only in 2004, some of the LG principles which include enhancement of actors’ capacity to engage, awareness of various governance levels, and recognition of all actors had already been addressed earlier in the TFCA discourses. The results provide insights into the temporal and geographical focus of publication activity in both academic discourses. For LG, 90% of all articles have been published between 2013 and 2022, whilst for TFCA it is still more than 70%. The number of TFCA publications have been consistently higher than the number of LG publications throughout the years since 2004 with a maximum peak of 21 in 2020, despite some fluctuation in this trend.

Fig. 3
figure 3

The figure shows row (i) the publication output for the two academic discourses from the time of initial publication of the TFCA in 1998, and the LG in 2004, (ii) Publication distribution and ranking by country for LG (a) and for TFCA (b), and (iii) the top five most occurring keywords in LG (c) and TFCA (d). The Total Link Strength (TLS) represents the cumulative strength or intensity of the links between keywords in the co-occurrence network within the bibliometrics software (VOSviewer)

Full size image

With regards to publications by location, the geographic distribution of the LG discourse is primarily rooted in the global North making 80% of the top 10 list globally. The Netherlands is by far the most publishing country in the LG discourse with a total of 34 publications. Other countries in the global North which have significant publications on this discourse in the global North include the United States of America (15), Australia (13), Germany (13), Canada (10) England (10), Sweden (9), and France (7). The LG debate is also well established in countries like Indonesia (18) and Kenya (11), which show a substantial number of publications despite not being geographically located in the global North. In contrast, 70% of the articles published on the TFCA topic are from the African continent, with South Africa (108) having the highest number of publications. The top African publishing countries in the list are specifically Southern African countries which include Zimbabwe (30), Botswana (25), Mozambique (11), Namibia (10), and Zambia (7). Despite the large volumes of publications in the TFCA discourse, LG has a broad spectrum of publishing countries (49) with each having at least 1 publication whilst TFCA has only 29 publishing countries.

Co-occurrence analysis

The co-occurrence analysis illustrates a distinct grouping of keywords emanating from both the TFCA and LG discourses (Fig. 4). The cluster ranking is based on the number of keywords assigned within that particular cluster, with the allocations as follows: cluster one (31), cluster two (23), cluster 3 (20), cluster 4 (19), cluster 5 (14) and cluster 6 (13). Clusters one (African Wildlife Conservation) and three (TFCA Wildlife Connectivity) are centred around topics close to TFCA whilst clusters two (Governance for Biodiversity Conservation) and four (Policies and Strategies) are strongly anchored in LG. While the first clusters are clearly delineated from each other, clusters five (Political Ecology) and six (Management of Protected Areas) link more strongly to both discourses by addressing the importance of different scales and actors (co-management/co-governance) and the role of policy. Figure 4 illustrates six delineated clusters followed by a detailed narrative from a systematic review of the literature (Table 2).

Fig. 4
figure 4

Keyword Co-occurrence map for LG and TFCA. The bigger the node the larger the weight and the closer the node the stronger the connection

Full size image
Table 2 Overview of key challenges and conceptual underpinning of each cluster. Conceptual underpinning articles directly address the central concepts or theories relevant to the cluster. Core articles present the narrative for each cluster and are based on high citations
Full size table

Cluster one: African wildlife conservation

Cluster one thematically focuses on wildlife biodiversity in African conservation areas revealing the intricate challenges, strategies, and triumphs that define conservation efforts across the continent. While the cluster provides context on the entire Africa, the main emphasis is geared towards Southern Africa (Michel et al. 2006; Loarie et al. 2009; Martin et al. 2011; Büscher and Ramutsindela 2015). The cluster underscores the significance of national parks and TFCAs, highlighting their significance for conservation and establishing connectivity through wildlife corridors. (Ros-Tonen et al. 2018; van Oosten et al. 2018). While corridors serve different purposes (e.g., wildlife migration, landscape connectivity) they play a vital role for wildlife to cope with climate change impacts (e.g., water availability, drought). The cluster pinpoints some of the key regions of biodiversity concentration such as the Okavango Delta and the GLTFCA. Cluster one also indicates how the success of wildlife conservation in Africa has resulted in the abundance of large herbivores that require large habitats and spatial resources (van Aarde and Jackson 2007; Gaughan et al. 2019). Increased landscape fragmentation, and the abundance of wildlife have resulted in increased human-wildlife conflicts (HWC) (Loarie et al. 2009; Kansky et al. 2021). Predators (Meer et al. 2016; Cushman et al. 2018; Dures et al. 2019) and large herbivores (e.g., elephants) (Loarie et al. 2009; Tshipa et al. 2017; Kansky et al. 2021), are increasingly roaming into settlements and agricultural land (discourses, pastures), exacerbating these conflicts. On the other hand, increased human activities such as farming and settlement development (due to population growth) and poaching (Büscher and Ramutsindela 2015; Ntuli et al. 2021) challenge the survival of several key species (e.g. African savannah elephant and large carnivores) (Loarie et al. 2009; Loveridge et al. 2022; Searle et al. 2020). These human activities increase the vulnerability of wildlife and threaten wildlife populations due to habitat loss, fragmentation, and restricted movement (Loarie et al. 2009; Young and Van Aarde 2010). Key species, such as lions (Panthera leo) and elephants are particularly vulnerable to habitat fragmentation (e.g., due to fences) and poaching, which results in lower levels of genetic diversity (Dures et al. 2019; Loveridge et al. 2022; Searle et al. 2022). Therefore, conservation areas in Africa are essential to address these problems (Ramutsindela 2017; Cushman et al. 2018).

Cluster two: governance for biodiversity conservation

Cluster two illustrates the role of governance in biodiversity conservation and facilitating agricultural activities It strongly illustrates the multi-actor, multi-tier, and multi-level characteristics of governing landscapes on the landscape scale (Dawson and Martin 2015; Elbakidze et al. 2010; Ros-Tonen et al. 2015; Visseren-Hamakers 2015). The growing demand for food and natural resources (e.g., rural land) has become a major driver for landscape change, affecting ecosystem services, rural land use, and livelihoods (Foli et al. 2018; Ros-Tonen et al. 2021). Approaches such as the integrated landscape approach have become instrumental in addressing these challenges in LG (Sayer et al. 2015; Arts et al. 2017; Angelstam et al. 2019). The integrated landscape approach is based on previous efforts to reconcile conservation with agriculture, forestry, and other land uses (Pedroza-Arceo et al. 2022; Reed et al. 2020; Ros-Tonen et al. 2021). The need to balance multiple objectives, engage relevant actors equally and address power as well as gender imbalances across sectoral, jurisdictional, and administrative silos has resulted in the evolution of LG (Elbakidze et al. 2010; Kusters et al. 2018; Mugo et al. 2020). LG challenges underscore the strong need for (vertical and horizontal) integrated policies, since sectoral approaches have failed to adequately address so-called ‘wicked’ policy problems (such as biodiversity conservation, often ignoring the livelihood needs of the poor), and the need to address inequalities and trade-offs among competing and/or conflicting actor interest (e.g. biodiversity conservation, hunting, settlements agriculture, mining) (Foli et al. 2018; Ros-Tonen et al. 2018; Riggs et al. 2021). Consequently, different actors are expected to collaborate, co-produce novel knowledge and integrate different knowledge types (Elbakidze et al. 2010; Pedroza-Arceo et al. 2022). It shows that local actors have a wide range of tacit knowledge on ecosystem dynamics and management practices (Elbakidze et al. 2010; Ros-Tonen et al. 2015). Integrating different knowledge types is considered a complex process due to three main reasons (i) questioning of local knowledge by incumbent knowledge types like policy knowledge or scientific knowledge (ii) challenge of articulation and comprehension of tacit knowledge which is often not documented (iii) diverse perception of knowledge from indigenous groups with different, assumptions, rules and starting points (Dawson and Martin 2015; Elbakidze et al. 2010; Reed et al. 2019; Ros-Tonen et al. 2015). Hence, landscapes and communities require adaptive capacity to respond and adapt to changing conditions (Folke et al. 2005; Ros-Tonen et al. 2014).

Cluster three: TFCA wildlife connectivity

Cluster three addresses the role of TFCAs in various aspects of connectivity relating to habitats and ecosystems (Roever et al. 2013; Elliot et al. 2014; Sulistyawan et al. 2019; Omoding et al. 2020). Wildlife corridors are considered essential for ecological connectivity across borders in TFCAs. While connectivity is essential, the co-existence of humans and wildlife in TFCAs has resulted in cases pertaining to zoonotic diseases such as Brucellosis (Gomo et al. 2012), Bovine Tuberculosis (Caron et al. 2016) and Foot and Mouth Disease (Lazarus et al. 2021) which threaten the survival of species. Human interference is considered a contributing factor to landscape fragmentation which in turn disrupts connectivity while destroying wildlife habitats (Naidoo et al. 2018; Lines et al. 2021a, b). Human population growth further causes fragmentation through the establishment of more settlements and agricultural land (Elliot et al. 2014; Bradshaw and Leonard 2020; Petracca et al. 2020) raising concerns over potential negative biodiversity impacts (Roever et al. 2013; Sinthumule 2017). The degree of range loss, fragmentation, and dispersal amongst species emphasize the importance of corridors in maintaining and re-establishing connectivity to protect wildlife (Roever et al. 2013; Elliot et al. 2014). TFCA governance is constrained by inter-state differences, power imbalances, and institutional tensions which affect institutional alignment across legal systems to achieve wildlife connectivity (van Amerom and Büscher 2005). Property rights arrangements further underpin power relations and determine the possibilities for establishing TFCAs (Lenggenhager and Ramutsindela 2021). Well-defined and recognized property rights provide a negotiation platform for TFCA establishment whilst unclear property rights create challenges, resistance, and disputes among actors, constraining the establishment of TFCAs (Lenggenhager and Ramutsindela 2021; Sjöstedt and Linell 2021). While state-owned land has been instrumental in the establishment of TFCAs, private land presents legal and financial constraints which make it difficult to negotiate in TFCA initiatives to enable wildlife connectivity (Sinthumule 2017; Lenggenhager and Ramutsindela 2021).

Cluster four: policies and strategies

Cluster four illustrates a strong institutional perspective on conservation that promotes the relevant role of policies and partnerships on landscape levels. The quest to foster integrated management of landscapes has sparked a growing interest in strategies like ecosystem-based management and eco-regional conservation planning (Olsson et al. 2004; Arts et al. 2017; Robinson and Kagombe 2018). They promote landscape resilience through afforestation, conserving indigenous forests, reducing deforestation, and improving access to clean energy (McNeely 2003; Murungweni et al. 2011; Kark et al. 2015). For policy and strategy development, different forms of partnerships between multiple actors are considered important (Olsson et al. 2004; Görg 2007). Partnerships establish interactions amongst actors from different sectors such as agriculture, forests, biodiversity, natural resource conservation to achieve multifunctional landscapes which provide food security, livelihood opportunities, and other ecosystem services such as carbon sequestration and cultural services (McNeely 2003; Plieninger and Bieling 2012; Arts et al. 2017; Kalvelage et al. 2021; van Oosten et al. 2021). In both LG and TFCAs, those partnerships must permeate political boundaries to establish and institutionalize the collaboration of actors spanning political borders if conservation approaches are to be effective (Kark et al. 2015; Njoroge et al. 2020). In TFCAs, partnerships are crucial for the conservation of flagship species (e.g., rhinos, elephants, leopards, lions, and buffalos) that require large areas for migration and also to secure genetic diversity (Lunstrum 2015; Purdon et al. 2018). The co-existence of humans and wildlife is an important concern in landscape management. Planners and conservation practitioners are increasingly faced with the challenge of reconciling past mistakes and developing governance and conservation strategies that respect different bundles of property rights of local communities. (Chirozva et al. 2013; Pricope et al. 2020). At this point, formal plans, strategies, and institutions must also meet informal norms, rules, and customs and form area-based co-management arrangements, that deliver the required integration. They are based on sharing of power and responsibilities between the government, communities, and local resource users, which is essential for joint learning and knowledge co-creation to sustainably manage the shared resources (Olsson et al. 2004; Kark et al. 2015; Portman and Teff-Seker 2017).

Cluster five: political ecology

Cluster five, ‘Political Ecology’, addresses power asymmetry, scale politics, decision making, and decentralised institutions (Görg 2007; Chiutsi and Saarinen 2017; Njoroge et al. 2020), as core characteristics. It focuses on shaping relationships and acknowledges the role of power relations condition in the creation, management, and governance of nature by different actors. (Büscher 2013; Buizer et al. 2016). Hence it recognises the political dimension as a formative element of governance arrangements and decision making (Görg 2007; Ide 2019; Linell et al. 2019; Movik et al. 2021). In this narrative, particular attention is paid to how the local actors are overlooked in instances of decision making and governance related issues (Castella et al. 2014; van Oosten et al. 2018). The socio-political relevance of local actors has become widely accepted and requires hybrid or polycentric institutional arrangements. There is a growing demand for authentic involvement of local and indigenous communities in decision making processes. This demand has strengthened their calls for official recognition of their territories and their rights to land and resources (Görg 2007; Plieninger et al. 2015; Chiutsi and Saarinen 2017; Movik et al. 2021). This is of particular importance in transboundary settings, that tie together different countries, institutional systems, legal instruments, and land-tenure systems (van Amerom and Büscher 2005; Lejano 2006; Munthali 2007). These transboundary settings can cause power struggles and/or power imbalances within and between different spatial and administrative scales and social entities/or communities (Lunstrum 2014; Sinthumule 2017). Hence, there is a need for continual improvement and development of novel transboundary governance approaches that support mutable, non-territorial structures and relationships among involved actors, rights- and title holders (Duffy 2006). However, a strand of literature in this cluster illustrates these challenges in the context of Southern African countries particularly those in the Southern African Development Community (SADC) region (Chiutsi and Saarinen 2017; Sinthumule 2017; Njoroge et al. 2020). It also focuses on the notion that TFCAs provide integrated management for regions that have encountered political conflicts and still experience tensions, whilst cooperating in transboundary conservation actions such as re-establishing migration routes (Purdon et al. 2018; Meyer et al. 2022). TFCAs are considered as frameworks that have been implemented to foster integration of biodiversity conservation and rural development to alleviate poverty which results in peaceful co-existence between countries (Milgroom et al. 2014; Ide 2019).

Cluster six: management of protected areas

Cluster six, ‘Management of Protected Areas’ revolves around ecosystem-based management (Robinson and Kagombe 2018), adaptive management (Riggs et al. 2021; Gonçalves and Pinho 2022), and the integrated management of natural resources and/or landscapes (Nzyoka et al. 2021; Ros-Tonen et al. 2021; Amaruzaman et al. 2022). It emphasises that integrated management approaches are essential for balancing environmental demands, land-use options, and socio-economic pressures (Acheampong et al. 2016) and that they contribute to conflict resolution (Amaruzaman et al. 2022). Integrated management strategies are considered to counteract habitat and ecosystem fragmentation and support the conservation of key species such as leopards and lions (Curveira-Santos et al. 2021; Rogan et al. 2022; Sultan et al. 2022) as well as key migratory species like elephants (van Aarde and Jackson 2007; Gross et al. 2022). Increasing pressure from developments like housing or infrastructure around protected areas such as TFCAs further strain human-wildlife relations in developing nations (Selier et al. 2016; Mpakairi et al. 2019). In addition to pursuing conscious land-use, knowledge of migratory routes and migratory connectivity are essential for developing and implementing governance strategies (Lines et al. 2021b; Tshipa et al. 2017). In TFCAs and various protected areas, one specific governance approach which increases tolerance to wildlife is community-based natural resource management which provides more sustainable mechanisms to live with wildlife (Kalvelage et al. 2021; Meyer et al. 2022). However, the possible establishment of such an approach hinge on the successful negotiations with property rights holders (Lenggenhager and Ramutsindela 2021).

Changes in the discourses over time

The analysis illustrates how the discourses have been changing over time based on the emerging topics within the publication timeline (Fig. 5). Both examined research discourses are strongly rooted in the topic of ‘conservation’ which is at the centre of the co-occurrence map. The oldest keywords within these two discourses address topics such as ‘Peace Park’, ‘Great Limpopo Transfrontier’, ‘Mozambique’, ‘Kruger-National Park’, ‘South Africa’, ‘Zimbabwe’, ‘War’, ‘Natural resource management’, ‘Disease’ and ‘Space’ in clusters one and three. The map shows that these topics emerged between 1998 and 2005. Between 2005 and 2015, a number of newly emerging topics gained significant influence as indicated by their node size and total link strength. Some noteworthy keywords in this period include, ‘landscape governance’, ‘governance’, ‘management’, ‘landscape’, ‘policy’, ‘politics’, ‘landscape’, and others. The analysis also shows that a large proportion of the latest emerging topics post-2015 are attributed to cluster two (Governance for Biodiversity Conservation) which is mainly rooted in the LG discourse. This cluster has the second largest number of keywords (23) assigned to it with eight of the keywords emerging as recent topics. These include ‘ecosystem services’, ‘initiatives’, ‘landscape sustainability’, ‘environmental issues’, ‘collaboration’, ‘redd-plus’, ‘Indonesia’, and ‘sustainable development’. The topic ‘ecosystem services’ had the highest level of occurrence as an emerging keyword within cluster two. In the TFCA discourse, some of the recent emerging topics post-2015 include: ‘Botswana’, ‘impacts’, ‘populations’, ‘benefits’, ‘natural-resource management’, ‘risk’, ‘extinction’, and ‘dispersal’. These topics within TFCA are only encompassed in clusters one (African Wildlife Conservation) and three (TFCA Wildlife Connectivity).

Fig. 5
figure 5

The diagram shows an overlay visualisation of keywords for both LG and TFCA discourses from 1998 to 2022

Full size image

Discussion

The discourses of LG and TFCA represent two distinct yet interconnected approaches to addressing conservation and landscape management. While both aim to foster sustainable land use and conservation, they exhibit both commonalities and differences in their conceptual frameworks, objectives, and practical implementations.

Commonalities and differences between the discourses represented in the clusters

Results show that the key commonalities between LG and TFCA discourses are (i) conservation, (ii) power, and (iii) actor roles. Both discourses underscore the essence of conservation which is central to TFCA and LG objectives relevant in clusters one, two, three, and four (Mugo et al. 2020; Ros-Tonen et al. 2021; van Oosten et al. 2021). Qualitative content analysis indicates that clusters two (Governance for Biodiversity Conservation), and four (Policies and Strategies) which lean towards LG share commonalities with clusters one (African Wildlife Conservation) and three (TFCA Wildlife Connectivity) which lean towards TFCA. The conservation aspect within clusters two and four pertains to the sustainable management of natural resources and ecosystems within various regions and landscapes (urban, rural, agriculture) (Ros-Tonen et al. 2014; Buizer et al. 2016; Carta et al. 2022). This is because the conservation aspects in LG are not constrained to one specific landscape but include a wide range with focus on resources such as water, forests, and biospheres (Robinson and Kagombe 2018; Hedden-Dunkhorst and Schmitt 2020; Best et al. 2021). In the TFCA discourse, which is mainly reflected in clusters one and three, conservation is expressed in transboundary contexts between multiple countries to protect biodiversity and promote sustainable development (Sinthumule 2017; Lenggenhager and Ramutsindela 2021). Conservation efforts in TFCA and LG however require navigating intricate power dynamics among diverse actors. Power is a common topic that is explored in discourses around TFCA and LG, represented in cluster five (Görg 2007; Ros-Tonen et al. 2018; Mugo et al. 2020). Articles highlighting power struggles in TFCAs are mainly related to two aspects: the distribution of property rights (Sinthumule 2017; Lenggenhager and Ramutsindela 2021) and on the other hand, the colonial history and legacies which play a significant role in shaping previous and current power dynamics within TFCAs (Sibanda 2015; Chiutsi and Saarinen 2017; Bourgeois et al. 2023). In LG discourses ‘power’ is strongly connected to aspects of actor-constellations and their roles and capacities in and resources for agenda-setting and decision-making processes (Lazdinis et al. 2019; Mugo et al. 2020; Amaruzaman et al. 2022). These diverse actor constellations play a crucial part in navigating competing and conflicting interests (Robinson and Kagombe 2018; Reed et al. 2019). The analysis shows that the roles of actors are well-established in both discourses: clusters one, two, three, and five acknowledge the relevance of involving different actor groups, indigenous communities in LG and/or TFCA management (Ros-Tonen et al. 2014; Langston et al. 2017; Best et al. 2021). They also hold specific knowledge types and expertise, relevant to ‘re-solve’ wicked problems (addressed in cluster two) and contribute ideas and strategies for sustainable development (Robinson and Kagombe 2018; Reed et al. 2020; Best et al. 2021). This ultimately leads to sustainable outcomes that consider the diverse array of factors at play within different landscape contexts. 2021).

Results also revealed disparities between the two discourses. Despite LG’s acknowledgment of the importance of collaboration and coordination amongst diverse actors (Opdam et al. 2016; Njoroge et al. 2020) the discourse falls short in terms of transboundary settings. While LG has found application across various contexts which include forests, agriculture, urban areas, and water, TFCAs have been notably absent from these discussions. This is related to the geographic focus (see Fig. 3, e.g., Australia, Europe, North America), where transboundary governance settings are or have been a lesser issue yet (e.g., Europe: with few numbers of transboundary nature conservation areas, particularly not on the scale of those in Africa). Given the increasing numbers of migratory large carnivores which traverse various protected areas (on national scale) particularly in Europe (e.g., Wolves, Brown Bear, Eurasian and Iberian lynx) and the associated wildlife-human conflicts (e.g., conflicts on livestock predation, tourism) (Ordiz et al. 2013; Bautista et al. 2019; Martínez-Abraín et al. 2023), European discourses can benefit from long-standing practise experience and academic knowledge from African large-scale protected areas. The increased call for rewilding approaches in Europe and North America (Trouwborst et al. 2017; Thulin and Röcklinsberg 2020; Papp et al. 2022), points to the relevance of learning from the TFCA literature and better linking these discourses. The TFCA-related discourses can be beneficial, emphasizing the need for addressing challenges arising from cross-country and cross-institutional collaborations and dealing with human-wildlife-coexistence in formerly human-dominated landscapes (Dressler and Büscher 2008; Chitakira et al. 2012; Muboko 2017).

Cluster two and four which are more inclined towards LG exhibit a strong governance focus. The qualitative content analysis highlights concepts such as adaptive governance, knowledge co-creation, and collaborative governance (Elbakidze et al. 2010; Dawson and Martin 2015; Ros-Tonen et al. 2015; Reed et al. 2019). Considering the diverse and intricate scenarios in which LG has been employed, effective governance approaches become imperative for managing conflicting interests within these landscapes (Ros-Tonen et al. 2018; van Oosten et al. 2021). Contrary to this discovery, clusters one (African Wildlife Conservation) and three (TFCA Wildlife Connectivity) are embedded in discourses strongly rooted in plant and wildlife ecology. Hence our results indicate a potential blind spot of LG on these issues. Considering the prevalence of governance-related challenges in TFCAs emanating from poor coordination and incoherent policies, LG can be essential in addressing these challenges due to its strong focus on (horizontal and vertical) coordination (Beunen and Opdam 2011; van Oosten et al. 2021). Conversely, TFCA can be essential in addressing challenges emanating from wildlife or cross-border (intra and international) issues adequately. Therefore, there is need for cross-fertilization of these two discourses to fully leverage the strategies, principles, and models that either of the discourses offer.

From conservation-centric to holistic approaches

Over the past years, topics within LG and TFCA have evolved increasing the scope, focus, and relevance of the discourses. The results from our visual overlay (Fig. 5) show that the discourses have evolved from a conservation-centred approach (Barnard et al. 1998; Westing 1998) to more integrated and holistic approaches in conservation (Chiutsi and Saarinen 2017; Sari et al. 2019; Best et al. 2021). This acknowledges the intertwined relationships between people and the environment (Chiutsi and Saarinen 2017; Sari et al. 2019; Best et al. 2021). Our findings correspond with previous work, such as Mace (2014) emphasizing that conservation incrementally shifted from ‘nature itself’ towards ‘people and nature’ (e.g., expressed in cluster four) thus illustrating a conceptual reframing of nature conservation towards more contemporary understandings of human-nature relationships (Flint et al. 2013; Mace 2014; Braito et al. 2017). The integration of the ‘human’ dimension is further expressed by the reoccurring themes such as ‘initiatives’, ‘collaboration’, ‘sustainable development goals’, and ‘landscape sustainability’. They demonstrate the evolution towards “people and nature” emphasizing more recent themes such as the links between human well-being and nature conservation (Ros-Tonen et al. 2015; Westerink et al. 2017). This evolution towards a more holistic approach has resulted in numerous practical applications (e.g., integrated land-use planning; community engagement and participation; ecosystem-based management) which entail the design and implementation of more inclusive conservation strategies (Visseren-Hamakers 2015; Gaughan et al. 2019; Chitakira et al. 2022). Furthermore, the evolution to a more holistic approach has driven the adoption of LG frameworks that facilitate the interplay between conservation and human activities (Flint et al. 2013; Braito et al. 2017; Sari et al. 2019; Best et al. 2021).

LG prominence in global North, TFCA in Southern Africa

Geographic location also plays a significant role in the publication output in both discourses. The publication distribution from the identified clusters shows that LG debates are strongly rooted in the global North. This is also due to robust governance, institutional structures, and well-established land use planning systems embedded in the spatial planning traditions of these countries (Plieninger et al. 2015; Wallner et al. 2017). Additionally, conventions in the global North such as the European Landscape Convention, played a significant role in advancing the concept of LG and human-nature relatedness (Pătru-Stupariu and Nita 2022). On the other hand, TFCA discourses reflect a Southern African-centred debate due to the region’s higher concentration of TFCAs compared to any other region globally. Hence Southern Africa has been at the forefront of transboundary conservation initiatives, emphasizing cross-border cooperation and collaboration (Duffy 2006; Chiutsi and Saarinen 2017). The difference and prominence of the LG debate in the global North and the TFCAs debate in Southern Africa are a result of response to the unique regional challenges, priorities, and approaches to landscape management and conservation. Whilst both academic discourses share common goals of sustainability and collaboration, their specific emphasis and strategies vary based on the regional context and the unique characteristics of the landscapes involved. With the growing concern of governance challenges within TFCAs in Southern Africa (Bhatasara et al. 2013; Chitakira et al. 2022) and the increasing wildlife population in the global North particularly carnivores (Trouwborst 2015), policymakers are calling for the establishment of transboundary conservation areas hence there is need for exchange of knowledge across these specific regions (Trouwborst et al. 2017; Papp et al. 2022). Principles developed in the context of TFCAs, which primarily focus on cross-border conservation across national boundaries, can be expanded beyond the cross-border context. LG can benefit from adopting these principles to address challenges arising in different landscapes where natural features serve as borders across a range of scenarios such as international, regional, or local. This cross-fertilization of knowledge, strategies, and practices can potentially enhance the efficacy of conservation and landscape management at the conceptual and practical levels.

Conclusion

The present paper investigates the narratives in the LG and TFCA discourses, and their commonalities and differences. The meta-analysis of 72 research articles resulted in six distinct narratives that are fuelled by debates on nature conservation, wildlife, governance, and integrated management. While the TFCA discourse has strong roots in conservation and ecological perspectives, it also acknowledges challenges such as human-wildlife relationships and conflicts. These topics are crucial for achieving the core objectives of TFCAs, particularly in relation to biodiversity, ecosystem management, and long-term conservation. However, they have so far fallen short of the actor and governance-oriented objectives of TFCAs, such as adaptive governance, learning, and knowledge co-creation. LG scholars have created robust knowledge and expertise on precisely these topics, which makes it sensible to bridge these discourses. At the same time, this bridging requires caution and sensitivity to whether their context is a good fit and suitable in other socio-cultural contexts, avoiding stereotype North–South ‘lecturing’.

LG has a strong foundation in the scholarly debates in the global North and should therefore not be adopted on a par-for-par basis in other socio-cultural and institutional contexts. In addition, LG has so far paid very modest to almost no attention to issues of transboundary and transfrontier landscapes, which we have identified as a blind spot in the LG discourse. On the other hand, the TFCA scholarship provides extensive knowledge and academically reflects experiences on human-wildlife conflicts, co-existence, and transboundary cooperation. These key issues are highly topical and relevant, at least in the current European context, where policy and conservation practice are confronted with increasing human-wildlife conflicts and require re-establishing the co-existence of humans with carnivores. This not only concerns issues such as livestock/herd protection, nature conservation, and tourism but also the management of (previously) human-dominated landscapes (e.g. (abandoned) alpine pasture systems). Increased efforts in rewilding in the European context make this South-North knowledge transfer particularly relevant. While this study specifically focused on LG and TFCA the results suggest the incremental integration of different approaches and scholarly debates. The results show that more recent topics focus on knowledge co-creation, indigenous knowledge, social equity, ecosystem services, learning, and climate change adaptation. Hence, other scholarly debates and conceptual approaches, such as environmental governance (Wilson 2019; Ruan et al. 2022), rewilding (Root-Bernstein et al. 2018) or transdisciplinarity (Gugerell et al. 2023), will become more prominent in the future. A potential next step would be to explore potential niches for integration of either of the discourses in practice as case studies.