Farmer A: “At one point, I lost hope and wilfully decided to stop cocoa farming.”

Farmer B: When I asked [Farmer A] why she had abandoned her cocoa farm, she pointed out cocoa farming was no longer profitable. With rising climate shocks, she was losing the money, labor, and time trying to rehabilitate her cocoa farm.”

NGO official X: ‘When I first met [Farmer A], she was desperate about her cocoa farming because the farm was not producing as much as she expected. She is now considered an interesting agroforestry champion … many of her peers now wonder why her farm has undergone such a tremendous change.’

NGO staff Y: ‘Agroforestry makes a difference in cocoa farms. The farms are rejuvenating and now giving the desired yield that the farmers really want to see.’

Farmer A: “If you are a farmer planning to quit farming cocoa, my advice is that they do not quit because it is highly profitable. I am happy to be a cocoa farmer and an Agroforestry Champion.” (Author emphasis)Footnote 1

Introduction: The global cocoa economy at the crossroads

How is cocoa agroforestry implemented under various initiatives in rural Ghana, and whose priorities do they serve and why? Navigating this question is essential for debating the future of cocoa in Ghana, where NGOs, cocoa companies, and the government are using agroforestry as a strategy to encourage farmers to remain in cocoa, as demonstrated by the introductory dialogue from a recent promotional video on cocoa agroforestry. Ghana’s recent agricultural survey suggests the country’s cocoa households may have declined from the widely recognized 800,000 in the 2000s to about 620,000 in 2017/2018 (GSS 2020, p. 145). Many factors contribute to the growing displacement of farmers from cocoa. This includes climate variability, pests and diseases, poverty, and land tenure (Amanor et al. 2021; Amanor 2010; Wessel and Quist-Wessel 2015). Illegal artisanal and small-scale mining of gold is another major issue driving farmers away from cocoa, with farmers coerced to sell their cocoa farms to miners for bulk cash (Nunoo et al. 2023; Siaw et al. 2023).

Many of the factors driving farmers away from cocoa are complex and interrelated. Addressing these challenges requires cross-cutting strategies. However, as demonstrated in the opening extract, many cocoa sector actors are using cocoa agroforestry to entice farmers—who have stopped or are considering divesting from cocoa altogether—and influence rural land-use choices profoundly. Arguably, these actors, including NGOs, cocoa and chocolate companies, and the government, do this to reinforce cocoa’s essential contribution to the global chocolate economy. Nevertheless, amid growing concerns about poverty, declining well-being among cocoa farmers (Amfo et al. 2020, 2021), and the mounting environmental impacts of cocoa farming systems (Ruf et al. 2015), tensions emerge between some sector actors’ use of cocoa agroforestry as an emplacement strategy and genuinely addressing the pressing needs and concerns of cocoa farmers.

I examine some of these tensions by questioning the theory and practice of cocoa agroforestry in Ghana. I do this by deconstructing cocoa agroforestry and examining equity in how stakeholders apply the concept and whose priorities it serves. The unit of my analysis is Ghana’s Juabeso/Bia Landscape (JBL), selected because of its history of cocoa production and prioritization as a Hotspot Intervention Area (HIA) by the Ghanaian government, NGOs, and cocoa and chocolate companies (MLNR et al. 2023, p. 21). Field observations of various cocoa agroforestry systems and analysis of focus groups and interviews involving cocoa farmers, NGOs, and public officials lead me to argue that cocoa agroforestry, as implemented in the JBL, seeks to primarily improve cocoa yield and protect the economic concerns of high-end cocoa sector actors. Current forms of cocoa agroforestry in the JBL do not adequately reflect on or address the primary concerns of cocoa farmers at the grassroots. Although such concerns are manifold—and include poor farmgate prices for cocoa, pests, disease, a declining labor force, and the rise of mining in cocoa farms—the issue of food insecurity is paramount in the landscape. Yet, unfortunately, cocoa agroforestry, as currently practiced, reinforces food insecurity in the landscape. A sustainable solution for farmers’ displacement from cocoa cannot be achieved by depending on cocoa agroforestry as currently practiced. Alternatively, it requires questioning why farmers are being displaced from cocoa and on what terms. This shift in understanding enables us to think about cocoa agroforestry differently, particularly what is prioritized, by whom, and to what end.

I recognize that cocoa agroforestry may mean different things depending on how its components are arranged across space and time. Thus, following this introduction, in “Analytical framework” section, I delve into the concept and develop a tripartite analytical framework (Andres et al. 2016; Gama-Rodrigues et al. 2021; Ntawuruhunga et al. 2023; Sonwa et al. 2019): (1) basic agroforestry, (2) intermediate agroforestry, iii) dynamic agroforestry, to unpack how different interests may be prioritized depending on various spatial and temporal configurations and the outcomes envisaged such as food, feed, fuel, and fodder. I then apply the framework to existing cocoa agroforestry practices in the JBL in “Results” section. The final part of the paper, “Discussion” section, situates the findings in the broader literature on sustainable cocoa transition. I argue that cocoa agroforestry cannot be taken as given. On the contrary, it is an analytical category to be debated and negotiated depending on what and whose needs matters the most.

Analytical framework

According to the World Agroforestry Centre (WAC) (2013 p.7), agroforestry is a ‘natural resource management system that, through the integration of trees on farms and in the agricultural landscape, diversifies and sustains production and builds social institutions.’ A synthesis of the literature on cocoa agroforestry demonstrates it covers a continuum of technologies and practices that vary based on the level of complexity of their components and interactions (Table 1). This malleable characterization raises three points essential for unpacking cocoa agroforestry: what qualifies as cocoa agroforestry, who defines it, and for whom.

Table 1 A summary of the forms of cocoa agroforestry
Full size table

Starting with what qualifies as cocoa agroforestry, two points are worth unpacking to guide the analysis. The first and most basic is the character of cocoa: Is cocoa a tree, a crop, or both? As a woody perennial with a main stem, supporting branches, and leaves, cocoa is technically a tree. However, it is also a crop, particularly in its adopted home of West and Central Africa, including Ghana, where it is cultivated mainly for its beans. This tree-crop distinction is essential because if cocoa is primarily perceived as a tree, then, by definition, cocoa agroforestry would be inconceivable without a crop or animal component. If cocoa is categorized solely as a crop, a tree component is required to construct a cocoa agroforestry technology. Despite the relevance of this fundamental distinction in influencing stakeholders’ imaginaries or visions of how cocoa agroforestry needs to look like in practice, it appears to be taken for granted in the literature on cocoa agroforestry. I do not conclude this debate a priori. Instead, I take a bottom-up approach, paying attention to what existing forms of cocoa agroforestry say about various stakeholders’ understanding of cocoa’s place in such systems.

The second point on what qualifies as cocoa agroforestry relates to the complexity of how different trees, crops, or animals are arranged spatially and temporally. Drawing on the literature, I use a tripartite construct - basic, intermediate, and dynamic - to characterize cocoa agroforestry, depending on the observed dynamism of its components, as described in Table 1. This distinction differs from works in the current literature, often framed around the binary construct of monocultures and cocoa agroforestry (Niether et al. 2020). Basic cocoa agroforestry systems combine trees/cocoa with crops or animals in the most foundational way possible. Intermediate agroforestry systems focus on specific species interactions. Finally, dynamic agroforestry focuses on building resilient systems responding to various shocks without losing their ability to deliver diverse regulating, provisioning, supporting, and cultural services over time.

Concerning who defines cocoa agroforestry and for whom, these are ultimately questions of authority and involve who wields the authority to influence rules, norms, and discourses that shape how cocoa agroforestry is conceived and implemented. Potential decision-makers around cocoa agroforestry include farmers, public institutions, NGOs, and corporations with different and sometimes conflicting interests. These actors are also embedded in differential power relations, employing various strategies to reduce and simplify socio-ecological problems to render them authoritative in offering their preferred management prescriptions (Scott 1998). Within this context, the concept of multi-dimensional equity (McDermott et al. 2013) provides a useful framework to examine how the different actors’ interests in basic, intermediate, and dynamic agroforestry intersect, converge, or diverge. These interests broadly cover various economic, social, cultural, and environmental outcomes.

Applying an equity lens to cocoa agroforestry enables me to focus on how cocoa farmers, as one of the key targets of equity in discourses on cocoa (Hütz-Adams 2022; McMillan 2001; Waarts et al. 2021), are involved in the design and implementation of cocoa agroforestry, the extent to which their local contexts are considered, and whether the benefits form cocoa agroforestry are distributed fairly. Participation, contextual recognition, and benefit distribution are distinct aspects of equity. However, they are also interrelated. Inference from the participatory governance literature indicates that involving local communities in designing cocoa agroforestry systems improves the chances of having their concerns and interests recognized, leading to tailored solutions that have a better chance of responding to their most pressing needs and concerns.

Thus, engaging this analytical typology, firstly, I parse out the forms of agroforestry implemented under current interventions based on ethnography and field observations conducted over several years in Ghana’s Juabeso/Bia Landscape (JBL), elaborated in “Research area and methods” section. Secondly, I analyze how current cocoa agroforestry proponents engage cocoa farmers as targets of equity in their design. Cocoa farmers are a target of equity because of several factors, including poor access to economic benefits from cocoa despite making efforts to improve their output (Bosselmann et al. 2024; Iddrisu et al. 2020; McMillan 2001). Finally, amid growing concerns about rising food insecurity among cocoa farmers (Amfo et al. 2021; Hashmiu et al. 2022), I employ the analytical lens of benefit distribution to discuss how well existing forms of cocoa agroforestry in the JBL influence cocoa farmers’ access to food as a basic need. This enables me to complement the existing literature, which, for example, raises critical concerns about rising poverty among cocoa farmers (Boysen et al. 2023; FAO and BASIC 2020).

Research area and methods

Case description—the Juabeso Bia Landscape

Commercial cocoa production emerged in eastern regions of Ghana around 1879. In The Migrant Cocoa Farmer of Southern Ghana, Polly Hill (1961) highlights how capital from enterprising migrants and their social relations of labor enabled a shift toward the middle and western regions soon after. The migrant cocoa farmer was motivated by profits, enhanced by access to fertile, disease-free land in the forest frontiers of the country. This trajectory established the cocoa in the JBL from the early 1900s amidst significant contestations between the colonial state and traditional authorities over who needed to wield the authority to preserve forests and enable sustainable cocoa production. Kumeh (2023) argues these contestations and associated struggles persist and form an essential social-ecological context within which land-use decision-making in the JBL may be discussed.

Nestled within the West Semi-Equatorial climate zone, the JBL is characterized by diverse social-ecological features, summarized in Table 2. The landscape has fertile soils, good rainfall distribution, and a humid environment suitable for farming. Predominantly rural (75.7%), agriculture is the primary livelihood activity in the region. Access to land is one of the limiting factors to farming food crops and cocoa in the JBL (Ajagun et al. 2022; Kumeh et al. 2022), with farmers documented to employ various strategies to farm in forest reserves, which constitute about 45.4% of the area’s estimated 244,000 ha.

Table 2 Selected socio-ecological attributes in the Juabeso/Bia Landscape
Full size table

Many stakeholders, such as the Ghanaian government, civil society, and cocoa companies, recognize the tensions between agriculture and forests in the JBL. Consequently, they have made the JBL one of the priority landscapes for land-use planning and investments in cocoa agroforestry (MLNR et al. 2023). To understand the various initiatives directed towards cocoa agroforestry in the JBL, I distinguish between additive and renovative mechanisms depending on how they reconfigure the various components of cocoa agroforestry. Additive interventions focus on already established cocoa farms, primarily integrating various trees into such farms. In contrast, renovative interventions completely remove old and/or diseased cocoa farms to establish new cocoa agroforestry from the ground up.

One example of the renovative intervention in the JBL is the National Cocoa Rehabilitation Programme (NCRP), facilitated by the Ghana Cocoa Board (COCOBOD). COCOBOD, through its Cocoa Health and Extension Division (CHED), uses the NCRP to support cocoa farmers in re-establishing old, swollen shoot-infested farms. Unlike the renovative actions under the NCRP, additive interventions occur primarily under the Ghana Cocoa Forest REDD + Programme (GCFRP), Cocoa and Forests Initiative (CFI), and various NGO interventions. GCFRP is led by the National REDD + Secretariat, which, in collaboration with NGOs and cocoa companies, establishes landscape governance institutions and supports farmers to maintain trees in their cocoa in order to qualify for performance-based carbon payments and non-carbon benefits for their contributions to climate mitigation and climate-smart cocoa production, respectively (MLNR 2017). CFI is a public–private partnership launched by the governments of Ghana and Cote d’Ivoire and 35 leading cocoa and chocolate companies to eliminate cocoa-related deforestation and restore degraded forests in cocoa-growing areas (MLNR et al. 2021, 2022). Table 3 is an overview of the various interventions within which cocoa agroforestry was engaged in the JBL.

Table 3 An overview of the cocoa agroforestry interventions observed in the JBL
Full size table

Data collection and analysis

I bring together data from four sources to demonstrate how cocoa agroforestry plays out under additive and renovative intervention in the JBL and discuss whose priorities are served. To start with, I reviewed written materials on the various interventions, as they reflect various actors’ stated perceptions about cocoa agroforestry. This includes GCFRP and CFI annual reports. Secondly, and with the verbal consent of respondents, I held five focus group discussions (FGDs) involving 14 NCRP beneficiaries and 18 non-beneficiaries to understand their views and interactions with the various interventions. All the NCRP non-beneficiaries indicated they had received seedlings from companies and NGOs to plant on their cocoa farms over the last two years. However, they were uncertain whether this came strictly from GCFRP or CFIFootnote 2 activities. Thirdly, I held key informant interviews with public officials (2), NGO staff (4), cocoa-buying companies (2), and village chiefs (2) in the JBL. The FGDs and interviews sought to understand the various actors’ roles and experiences with the NCRP and the other inventions and how they addressed farmers’ needs in JBL. Finally, I draw on ethnographic materials collected through cocoa farm observations and resource mapping from my documented longstanding engagements with cocoa farmers in the JBL between 2014 and 2022 (Asumang-Yeboah et al. 2022; Kumeh 2023; Kumeh et al. 2021, 2022). The research could have been improved with the collection of biophysical information, such as the relative number of trees on the various cocoa agroforestry farms. However, such data is not included in this study, which focuses on the social aspects of the systems observed.

Notes from the interviews and focus groups were analyzed to identify themes around participation, consideration of farmers’ operational context and needs, and their access to various benefits from cocoa agroforestry within the equity framework. I organize the findings by dissecting the origin, actors, and interests within the various renovative and additive cocoa agroforestry initiatives, the forms of cocoa agroforestry they deliver in practice, and the distribution of their associated outcomes. In presenting the results, I periodically quote materials from the FGDs to convey the views of the various respondents engaged.

Results

Within the JBL, various forms of cocoa agroforestry were observed to be implemented (groups of) actors such as cocoa farmers, agricultural extension agents, staff of cocoa companies, and forestry officials. Except in a few cases where free-range chicken and pigs were observed on cocoa farms within a couple of kilometers from village settlements and one case where two cows were tethered and fed with fodder from nearby fallow lands, livestock was not managed as a core component in cocoa agroforestry practices observed in the JBL. Cocoa, shade trees, and food crops as the main components were seen to be configured in various ways over time. At one end of the continuum, I distinguished farms dominated by cocoa and interspersed with native trees. At the other end were farms dominated by food crops and young cocoa trees (Fig. 1). As I will demonstrate in the results, the interests or priorities of various (groups of) actors entail one of the deciding factors of not only the complexity of the cocoa agroforestry in the landscape but also the kinds of benefits farmers can access over time.

Fig. 1
figure 1

Different forms of cocoa agroforestry. A renovated, young cocoa agroforestry (under a year old) with plantain, cocoa, and shade trees (a), and an additive cocoa agroforestry farm with a few shade trees and no crops in the cocoa understorey (b)

Full size image

Cocoa agroforestry under renovative initiatives with a focus on the NCRP

Historically, cocoa rehabilitation has been used by the Ghanaian government as a “reset mechanism” to tackle pests and diseases in cocoa, as well as replace old, “less productive” farms (from a cocoa beans output perspective). The cocoa swollen shoot virus disease (CSSVD) is the disease of greatest concern due to its acuteness and lethality. CSSVD-infested cocoa trees eventually die, with the eradication of diseased farms being the only control currently available (Ofori et al. 2022). The NCRP is used for this purpose. COCOBOD (2020a, p. 22) holds the NCRP as one of its cocoa productivity enhancement interventions.

Origin, actors, and interests

Cocoa rehabilitation is not new in Ghana. Typically, the cocoa rehabilitation is preceded by the cutting of diseased CSSVD farms. This has historically been a source of conflict between the ruling class and local actors, especially cocoa farmers. The latter establishes cocoa farms using credit and labor to secure lands as a long-term investment and means of passing down wealth to succeeding generations. Many cocoa farmers in the JBL are migrants from other parts of Ghana. Migrant farmers’ ability to retain control over their cocoa farms and secure land access to land under customary arrangements is contingent on maintaining cocoa trees on the land (Boni 2006). Thus, cutting their cocoa trees constitutes a serious threat to their livelihoods and well-being.

During the colonial era, when Ghana (then Gold Coast) was controlled by British, the colonial administration used coercion as its primary tool to destroy CSSVD-infested cocoa farms. Subtle forms of resistance by cocoa farmers to the destruction of their farms in the 1930s were silenced through legislation such as the Swollen Shoot Disease of Cocoa Order 148 (1946). The British colonial Governor E.N. Arden-Clarke called this piece of legislation “compulsion by consent” as it enabled the colonial administration to use security forces to destroy CSSVD-infested cocoa farms with or without the consent of cocoa farmers (Danquah 2003). The British colonial administration’s main interest in cocoa was securing foreign exchange and raw materials industries to feed the chocolate industry in the UK and beyond.

Considerable grassroots resistance, including violent clashes between staff of the Cocoa Rehabilitation Department and a coalition of cocoa farmers, local chiefs, and nationalists to the “compulsion by consent” approach, resulted in the dissolution of the order in 1951. Since then, various post-colonial government administrations have employed different strategies to compensate farmers to avert tensions associated with cutting down diseased cocoa farms. These include providing farmers with inputs to re-establish their cocoa farms using loans from multilateral banks. For example, cocoa rehabilitation programs across Ghana, including the JBL landscape in the 1980s and 1990s, were by the World Bank and the African Development Bank (African Development Bank 2002; World Bank 1987).

The current NCRP in the JBL is part of a national intervention that the government first piloted in the 2016/2017 season. The government scaled it out in 2020 using USD 200 million out of a USD 600 loan procured from the AfDB, the Industrial and Commercial Bank of China Limited, and Credit Suisse (COCOBOD 2020b).

Form(s) of cocoa agroforestry

Under the NCRP, district Cocoa Health and Extension Division (CHED) officials identify and work with local contractors to cut down and fumigate old and/or diseased cocoa farms. COCOBOD pays cocoa farmers and their landowners GHS 1000 (USD 78)Footnote 3 for every ha of cocoa farm cut for rehabilitation. It also pays laborers to plant plantain suckers, hybrid cocoa seedlings (1100–1200/ha), and shade trees on disinfected farms. The plantain provides temporal shade and protection to the cocoa seedlings. Tree seedlings observed on the rehabilitated farms were predominantly Terminalia spp. and Khaya ivorensis. CHED officials argue that they aim to maintain between 15 and 18 trees to form an overstory based on recommendations of the Cocoa Research Institute of Ghana (CRIG). Furthermore, CHED officials held that the COCOBOD pays all costs (inputs, labor) for maintaining the farm for two years when it relinquishes control over the renovated farm to its owners. Apart from providing consent to be involved in the NCRP, cocoa farmers observed that they generally have little say in the spatial and temporal configuration of the agroforestry components under the NCRP. None of the rehabilitated cocoa was fruiting during the time of data collection. Food crops within the established farms were, however, yielding.

Cocoa agroforestry farms observed within the NCRP may be described as intermediate, with the complexity likely to reduce considerably as the cocoa matures and takes over the most space within the system (Fig. 1a, b). The current approach appears designed primarily to secure cocoa beans as the main output over time, and all other components are configured toward reaching this goal. Two factors may explain this observation. Starting with trees, old-growth trees identified in diseased cocoa farms were felled or maintained based on CHED officials’ assessment of their potential to harbor mealybugs that transmit the CSSVD. This led to the felling of several tree species including Ceiba pentandra, Pouteria altissima, Morinda lucida, Spathodea campanulate, and Albizia zygia. Some of the standing trees observed on the farms included Newbouldia laevis, Pycnanthus angolense, Terminalia spp., and Alstonia boonei. The felling of diverse native trees and replacing them with the choice between a few tree species could impact future biodiversity in the newly established cocoa agroforests, given that food crops and cocoa trees may outcompete, or hinder the natural regeneration of various indigenous trees.

The second point concerns food crops. There is a strong supply of plantain and other food crops during the initial years of establishing the cocoa agroforestry. However, this appears to wane as the cocoa trees become established. Thus, a food crop production dilemma under the NCRP is that it contributes directly to cocoa households’ food availability during its initial years but requires such households look elsewhere for food as their cocoa matures.

Distribution of outcomes

Who gets benefits and when from cocoa agroforestry established under the NCRP depends on their role within the system. From the onset, various individuals and companies gain short-term employment, for example, through contracts to raise tree seedlings and plant suckers. Similarly, people in local communities gain employment as they are hired to fell and disinfect the disease-infested cocoa and shade trees, as well as agricultural practices such as lining and pegging and weed control. The state’s interest regarding benefits is medium to long-term, primarily access to quality cocoa beans for the international market. The Ghanaian government, through COCOBOD, is instrumental in setting cocoa prices, which enables it to tax cocoa farmers directly. Historically, cocoa farmers receive less than 70% of the free-on-board (FOB)Footnote 4 price for cocoa (Frimpong-Ansah 1992; Kolavalli and Vigneri 2017). While this invariably enables the state to benefit from the cocoa trade, its capture of benefits pales in comparison to cocoa and chocolate multinationals and retailers who retain most of the profits from the cocoa-driven chocolate economy (FAO and BASIC 2020).

Valorizing farmers’ benefits is quite challenging due to the opportunity cost of making their farms available for renovation. It also depends on the type of farmer, i.e., sharecroppers, resident farmers, or landowners. Irrespective of the difference, the main benefits include initial payments per hectare cocoa farm put up for renovation (to cocoa farmers and/ or landowners), employment and monetary payments to laborers, access to food crops by farmers during the establishment phase of the cocoa agroforestry (Fig. 1a), and cocoa beans derived by sharecroppers and landowners in the medium to long term. Overall, specific cocoa farmers’ ability to benefit from their renovated farms depends on many factors, such as access to inputs (fertilizers, fungicides, insecticides) and labor, CSSVD recurrence, and climate risks.

Figure 2b–d, an annual timeline developed from focus groups, field observations, and interviews, offers a broad overview of the relative supply of food and cocoa over time from cocoa agroforestry, as well as labor requirements in the JBL. This is useful in understanding the distribution of potential benefits and costs of farming in the JBL. The figure indicates that labor requirements are fairly similar for cocoa agroforestry and mixed cropping involving Musa paradisiaca (plantain), Manihot esculenta (cassava, manioc), Colocasia esculenta (cocoyam), Solanum lycopersicum (tomato), Abelmoschus esculentus (okra), and Solanum melongena (garden eggs). However, benefits from food crops tend to have a wider spread than cocoa over time, an issue that is elaborated in the next section.

Fig. 2
figure 2

Timelines of labor, food, and cocoa beans supply over a typical year in the JBL. An increasing number of bars indicates greater requirement or availability. Labor required for managing a matured or additive, renovative cocoa and mixed cropping farms (a), supply of cocoa from an average matured cocoa agroforestry farm (b), food crops availability from matured agroforestry farms, mainly Colocasia esculenta and Dioscorea villosa (c), food crops availability from a renovative cocoa agroforestry plots [during the establishment phase], also applicably to food mixed farming in the Krokosua Hills Forest Reserve within the JBL

Full size image

Cocoa agroforestry under the additive initiatives

Additive forms of cocoa agroforestry have emerged over the last decade partly in response to concerns about the impacts of climate variability, particularly heat stress, on cocoa yields. Proponents and investors in additive cocoa agroforestry interventions argue their investments are essential to mitigate climate change, sustain cocoa yield, and improve cocoa farmers’ well-being over time. There is, however, a poor understanding of how their intervention configures cocoa agroforestry in practice. In addition to drawing attention to inherent shortfalls associated with their progressive nature, the insights below lay the foundation for further research connecting the policy and practice of cocoa agroforestry.

Origin, actors, and their interests

Additive forms of cocoa agroforestry became popular in the late 2000s. Historically, when establishing cocoa, farmers reserved and tended native trees considered useful for shade, medicinal, and other values. The confluence of two factors disincentivized the retention of such trees on cocoa farms. The first involves complexities surrounding tree tenure in Ghana. Legislative reform under Ghana’s Concession Act of 1962 (ACT 124) confers property rights over naturally regenerated trees to the state. This mandate introduced a significant risk to all forms of agroforestry on farmlands, as it enabled forestry officials to issue permits for harvesting trees, including cocoa, on farmlands. In harvesting such trees, cocoa farmers were often not compensated properly (Hirons et al. 2018). Besides, it appears contradictory that the state encourages farmers to maintain trees on their farms on one hand while issuing permits to harvest matured ones on the other. Disgruntled cocoa farmers are widely known to use various techniques to eliminate trees on their farms, including burning large trees at their base, debarking trees, applying herbicides to the stem, and uprooting tree saplings.

The second factor concerns a policy failure. To improve cocoa productivity, COCOBOD incentivized the establishment of so-called hi-tech cocoa monocultures since the mid-1980s. This was based on the assumption that using hybrid, sun-loving cocoa will boost productivity over time. However, the onset of climate impacts and its devastating impact on cocoa monocultures led policymakers to reconsider their decision (Asitoakor et al. 2022a, b). This was catalyzed by a rise in scientific evidence that cocoa agroforestry systems were more resilient to heat stress and thrived better in the long term (Asitoakor et al. 2022a, b; Blaser et al. 2018; Middendorp et al. 2018). While there are disagreements between scholars and practitioners on the optimal shade levels for effective cocoa agroforestry, the widely recognized policy position of COCOBOD is for farmers to maintain 15–18 shade trees per hectare of their cocoa farm (COCOBOD et al. 2016, p. 23). Additive cocoa agroforestry interventions emerge as a path to reach this goal in already established cocoa monocultures.

Additive cocoa agroforestry interventions primarily supply cocoa farmers with seedlings to integrate into their farms. Within the JBL, this is stimulated by the CFI and the GCFRP. Both interventions are implemented in the JBL by a wide coalition of actors, including COCOBOD, the Forestry Commission, Touton SA, Cargill, Mondelez, Nature Conservation Research Centre, Tropenbos Ghana, and SNV (a Dutch NGO).

Form(s) of cocoa agroforestry

GCFRP and CFI-related investments in the JBL focused primarily on boosting the tree count on existing cocoa farms in the landscape. Similar to the renovative systems, basic and intermediate cocoa agroforestry were characteristic of these investments. Farmers reported receiving seedlings from NGO officials, cocoa purchasing clerks, and COCOBOD extension agents to integrate into their farms. The various stakeholders did this conveniently. For example, farmers pointed out that no prior assessments of the shade status or number of trees in existing cocoa farms were done before they were supplied with the seedlings. Questions about whether the various stakeholders supported farmers in improving food production in the existing cocoa revealed they had not. Where food crops, e.g., cassava, plantain, maize, and cocoyam, were spotted in such systems, farmers argued they had cultivated them to provide temporary shade while replanting dead cocoa trees in such spaces. Multiple farmers reported recycling Dioscorea villosa underneath their cocoa understorey for food. However, they were concerned that the growing use of herbicides to control weed imperiled the practice and by extension their access to food on matured cocoa farms.

Distribution of outcomes

Cocoa beans are the most dominant element from additive cocoa agroforestry observed and most prevalent from late September to mid-January (Fig. 2b). During focus groups, cocoa farmers expressed two main concerns about the distribution of cocoa beans, namely: (1) a lack of cash flows during critical periods of the year, especially late May to mid-August, (2) limited supply of food from the established cocoa farm. May and August mattered to the participants because, during this period, cocoa farmers require considerable resources to manage their farms, including weed control, fertilizer application, and pests and disease management, especially the black pod disease. Left unmanaged, many farmers noted the black pod disease has a devastating impact on their cocoa yield (Fig. 3).

Fig. 3
figure 3

Farmers drying cocoa beans in the JBL. Black pod disease-infested cocoa beans being dried at an aggregation point in Sayerano (left), healthy cocoa beans being dried by a farmer in Mafia (right)

Full size image

Concerning food, many cocoa farmers pointed out that their inability to access food crops from matured cocoa farms increases their exposure to food insecurity, especially during the lean cocoa season from late December onwards. During the focus groups in the Juabeso area, several farmers observed that illegal farming in their adjoining forest reserves has become their main path to gaining access to food. Others access food by buying typically buying bags of rice on credit. However, they point out that this is now a major bottleneck to their well-being, as “food-loan sharks” charge exorbitant interests:

Debt is killing us as cocoa farmers; debt from buying food. We do not have any land for producing food apart from the Krokosue [sic] (Forest Reserve). However, if you go there, the forestry officials will arrest and beat you and destroy your food crops. We end up buying bags of rice on a credit basis from traders; we must register for this in advance. It is the same as the one-for-one [sic]; [facilitator: one-for-one?]. Yes, it is like taking a loan here; you repay twice the amount upon harvesting your cocoa. If you default, it doubles on a compound interest basis the following season, and so on. They are killing us here. However, we don’t have any alternatives. (FGD, Asempaneye, JBL).

When asked why not cut down some of their cocoa trees to create space to grow food crops in their mature cocoa agroforestry, all farmers argued it was inconceivable for them to do. The reasons they cited for this were the risk of losing their farms due to close ties between the existence of cocoa on the land and tenure security, sharecropper/tenant farmer and farm owner relations forbidding the former from destroying cocoa trees, the nature of cocoa as a source of inheritance, and a sense of attachment, personal and national pride associated with growing cocoa. Many of these factors indicate tensions in land-use decisions between securing various forms of cocoa agroforestry and fostering local, household food security directly from the JBL. I examine some of these tensions, situating my findings in the broader literature on cocoa agroforestry.

Discussion

The various forms of agroforestry demonstrated in the results, along with the actors involved, and their interests and priorities raise multiple tensions and questions about who cocoa agroforestry is implemented for within the JBL. Renovative cocoa agroforestry under the NCRP appears to appeal to the immediate food needs of cocoa farmers and to fix farmers into cocoa for long-term access to cocoa beans, which is required to boost the Ghanaian government’s foreign exchange earnings and the profits of cocoa and chocolate companies. The realization of these economic gains over time, as observed with the additive initiatives, displaces food and risks local food insecurity, especially during periods of the year when farmers are unable to access income from cocoa. This section parses out this inherent tension and the factors that precipitate it. I argue that finding a balance requires paying greater attention to the context of cocoa farmers and engaging them actively in rethinking cocoa agroforestry configurations that meet their food and other livelihood needs year-round and over time.

Tensions between renovative and addictive initiatives to implementing cocoa agroforestry

The renovative and additive approaches to cocoa agroforestry observed in the JBL appear to ultimately prioritize mass-producing cocoa beans in comparison with other products such as food, fodder, or fuel. This leads to basic, less diverse agroforestry land-use practices that are perhaps misrepresented in how various actors frame cocoa agroforestry in sustainability debates. It may be argued cocoa agroforestry under the NCRP contributes to local food availability at the initial stages. However, a massive shortfall emerges as soon as the cocoa matures. This shortfall is largely unaddressed by additive forms of cocoa agroforestry in the landscape, resulting in several negative spillovers, including food insecurity-driven forest conversion by cocoa farmers in the landscape. Several interrelated factors contribute to this, including mismatches between cocoa income and cocoa households’ food and nutritional needs, a seeming lack of agency among cocoa farmers created by sunken costs associated with establishing cocoa farms, and the influence of power imbalances between cocoa farmers and advocates of cocoa agroforestry in determining decisions about what and for whom the JBL is administered.

Starting with the latter, the government of Ghana and cocoa and chocolate companies are primarily interested in the continuous supply of cocoa from the JBL. The state asserts this with its sector strategy aimed at more than doubling current cocoa productivity per unit area amid climate impacts. Irrespective of variations in their starting point of renovative and additive initiatives, the end goal is to maintain 15–18 shade trees per hectare of cocoa, which the state deems suitable for climate-smart cocoa production (COCOBOD et al. 2016). Within Ghana’s cocoa sector framework, climate-smart cocoa is cocoa that is resilient to climate risks such as heat stress, droughts, and pollinator decline. Some studies have questioned the adequacy of the prescribed number and types of trees used in cocoa agroforestry to engender climate-resilient cocoa (Abdulai et al. 2018). More than planting a specific number of trees, some scholars call for moving toward a shade-tree cover that is under 30% for optimal outcomes in terms of cocoa productivity and climate and biodiversity benefits (Blaser et al. 2018). While useful, much like the government, scholars tend to discuss shade regimes within cocoa agroforestry, mainly in relation to cocoa productivity over time. Food crops are often relegated to the margin or treated casually as a by-product of establishing new cocoa farms despite considerable evidence that cocoa cultivation displaces food production in Ghana (Ajagun et al. 2022; Asubonteng et al. 2018). The apparent neglect of long-term food production in cocoa is not only a policy problem created by the government’s power over cocoa farmers; it is probably reinforced by scholarship that centers cocoa productivity and climate resilience in discussions on cocoa agroforestry systems.

One potential explanation for why policymakers, practitioners, and scholars appear fixated on cocoa productivity in cocoa agroforestry concerns the nature of cocoa as a cash crop and the underlying logic that income from cocoa will enable farmers to access diverse food sources. The findings suggest that stakeholders focus on cocoa as a crop component of cocoa agroforestry, manipulating it with a few shade trees in trying to validate the agroforestryness of their practices. This orientation appears to be problematic because cocoa farmers in the JBL do not interact with cocoa as food directly; they primarily exchange cocoa for cash. There is nothing wrong with securing income from cocoa per se; it has been argued cocoa farmers rely on “bulk money from their cocoa beans to service debts, pay school fees, and renovate their houses” (Bymolt et al. 2018, p. 144). A challenge emerges, however, when cocoa households are over reliant on income from cocoa to eat. Such a reliance increases their vulnerability considerably. This was illuminated during the COVID-19 pandemic, where delays by the government to secure its yearly syndicate loans to buy cocoa beans left so many cocoa farmers with cocoa beans but no money to buy food (Asante-Poku and Van Huellen 2021).

Moreover, it is also questionable whether farmers with smallholdings (< 2 ha), as is the case of most cocoa farmers in Ghana (GSS 2020), generate sufficient income from their cocoa beans to overcome the lurking threat of food insecurity in their households. Hashmiu et al. (2022, p. 1) put this succinctly when they write following their analysis in 408 cocoa households that ‘income from cocoa alone is not high enough and widely distributed to guarantee food security’. This observation is not peculiar to farmers with basic cocoa agroforestry farms. In contrast, similar observations have been made for farmers involved in cocoa certification programs that tend to emphasize having more shade trees in cocoa farms (Iddrisu et al. 2020).

If cocoa agroforestry is not properly addressing farmers’ food needs, why do they remain in cocoa? This is the third intersecting point. Indeed, as demonstrated in the results, some cocoa farmers are giving up on cocoa where conditions allow. These conditions include, for example, the availability of alternative land-use options, especially opportunities for gold mining on existing cocoa farms (see Fig. 4).

Fig. 4
figure 4

Active surface gold mining on a cocoa farm in the JBL

Full size image

Where this deceptively lucrative option is unavailable, it appears the intersection of multiple cocoa sustainability initiatives and campaigns by NGOs, COCOBOD and cocoa companies -as demonstrated in the introductory quote- along with opportunities for growing food crops in adjourning forest reserves, farmers’ perception of cocoa as a means for inheritance, and personal attachment to the cocoa economy affect cocoa farmer’s ability to exert their agency to diversify and improve the integration of food crops on their existing cocoa agroforestry systems. On the availability of forest reserves for food production, a recent assessment within the CFI concluded that food crop production is the single most important driver of forest conversion in the JBL (MLNR et al. 2023, p. 21). As raised in the results and reported elsewhere, cocoa agroforestry reinforces this by “eating up arable land” outside of forest reserves (Ajagun et al. 2022). Thus, unless stakeholders can find ways to accommodate food crops more permanently in cocoa agroforestry, this trajectory is likely to continue if not exacerbate. I lay out some of the policy and research implications of these findings before ending with some concluding thoughts.

Policy and research implication of findings for more equitable cocoa agroforestry

Many insights may be drawn from the findings to improve equity in cocoa agroforestry. However, I will direct attention to two main points. Firstly, policymakers and companies need to recognize and pay attention to the contextual experiences of cocoa farmers, particularly regarding tensions between cocoa production and food insecurity in cocoa households. This could be achieved by creating more space to understand cocoa farmers’ needs and working with them to rethink cocoa agroforestry technologies that contribute to addressing them. Finding ways to better accommodate more food production in existing cocoa agroforestry systems may influence cocoa outputs. However, it is also likely to make such farms more dynamic and resilient. This does not mean romanticizing long-term food production in cocoa. In contrast, it is important to recognize there may be critical tensions in integrating food crops within the cocoa understory. In a recent systematic review of agroforestry studies in tropics and sub-tropics, Kumar et al. (2024) found that 113 out of the 145 tree-crop combinations reviewed found negative effects of the understory crop on tree crop yield. However, overall tree-crop output was observed to have been maximized for most of the systems.

This suggests that the question of food in cocoa agroforestry is not necessarily whether to improve food in cocoa understory or not. Instead, it is about whose interests and needs matter enough to be prioritized. As demonstrated in the findings, the prioritization of cocoa over time appears not to be serving cocoa farmers well in the JBL. Thus, perhaps it is time to try something new. Initial indications from the findings are that Dioscorea villosa and Colocasia esculenta are compatible as understorey food crops for cocoa. Moving forward, more research is needed to better understand locally relevant crops that thrive well with cocoa and shade trees, as well as the gendered labor allocation options that offer beneficial outcomes for cocoa households when they integrate significantly greater food crops in their cocoa.

Meanwhile, a lack of evidence should not be used as a political excuse to delay action, given we are only six years away from meeting our global commitment to end hunger within the sustainable development framework. Alternatively, policymakers, NGOs, and cocoa and chocolate companies alike are encouraged to embrace experimentation and adaptive learning to foster positive change over time. By way of practical pathways, policymakers could consider mandating a food reserve in all cocoa farms. This means reserving a fixed area within cocoa farms for food crop production, which could be achieved by integrating it into the national sustainable cocoa standard within the broader framework of the ARS-1000 regional standard on sustainable cocoa. Such an approach could also strategically position the country to address displaced deforestation created by the conversion of forests to produce food and improve the country’s commitment to reducing deforestation locally and getting ahead of European Union Deforestation Regulation requirements. Another option would be to introduce short-term food subsidies for cocoa households in the landscape, especially between late May and late July. These are essential to save lives and protect cocoa farmers from the indignity of falling into debt distress from the exorbitant interests on food loans in the study localities. Additionally, they also provide a pathway for tackling the contributions of cocoa agroforestry to deforestation from forest conversion to feed cocoa households.

The second point is research-oriented and in two directions. To start with, more research is needed to understand how planting a few selected tree species in renovative and additive cocoa agroforestry systems impacts biodiversity and CSSVD reinfection on cocoa farms and the landscape at large. An earlier study by Anglaaere et al. (2011) found tree diversity declines significantly as one moves from natural forests through fallow lands to mature cocoa farms to newly established cocoa. While this study observed the obliteration of native shade trees was done with the argument to prevent future CSSVD reinfection, other studies report high reinfection in newly planted farms CSSVD (Ameyaw et al. 2015; Amon-Armah et al. 2021), leading to calls for more integrated ways of managing the virus. Some authors have demonstrated that using cocoa swollen shoot virus immune crops such as oil palm, citrus, and kola as natural barriers reduces CSSVD reinfection (Domfeh et al. 2016). More studies in this direction are encouraged to inform more targeted native tree management in CSSVD-infested cocoa farms and efforts to diversify food production in cocoa farms.

The other aspect of research concerns the need for a more systematic analysis of power in cocoa agroforestry systems to better understand how it influences decision-making and forms of control over how and for whom land resources are managed. This is necessitated by how food access is a critical challenge among cocoa farmers but appears to have been overlooked for so long in the interests of producing cocoa to meet the needs of external actors. Given the limited analysis of power in the existing literature on agroforestry, perhaps organizing future work under the analytical handle of critical agroforestry could render this more legible. As a preliminary definition, critical agroforestry may entail the analysis of how power differentials or imbalances in society influence decision-making processes and resource distribution within agroforestry systems. This involves analyzing how decision-making authority is distributed among actors in polycentric systems, including farmers, policymakers, civil society, and companies. Understanding these power dynamics is crucial for ensuring that the benefits of agroforestry are equitably distributed, avoiding situations where certain groups have disproportionate control over resources or decision-making processes, and ultimately fostering a more inclusive and sustainable approach to agroforestry.

Conclusions

Amid the growing promotion of cocoa agroforestry as an essential solution to multiple problems in the cocoa sector, including poverty and hunger among cocoa households, this study sought to examine equity in the delivery of such systems. By deconstructing cocoa agroforestry and parsing out how cocoa, shade trees, and food crops intersect over time, I have demonstrated a misalignment of interests and priorities between the Ghanaian government, cocoa companies, and farmers involved in renovative and additive forms of cocoa agroforestry in Ghana’s Juabeso/Bia Landscape. Additionally, neglecting long-term food production in cocoa agroforestry suggests producing cocoa for profits has come to take precedence over meeting the basic food needs of cocoa households in the landscape. Yet, amid the insufficiency of income from cocoa to cover farmers’ year-round food needs, cocoa agroforestry has arguably become an indirect driver of deforestation, as it pushes farmers to convert their adjourning forest reserves to meet their food needs. Addressing food production in cocoa understorey is thus not only a means to addressing food insecurity and hunger in cocoa households but an essential means for overcoming deforestation. Achieving this requires reconfiguring current power relations to give cocoa farmers more space to influence decisions on what and for whom rural land resources are managed. Critical agroforestry provides for a more deliberate thinking about how power shapes choices among power differential actors, improving the chance to break the hegemony of term in pursuit more equitable agroforestry options that works for people and nature.