Plantations, women, and food security in Africa: Interrogating the investment pathway towards zero hunger in Cameroon and Ghana

In this paper, we critically engage with SDG-2- Target 2.3 and SDG-5 to examine how and why large-scale agricultural land acquisitions modify the social relations of women’s food access. Adopting qualitative methods and from agrarian political economy perspectives, we assess the implications of plantation investments on food security and gender equality in Cameroon and Ghana. The study draws on the impacts of various plantation schemes in Cameroon and Ghana to argue that the current all- encompassing framing of the SDG-2, whereby support for small-scale production appears to co-exist with the promotion of corporate-led agricultural investments, tends to create a vague interpretation of food security, even when capitalist motives override the interests of rural working men and women. In both countries, plantations have been characterised by displacement, reduced food production and competi- tion over land resources in rural contexts where women bear the burden of social reproduction, particularly in subsistence and food provisioning for their households. Restricted access to farmlands hampers petty commodity production, and provokes rural exodus among farming populations in Cameroon who compete with pastoralists for land; and in the Ghana case, where settlers and migrants compete with dis- possessed indigenes for arable lands. The study shows the inherent contradictions and tensions within the narratives of sustainable development and projects which are considered as potential pathways to the SDGs. One of such tension areas is between local food security- which is a gendered role for women in West/Central Africa – and the accumulation imperatives of capitalist investors. We iterate that achiev- ing zero hunger requires gender-inclusive land and labour policies that recognise, protect and empower small farmers and women, and create opportunities for local and national food self-sufﬁciency. (cid:1) 2020 The Authors. Published by Elsevier Ltd. ThisisanopenaccessarticleundertheCCBYlicense(http:// that the of


Introduction
Africa remains the most targeted region for large-scale land acquisitions (LSLAs) by foreign and national investors in the global South. For instance, over 3 million hectares of agricultural land concessions; 2.2 million in Cameroon and 1.1 million in Ghana have been acquired for biofuels, forestry, food crops, oil palm, and rubber, mostly for exports (Land Matrix, 2020) 2 . This ongoing rush for land in Africa has heightened debates on land tenure governance, rights to land, issues of dispossession, and sustainable development. Notably, in (West) Africa where small scale and family farming systems depend significantly on women's labour, the surge in LSLAs is shaping and redefining women's land rights and their role in domestic food security (Fonjong, 2004;Julia & White, 2012). The Consolidated Approach to Reporting Indicator of Food Security for Cameroon indicates that 35 per cent of the country's population is food insecure, 49 per cent marginally food secure and 16 per cent severely food insecure in 2017 and at least one in three children continues to suffer from chronic malnutrition (De Schutter, 2012). Ghana, on the other hand, though rated among the most food secure in Sub-Saharan Africa, 5% of the population is food insecure, and about 2 million people are predisposed food insecurity, not neglecting the wide in-country disparities and seasonal variations (Global Food Security Index, 2017). It is against this backdrop that this paper uses the example of Cameroon and Ghana to examine how and why women's access to land and their contribution to food security is affected by large-scale land acquisitions. It also demonstrates how efforts to achieving food security also require progress in gender equality through gender inclusive land and agricultural policy frameworks that recognise and empower women and small farmers.
The emerging evidence is that, regardless of the usual promises of jobs, the gendered nature of some tasks and models of farming, skill requirement, and other conditions may not necessarily increase incomes and open windows of job opportunities (Gyapong, 2019;Hall, Scoones, & Tsikata, 2017). Although Fonjong et al. (2007) and Fold (2008) found significant integration of rural labour on some plantations in Cameroon and Ghana respectively, they also emphasised the differentiated terms of incorporation and unequal opportunities for different social groups based on gender and access to land. In effect, context matters, and as we show in the subsequent sections, the food security implications in Cameroon and Ghana are strongly linked to both the accumulation logic of capitalist production and policy shortcomings. This situation undermines the underlying gendered nature of local resource ownership and production systems that define production and food security.
The study explores two broad, interrelated aspects of gender relations in communities affected by plantation land deals: how domestic and capitalist relations of production structure; reinforces inequalities in access to and control over land and labour, consequently the implications for food security. The study sheds light and contributes to critical debates on with the complexities and gender differentiated impacts of the global rush in rural African communities. The emphasis on women's experiences in securing their households food security amidst the increasing capitalist enclosures for plantations, contribute to knowledge, alternative policy pathways and also in the light of the Sustainable Development Goals (SDGs) which calls for reduced inequalities between men and women in the access to resources.

SDG-2: A political economy of food security and social reproduction
The most widely accepted definition of food security is 'when all people, at all times, have physical and economic access to sufficient, safe and nutritious food that meets their dietary needs and food preferences for active and healthy life (FAO, 1996). This understanding of food security encapsulates ideas and policy shifts over the past four decades as linked to post World War II food supply deficiencies, food price fluctuations and global hunger. For instance, between the 1940s and 1960s, many states focused on their national food security by protecting and prioritising their domestic markets more than external trade. In the 1980s when the notion of access, emanating from Sen's (1981) groundbreaking work on famines came to the foreground, it also marked the beginning of the present corporate food regime (McMichael, 2009). In recent times, there has been a growing political agenda by civil organisations and radical agrarian justice movements which have brought prominence to right-based framings to address the inequalities in access to food (Jacobs, 2013). Such political framings, like food sovereignty, have often been debated in opposition to food security, but as Clapp, (2014, p. 210) rightly indicates, there is the need to guard against '. . .a false binary that conflates the current mainstream normative agenda with the more descriptive and open-ended concept of food security. ' Although there is a consensus around the four pillars of food security-availability, access, utilisation and stability, the productivist (availability) framing appears to dominate discourses and policies; putting at the margins, access, distributional and powerrelated issues (Tomlinson, 2013). Indeed the SDG-2 goal of ending hunger, achieving food security and improved nutrition and sustainable agriculture by 2030 recognises the interlinkages within the food system whereby access, diet, health, and nature are central. Nonetheless, most of the targets and indicators appear to be underpinned by a narrow premise that the challenges of access to quality food are mainly tied to low productive capacities and the poor functioning and access to commodity markets (see targets 2.3, 2.4, 2A, 2B, and 2c). Food security in Africa transcends hunger, nutrition and production (Amanor, 1999;Tomlinson, 2013). Some have argued that the four billion tons of the annual global food production is sufficient to feed the world but for the problems of access, distribution and waste (Hughes, 2016). The structural and relational dynamics of global and local food politics of access in this present 'neoliberal' hegemony cannot be overstated; yet, only partially touched on in the targets and indicators of the SDG-2. Target 2.3 3 , should be the space for engagement with these power and land access related issues. However, like the other SDGs, the indicator for achieving this target is also expected to be measured by the volume of production and income from production. It presupposes that to achieve food security, rural agricultural lands have to be in constant use-thereby overlooking traditional/family farming practices like shifting cultivation, land rotation, hunting, fuelwood collection by women, and grazing. It also begs the question of sustainability and intergenerational land use access.
While we acknowledge the role of plantation agriculture for scaled-up production, we probe into the gender implications of plantations and food insecurity through the lenses of land access, wage labour and social reproduction. As we discuss the SDG-2 in this paper, we do so critically from an agrarian political economy perspective in examining the structural and relational factors that modify access to locally appropriate food when investors acquire large-scale rural, forest and smallholder lands for plantation agriculture. Rather than following the indicators of target 2.3 rigidly, we are inspired by Ribot & Peluso (2003) to define access as the 'the ability of present and future generations to benefit from land resources' under plantation agriculture. From this view of access, we unpack the power relations that determine and hinder access, and how women lose their existing benefits from land, e.g. for food production and other monetary benefits (Hall, Hirsch, & Li, 2013). Borras and Franco (2012) also give insights into the direct impacts of land use changes associated-food to food, food to biofuels, food to non-food and non-food to biofuels. Thus, depending on the models, crop type and purpose of the large-scale investment, local food systems could be directly displaced or indirectly affected by the intensification of forest, marginal or the so-called idle lands which could have implications for future generations (Hall, Scoones, & Tsikata, 2017). Another important subject is how plantations and the sudden commodification of land through large-scale land transactions for plantations can modify existing customary land tenure systems. While the customary tenure system provides women with usufruct rights to produce food for household consumption, this system in itself cultivates gender disparity in access to and control over land. Indeed rural farm households in western Africa, do not organise the distribution of resources based on norms of sharing and pooling (Whitehead & Tsikata, 2003;Whitehead, 2009) and so customary land institutions are not necessarily sites of equity. Thus, we emphasise that when customary lands are deeply integrated into capitalist land markets, not only does it exacerbates pre-existing inequalities, but also breeds new forms of control, often to the disadvantage of women (Razavi, 2009). Therefore questions of rights over land, the existing tenure institutions, patterns of inheritance and how benefits from land are determined are all essential contextual issues for examining gendered relations in the course capitalist land transactions for planation (Yaro, Teye, & Torvikey, 2017). 3 By 2030 double the agricultural productivity and the incomes of small-scale food producers, particularly women, indigenous peoples, family farmers, pastoralists and fishers, including through secure and equal access to land, other productive resources and inputs, knowledge, financial services, markets and opportunities for value addition and non-farm employment Lotsmart N Fonjong and Adwoa Y Gyapong World Development 138 (2021) 105293 In light of recent plantation land deals, the labour question serves as an important link between women's income and food security. However, wage labour is another area that the SDG-2 is almost entirely silent about, and access to markets is again narrowed to the income of small-scale food producers or farmers, not farmworkers. The labour question can be analysed within the dynamics of investors' land and labour demand. When massive land is needed (but not necessarily labour), there are implications to the food demand ability of the displaced and surplus agrarian population. When both land and labour are needed, it begs the question 'who does what' and 'who gets what' (Bernstein, 2010). These questions provide an excellent framing to examine the extent of women's incorporation into plantations, the windows of opportunity for increased income in comparison to men, and the implications for both food market access and their own production.
Feminist agrarian political economists, building on the works of Marx, (1977) have in the past decades, highlighted the centrality of women's work in the household economy, and even in recent times, their incorporation into the capitalist production through wage labour (Fraser, 2014;Bhattacharya, 2017). Women generally spend more of their income and time on food crop production. As small farmers, their access to land is defined by their relationship to men who control land and other resources. Regarding land use, Doss, Summerfield, & Tsikata (2014) indicate that in the course of large-scale land deals, women lose their tenure security through both the land transfers and entitlements to remaining lands thereby leading to landlessness and unemployed poor women who can neither grow nor afford food. Thus the connecting string to land access and labour relations is the centrality of women's role in social reproduction whereby food provisioning is key and how capitalist land transactions and the relations of production can deplete women capabilities of maintaining their households (Fernandez, 2018). In this study, the concept of social reproduction provides analytical insight in examining the dynamics of dispossession, wage labour relations on the plantation and labour on family farms. It guides understanding into how dispossession affects women farmers and their sources of subsistence and the interrelated process that contribute to the reconfiguration of land access and entitlements.
In central and western Africa, while the production systems of export crops like cocoa have continued in the domains of smallholders, others including oil palm and rubber were central in large-scale schemes. Plantations have historical antecedents in European colonial rule in Africa. In the 1930s, many colonial administrations experimented with the modern food system as a means to advance their agricultural export policies (Amanor, 1999). In Cameroon, the first commercial plantations were established in 1907 by the Germans and continued under the French/ British regimes until independence in1960 when they were nationalised. The introduction of plantations in Ghana can be traced even further earlier to the Dutch influence in the early eighteenth century, and their establishment expanded from the late 1800s under the British colonial administration. Oil palm plantations, for instance, served and continue to meet the pressing demands of the chemical, cosmetics and pharmaceutical industries in Europe (Huddleston & Tonts, 2007).
Upon independence, both Cameroon and Ghana had inherited an economy dependent on cash crop exports, yet without the expected trickle-down benefits to the local people's food security. While at the same time, food importation has been significantly high in both countries. In Cameroon, some of these colonial plantations including Société des Palmeraies and the Commonwealth Development Corporation later called Société Camerounaise des Palmeraies (SOCAPALM), and the newly independent government took over Cameroon Development Corporation (CDC) for both exports and national food security goals (Carrere, 2010). In the newly independent Ghana, the state experimented with modernisation projects, and established new plantations, mainly for oil palm under compulsory customary land acquisitions from peasants. These land acquisitions occurred mainly in the southern part of Ghana where land markets were more developed and lands more fertile than in the north, for the cultivation of prominent tree crops including oil palm, rubber and many other food crops 4 . Many of these large-scale land acquisitions occurred in the 1970s the food self-sufficiency vision dubbed 'operation feed yourself''-ideologically in line with the import substitution industrialisation policies of Ghana's first President, Kwame Nkrumah (Amanor, 1999). These developments were reinforced by some success stories of plantations in neighbouring Cote d'Ivoire and with funding support from the World Bank (Akoto, 1987). At the height of liberalisation ideologies and low success rates of state interventions, between late 1980 and 2000, the private sector became prominent. These transformations advanced at a time when international debates, foreign policies and development assistance sought to tackle poverty and economic stagnation in developing regions by setting conditions for alternatives development strategies-primarily centred on markets (Daniel, 1983). For instance, the Cameroonaise de la Banane plantation was sold to French-owned Plantations de Haut Penja (PHP) in the late 1980s and the privatisation of the Ghana Oil Palm Development Corporation since 1995. Both countries have seen a decreasing role of the state in existing plantations schemes. At the same time, newer ones are also driven by the private sector with the state playing enabling roles through supportive investment regulations and agricultural policies.
Even though the peasant production has prevailed alongside plantation schemes in both countries, successive governments have sought to promote food self-sufficiency and rural development through a transformation of the existing production system to favour mechanised, modernised and large-scale production schemes (Gyapong, 2019). Since the 2007/2008 food crisis, we have seen the resurgence of capitalist investments in plantations schemes in Africa, not only by national investors but from transnational corporations and agribusinesses operating independently or interlaced with host governments. In Cameroon, hundreds of thousands of land hectares have been acquired by transnational investors for oil trees like palm and rubber and forest logging 5 . Investors in Ghana seem to have more diversified production, engaging in various food crops and forest produce. Oil palm in particular, which serves domestic food markets in Ghana and Cameroon, is also of material flexibility in its multiple uses for food, feed, fuel and industrial material makes it attractive to investors because even as a single crop, it supports the diversification of investment markets (Borras et al., 2016).
This rush for land and plantation investments have spurred debates broadly between critics, especially agrarian justice organisations such as the radical anti-land grab campaigns of the food sovereignty movement and optimists who support and envision responsible agricultural investments in Africa (Deininger & Byerlee, 2012;Jacobs, 2013). Interestingly, the investment incentive essential to the implementation of the SDGs provides a layer of legitimation for plantation projects that are deemed to promote 4 While land tenure is mainly customary in Ghana, the development of land markets in southern Ghana is strongly linked to the advancement of cash crops like cocoa, oil palm, rubber, and forestry products like timber which are supported by more favourable climatic conditions in the south than in the north where there is prolonged dry conditions. Nonetheless, there is also a growing peri-urbanization, commercialization and interests in non-traditional export crops which are changing the dynamics of land tenure and increasing land expropriation in northern Ghana (see Yaro, 2012 food security. The zero hunger goal centralises production, which requires increased international cooperation. Also, discussions have arisen around Sustainable Development Investments (SDIs) aimed at steering 'foreign direct investment towards the achievement of the SDGs. Governments are, therefore, admonished to create enabling environments and appropriate levels of access to private and foreign investors for SDG-related sectors' (UNCTAD, 2017, p.3). In Cameroon, the government policy of 'Second Generation Agriculture' following the 2011 National Agricultural Show and tax breaks for agro companies target agricultural modernisation intending to increase employment, productivity and food sufficiency. Ghana, under President Akuffo Addo, has recently launched the flagship policies dubbed ''One District, One Factory" and the program for ''Planting for Food and Jobs" to provide agricultural investment incentive schemes, tax reductions for agroexport businesses and financial institutions that support agribusinesses. These projects do not only create an enabling environment for agricultural investments but are explicitly linked to achieving the SDGs (Government of Ghana, 2019). Indeed, we do not disregard the numerous and ongoing pressures on investors to consider global climate change, rural land rights and family farms in Africa (Graeub et al., 2016). Plantation agriculture, when well-thoughtthrough and executed within the framework of promoting equal opportunities for income and food, could be an essential pathway to development in Africa. In principle, the prospect of such benefits emanate from the direct and spill-over effects through access to farm and off-farm employment opportunities, increased income to purchase food, technology transfer, the extension of rural infrastructure, and improved access to basic services (Byerlee & Deininger, 2013;Deininger & Xia, 2016). However, such prospects rarely materialised, and there is the risk that the all-encompassing intentions of the SDGs leave room for vague interpretation of sustainable development and a surface view of food security. Blesh et al. (2019) have highlighted this contradiction within SDG-2 whereby food sovereignty framing or small scale family-farming intentions exists alongside strong support for corporate interests. On the one hand, it raises fundamental questions on the internal coherence and logic within the goals, thus the feasibility of achieving the SDGs. On the other hand, it questions the extent to which plantations can displace women, small farmers, from the land and still enhances sustainable and stable access to quality food for all, especially women. The role women in agriculture vary widely by region, age, ethnicity and social class, yet women tend to dominate subsistence agriculture, non-timber forest products enterprises, conservation and marketing of agri-food products (FAO & African Union, 2018). In rural West-Central Africa, men are mostly engaged in cash cropping while women remain central to their households' food security through small scale farming. Even beyond production, studies have shown that household food security is not only influenced by total household income but the proportion of income controlled by women (Kennedy & Peters, 1992;Molua, 2012). Nevertheless, both statutory and customary land institutions as well as agricultural policies in Africa appear to benefit men more than women. The increasing privatisation and commodification of land consolidate men's control of land (Mariwah, Evans, & Antwi, 2019); governments' disproportionate support for cash and export crops indirectly skews income, land access and property rights towards men; and depending on the crops that are promoted and the models of production, present differentiated and gendered opportunities in wage labour (Hall, Scoones, & Tsikata, 2017). For instance, even in contract farming schemes, which are supposedly intended to incorporate smallholders incorporate and agribusiness schemes, reports show that female farmers are primarily excluded because they lack secure control over land, family labour, relevant resources needed for their effective incorporation (FAO, 2011).
Until the 1990s, Cameroon enjoyed considerable national food self-sufficiency with enough production of cereal, tuber crops, vegetable and fruit to satisfy domestic demand. Madeley (1987) attributed this success to the favourable climate and farmer-centred policies. The government, until the late 80s, pursued a comprehensive agricultural policy that motivated women in food production. In the past three decades, Cameroon has seen the collapse of many such smallholder support schemes and high costs of living. A growing climate crisis in the Savanna and Sahel agro-ecological zones and Boko Haram insurgency in the Far North all which put additional pressures on women's access to land (Bamenjo et al., 2019). As export-led agricultural plantations continue to expand driven by their speculative motive in an effort to control as much land as possible, what is left become subject to competition between cash and food cropping; and farmers and pastoralists; locals and refugees. Ghana enjoys a fairly more stable development. Between 1991 and 2006, extreme poverty was reduced from 36.5 to 18.2 per cent, and by 2013, the country had attained the Millennium Development Goal (MDG) target of halving poverty and hunger by 2015 (FAO 2015). Ghana is almost self-sufficient in maize production (Yaro, Teye, & Torvikey, 2017). However, the same cannot be said for other major cereals like rice as well as tomatoes which can be produced by women if they are adequately supported.

Study methods
This is a two-country case study that draws on findings from qualitative studies conducted on plantation investments in Cameroon and Ghana between 2014 and 2019. Although fieldwork in both countries is slightly different in terms of scope and approach, the objective, focus and technique used for data collection are fundamentally the same. That is: to understand the dynamic, processes and implications of LSLAs on women and food security by interviewing actors, including corporate management, government officials and local community members/women. While 148 interviews (78 from Cameroon and 70 from Ghana) provided a panoramic country view of the drive for LSLAs and how it is affecting women and national food production, the additional survey of 200 people in Ghana underscore the detail experiences of affected communities.
Unlike the Ghana case which was an in-depth study of a major oil palm company in a district, the Cameroonian data was obtained from horizontal research on four major companies in the North, South West and Littoral Regions cutting across the savanna, forest and maritime agro-ecological zones. The agro-companies surveyed were the Sithe Global Sustainable Oil Cameroon, Société des Plantations du Haut Penja, Cameroon Development Corporation, and Ndawara plantations. They produce oil palm, rubber, tea and banana, and similar to Ghana, they acquired forest and farmlands in rural agrarian communities in Cameroon. The 78 interviews conducted between March and December 2014 included 38 rural women, 4 top management officials from different plantation companies and 36 public and traditional authorities. Respondents were purposely chosen-to include people affected by the land acquisition and other key players in the transaction processes. Although separate interview guides were administered to the different groups of participants, the questions were generally framed to understand the process of LSLA and actors, the impact of the activities on gender relations, food security, and local development. Field notes, data from portals/reports of government and other development partners have been integrated into our thematic analysis.
In Ghana, the fieldwork was undertaken between February 2018 and March 2019. The main company involved in this study Lotsmart N Fonjong and Adwoa Y Gyapong World Development 138 (2021) 105293 is the Sithe Global-Sustainable Oils Ghana (SG-SOG)-an oil palm agribusiness firm that is affiliated to one of the Cameroonian firms presented in this study but currently operating in Ghana under new management, and now named as the Volta Red Farms. In 2008, the SG-SOG-Volta Red oil palm plantation was acquired from families of the Ntrubo clan in the Nkwanta South Municipality of the Volta region (Oti) of Ghana, spanning an area of 3750 ha at the eastern border of Ghana. The catchment communities visited include Brewaniase, Fankyenekor, Abuburuwa, and Dodo Tamale, but also a third of the farmworkers are migrant labourers from the region. More than 80 per cent of the employed farmworkers participated in a socio-economic survey that assessed their access to land, working conditions and other livelihood and demographic data. About 70 Semi-structured interviews were also conducted with farmworkers, including more than 50 per cent of the female workers, supervisors, and management. Aside from farmworkers who were affected by the land deal, dispossessed families who do not work on the plantation were involved in interviews and focus group discussions. Interviews were also conducted with traditional authorities, officials at the regional Lands Commission, the local government, and the Ghana Agricultural Workers Union.
Using thematic analysis, we were able to identify common themes from both studies. These themes have been reported in the findings while occasionally laying emphases on particular case study as emerged from the data. The subsequent section discusses these with a focus on the exclusionary processes against women in their access to farmlands; compensation and rents; and income from wage labour and how these affect women role in social reproduction, particularly regarding food provisioning in the aftermath of new plantation investments.

The dynamics of displacement, resource access and food production
Many rural communities in Africa rely on forest lands to produce food and fuel, but in ways that do not undermine the biophysical environment in the long run. As we observed in Cameroon, SG-SOC, CDC, PHP and Ndawara have placed restrictions on women's rights or access to the forest in Kupe Manengouba, Ndian, Mungo Fako and Ngoketunjia Divisions. Nontimber forest products (NTFPs) such as wild vegetable, bush mango, eru, ngabi oil, njansang and cashew have for centuries served as relevant sources of food and revenue for forest communities, especially women. In Ndian and Kupe Manengouba Divisions, both men and women were affected in this regard. One woman from the Nguti village recounted ''. . .we have been forced to abandon the collection of NTFPs because of restrictions into the forest. . .. the income we used to get from this activity is gone. . ." Another woman noted that one of the NTFPs rich in food nutrients oil, locally called ngabi oil extracted from the forest is no longer available in the market because the forest has been destroyed by plantations (Fonjong, Sama-Lang, Fombe, & Abonge, 2016). The impact of the loss of community rights to access these NTFPs did not fall on the villagers alone but also on urban consumers where the products became scarce and expensive to the average household.
Again, fuelwood is the most important source of energy for cooking in Cameroon for 70.3 per cent of the population and 72.2 per cent for female-headed households while gas, coal and oil stove represent the rest (World Food Program, 2017). Following the enclosures of forest lands, women's access to fuelwood is restricted, and the situation demands that they cover longer distances across extensive plantation farms. In the South West Region where women use much energy from fuelwood to process cassava into flour 'garri' (important staple food for both domestic consumption and markets), scarcity of farmland and fuelwood in these traditional zones of production have indirect effects on its quantity, quality, price; and of course, on poverty alleviation. In general, the replacement of natural forest around villages by new plantation cover has made the sale of fuelwood a lucrative business for some men and youths who go far into the forest or cut down abandoned aged rubber trees from surrounding rubber plantations. Where women cannot afford the cost of fuelwood, they resort to cutting down trees planted as windshield against soil erosion or wind destruction around banana plantations, breeding constant tension between the villagers and plantation authorities. Table 1, which provide an overview of the land use activities reported by 68 respondents (women, men and traditional authorities) prior to the acquisition of their land for plantations reinforce studies that have highlighted the gendered differences between men and women's land use. As Table 1 indicates, corporate land takeovers affect men and women differently. Regarding changes in land use activities, women pay a heavier price than men in the area of food security because twice the numbers of women than men use the land to grow food while men are mostly affected in their cash crop production. Expressing her displeasure against the enclosures, the president of Ndian Women's Forum; a local initiative group-''. . .the women are not happy at all. . . our land has been taken. . .we no longer have farmland to grow food to feed our children. . . women now travel long distances, to cultivate in areas that do not have roads. . .Food is scarce today, and the women are anxious. . ." In the locality of Pendamboko, where the CDC is expanding its plantations, most women have resorted to farming on marginal land.
In the Ghanaian case, there are some variations in the dynamics of dispossession and displacement. It is primarily linked to the relative land availability, access to alternative lands, and the predominance of very peasant scale farming, sharecropping and mutual support farming systems. In total, the affected Ntrubo clan has been grouped into fifteen extended family groupings for rent distribution. Among these groupings, three extended families were completely dispossessed, with no access to alternative family lands-although this number conceals the pockets of nuclear families (among the thirteen) who lost all their lands and the actual count of the numerous affected generational groups (exceeding 500) in these high household population densities. Attempts by the company to provide alternative lands to those dispossessed has come to a standstill due to the remoteness, size, and tensions between affected families. Even though some were not entirely dispossessed, they still face challenges with access-size, distance and suitability; and control-the autonomy of production, and modified rules of entitlements to these alternative lands. In this process, women were immensely affected. Women in these communities take pride in their ability to farm independently, no matter how small, and do so alongside their family farms or intercrop in them (see also Amanor-Wilks, 2009). While cash and export crops like cocoa are predominantly cultivated by men, similar to the Cameroon context, food crops like maize, cassava, plantains, vegetables and legumes are vital to women's daily household reproduction. They do so, not only as a household food stock but also for some economic benefits through processing and trade. In the light of the land takeover for the establishment of the plantation, women from migrant sharecropping households and indigenes those without access to suitable and proximate alternative family lands bear the brunt of food insecurity. In completely dispossessed families, they now have to acquire land from other members of their clan on sharecropping terms-a practice that is conventionally reserved for migrants and settlers but now applies to rent-receiving families. For these dispossessed natives, this severely restricts the kinds of crops grown, the scale and share of the harvest. Most women in such families, under restricted access, have limited abilities to engage in their own farming. While some have resorted to only intercropping, others have been compelled to Lotsmart N Fonjong and Adwoa Y Gyapong World Development 138 (2021) 105293 seek other lands further away from their new family sharecropped lands thereby affecting the extent of family labour support, thus production. In words of a dispossessed woman, on their own land, ''I used to farm vegetables, corn, groundnuts, maize and my husband also cultivated these [above] in addition to maize and rice and plantains. I usually marketed my husband's produce for him, and I keep my own income. After the takeover, I now sharecrop 1 acre of rice, groundnut and corn and this land is distant from my husband's sharecrop land" 6 For women with access to alternative family hands, they complained about how long distances and unfavourable soils affect their production. Not necessarily like the case of marginal lands in Cameroon, their main concern relates to how the unsuitable soils dictate what can be cultivated, particularly when they do not support cassava and maize production. In effect, the challenges faced are not limited to dispossession in its traditional sense; rather, the other social dynamics of access to alternative lands. The narrative below of a completely dispossessed family head illustrates the dynamics of food insecurity: ''We used to live and farm on the acquired land. We had much cocoa, 50 trees of oranges, cola, oil palm and other food crops. We had allocated portions of the land to migrant sharecroppers who farmed cassava and corn. We often faced the risks of bushfires which is one of the reasons we gave out the land. They promised us land at Abuburuwa, but that land is 'not sufficient, so we have been farming on my wife's family land. Now we farm 1 acre at a time, but previously we could do more-up to 2.5 acres because of the family labour support, and also that land was much bigger, more fertile and fresh. My two wives used to do their individual farms alongside and also intercropped his farm with vegetables, but they do not farm any more. When we lived on the farm, we benefited more; we did not have to buy meat. Now we cannot get many things, we buy more, and I cannot labour on the plantation for income because I am old 7 ." Many land deals neglect or do not take into cognisance the water rights of affected communities and are often done in areas where the investors have full control to access to water (Mehta et al., 2012). Fonjong and Fokum (2015) observed that water is a crucial consideration for large land acquisitions. In Kupe Manengouba of Cameroon, the natives accused SG-SOC of stream pollution and reducing the number of potable sources of water. In Fako, many local residents have struggled with access to potable water for eight years, yet the surrounding CDC plantations are well irrigated. Besides, in many plantation communities, rivers, streams, and vegetables have become unfit for human consumption because of contamination and pollution from the aerial spray of plantations crops. Several cases of where local water sources have been abandoned were reported by communities living around the PHP plantations. Women and children have to walk further afield daily to collect fresh water for their households. Time and distance covered to fetch water and fuelwood are not only a burden to the women but also economic time not well spent that could have been invested in other productive activities to support their families or boost the economy.
The growing landlessness in Talangye village (Kupe-Managouba) has provoked high incidences of rural exodus while in other localities; it has led to forceful and illegal occupation of parts of the CDC land concession, triggering constant tension between the villagers and the corporation. The situation is different for displaced women in Penja, who have been affected by PHP. Compelled by their difficult situation, PHP has allocated just about 60 ha of the uncultivated parts of their concession to over 1500 organised women to farm. However, the 'opportunity' is accompanied by restrictions on the types of crops and seasons they can cultivate, as the company reserves its rights over these plots. To the interest of the company, the women are permitted to grow only vegetables, legumes and cocoyams-crops that do not use up excess soil nutrients and like beans, a good source of nitrogen. The company however prohibits the cultivation of maize and banana for their own interests; to control the spread of transmittable diseases that may affect the yields of their plantations. Underpinned by similar capitalist logic, the farmworkers in Ghana were temporarily granted land access to the plantations purposely for weed control.
As investor interests in plantations continue to rise, remaining lands are subject to competition between users. In Boyo and Ngoketunjia, (North West region, Cameroon), the expansion of the Ndawara tea plantations has amplified the number of resource conflicts recorded and the destruction of women's food crops by pastoralists. Again, women and other small farmers are always the losers when such conflicts are adjudicated by administrative authorities which in most cases are either owners or corrupted by wealthy cattle owners. Over 51percent of women interviewed reported that their lands had been taken over by the Ndawara Plantations; 49 per cent had been evicted by pastoralists, without commensurate compensations.
Similarly, there is a growing resource competition between dispossessed natives, settler sharecroppers and migrants over the few suitable and accessible lands in plantation areas in Ghana. The survey of 200 farmworkers showed diversity in the longevity of sharecropping arrangements ranging from under a year to approximately twenty years, variations in the levels of autonomy of production, and share agreements. First, this limited land access is modifying farming practices, leading to shifts from land rotation and shifting cultivation to more intensified land-use patterns. Second, and significantly, is the changing social relations between landowners and tenant farmers or sharecroppers. Historically, the allocation of lands to settlers and migrants for food crops formed part of the moral economy around subsistence. As such, sharecropping arrangements for food crop cultivation are usually not strictly defined and often depends on the gender and age of the tenant, years of tenancy, the types of crops grown, and other social relations. While this system is not entirely erased, this culture is seriously threatened, thereby affecting women who often farm tiny sizes of land (under an acre). Third, the limited access to land and food, and the growing cash economy is also contributing to competition and contentions at the household level; among husbands and wives over the distribution between household food stock and a market share. This is because families have different dynamics in the usage of their family farm yields. For instance, if the woman is responsible for the protein (meat and fish), then ideally, she expects to market a vast proportion of the yield of her independent farm, while they depend on the family farm produce for important staples.

Commodification and the reconfiguration of Women's land entitlements
the rise in investments in rural farmlands and their commodification alter land relations in complex ways, Crop based compensation creates a false dichotomy between 'cash' and 'food' crops and has gendered implications. Mainly, traditional cash crops locally referred to as 'property' crops-cocoa, oil palm and citrus were considered for compensation and in some cases some wild crops. They are mostly cultivated or maintained by men. Most women intercrop vegetables in their family farms, but vegetables are often undervalued, thus uncompensated. For these systemic reasons, coupled with patriarchal practices of resource distribution, women were excluded from compensations. In Ghana, the women of both native and settler sharecrop households have been affected by the non-compensation of food crops. At the initial stage of the land acquisition, they were given a short period to clear their foodstuff from the land. This situation created temporary oversupply and wastage of some food crops like cassava, yams, and vegetables while others made losses from crops that were not due for harvest. Cassava, cocoyams and other tuber crops which are sources of food and cash regenerate over time. Therefore, most women complained about the lack of access to the plantation to harvest the remnants. However, as observed during fieldwork, farmworkers are occasionally allowed to harvest some of these remnant food crops, a situation that sometimes creates tensions between them and the dispossessed landowning families.
The sudden commodification of land has also generated a new system of control, inequality and the reconfiguration of existing land relations (Berry, 2017). Whereas having a land title is not pre-condition for compensation in Ghana, in Cameroon, compensation is often paid to those displaced from their ancestral land. The Cameroon Land Ordinance requires that compensation can only be paid for crops and other developments on land that has been registered or titled. In the light of this statute, women can only receive compensation for crops destroyed since mostly they work on national land for which they have no titles. Even in such circumstances, NTFPs are not compensated, and patriarchal institutions often discriminate against women. In both countries, base compensation rates emanate from obsolete government rates, unknown to farmers. The loss of future crop yields and displacement are not factored in compensations. In both countries, male household heads are gained access to compensations under a flawed trickle-down assumption (Udry, 1996). Seething against such discriminatory practices, a local government official in Cameroon expressed ''. . .there is no respect for women's land rights even as users. . ." Women/widows who receive direct compensation receive as derivate users and not on their own rights. In some other cases as in Ndian where most land deals are shrouded in secrecy, women and other victims received no compensations and the vast proportion directed to elites, traditional rulers and male household heads. In Ghana, the SG-SOG lease was acquired on a rental basis for a period of 50 years. The annual rent of 5USD per hectare is in itself a manifestation of the inequalities of market-led land transfers in remote rural communities where deprivation, poverty and limited access to information affect their decision making (Akram-Lodhi, 2007).
Given this meagre rent, women, including those who were utterly dispossessed and those previously entitled have now been excluded by the family representatives from such monetary benefits. The rent list is premised on the company's survey of these customary family lands. It has become an improvised land tenure 'security' tool and a benchmark to forcefully 'resolve' pre-existing land contestations. Under the patrilineal land inheritance system that is practised among the Ntrubos, there are several layers of derived and direct land access and control for women. However, daughters, sisters and wives of family heads/representatives have been almost entirely excluded from access to rents. On the rent list, their entitlements are buried under the names of their brothers, uncles, families, and husbands. In families where brothers have living sisters, the rent list has captions like 'so and so brothers' or 'family land' where men control the rents. From Fig. 1, we show how the land access system has been reduced to benefit a few groups of people, mostly men. Interestingly, in several instances, the previous generation women-especially living paternal aunties, of family heads, or their successors have managed to secure access rents as compared to the current generation women. Such a situation has implication not only for access to income to purchase food, but also very significantly the future land use and ownership rights for these excluded women.

Gendered employment, income disparities and food security trade-offs
Proponents of plantation investments do so not only from a productivity lens but also the jobs they create. To a large extent, the cases from Cameroon and Ghana corroborate with emerging evidence on the limited and unequal job creation opportunities for men and women, and the unfavourable conditions of labour (see Behrman et al., 2012;O'Laughlin, 2017). Different crops types create different opportunities for men and women. In these two country studies where oil palm and rubber are predominant, gender represents a key factor for differentiation in terms of tasks, contract type, and skill. Some jobs were created, yet investors', like CARGILL's projection of 10,000 jobs in Cameroon are not only mere imaginaries but also less beneficial for women (Palm Watch, 2013). Though 51 per cent of jobs created by CDC in Mungo Division went to women as slashers, carrier and harvesters, they occupied just 4 per cent supervisory jobs dominated by men. A local villager elder in a slightly exaggerated tone decried that ''. . .All those working on the plantation are migrants. When jobs are to be provided to the natives, there are mostly unskilled jobs. . .' However, many villagers do not have relevant skills for white-collar positions. Majority of these jobs were also seasonal.
The SG-SOG-Volta red concession in Ghana, employs approximately only 50 women and 150 men on the 3750 ha of land. Brewaniase, the plantation site is the second-largest community in the municipality and home to at least 1500 women within the labour force (2010 population census), excluding others in the catchment villages. Indeed, in other crop sectors like horticulture and fruit production in the Volta Region of Ghana almost two-thirds of the farm labourers are women because of the apparent gender 'friendly' tasks like mulching, cutting, packing, and fruit picking (Tsikata & Yaro, 2014). This contrast with the oil palm sector, where the main tasks that women are engaged in are loose (palm nuts) picking while men can have opportunities in twelve different tasks . Nonetheless, across crop sectors, men take up the most highly skilled and administrative jobs due to their better access to education and training. Sixty per cent of the women farmworkers have up to primary level education or none, whereas that of men is just about 30 per cent. Another issue is the casual and insecure nature of job contracts, which happens to affect women the most. Advocates of labour casualisation stress its importance for income diversification and even the potential prospects for labourers to transfer new technologies to their own small-scale farms (Deininger & Xia, 2016). However, it has become a means by which investors cut costs, evade worker safety and insurance responsibilities, and prioritise their business operations over many workers (Arnold & Bongiovi, 2013). Women farmworkers in both countries have insecure employment. In Cameroon, many are underpaid, without employment contracts and abusively laid off.
Nevertheless, per the laws in Ghana, women work under biannual renewable contracts that have no benefits for stability. Seven out of ten men are casual workers, and the figure goes up to 90 per cent for women. During the off-peak oil palm season which also doubles as the dry season, more women than men are laid off due to the gendered differences in opportunities to switch from harvest related tasks to general farm maintenance, e.g. from harvesting, weedicide application and loading to pruning, fireguard, irrigation among others.
Due to the low incomes and the gendered childcare responsibilities, most women in both countries are unable to save enough to purchase sufficient food in the lean season. However, unlike in Ghana where more than 80 per cent of the workers maintain their culture of subsistence farming and also to make up for their low food demand, the situation is different for their counterparts in Cameroon who work far longer hours in the plantation that legally required Over 90 per cent of female farmworkers in Ghana cultivate up to an acre of land. Most farmworkers maintain their minimum stock of maize, rice, sweet potatoes and cassava for their households even though most women indicated a reduction in their farm sizes due to the wage work. They are also able to maintain access to food even when they are lay off. Unlike maize which requires complete clearance upon harvest, cassava as a tuber crop can be uprooted as and when needed relatively throughout the year-explaining the slight variation in their access to maize and cassava in Table 2. The major cassava season, however, coincides with the peak period of the oil-palm where women can be guaranteed more stable and relatively higher incomes as compared to income from cassava processing in unstable markets. A worker expressed 'foodstuff has no price anymore, so you may end up losing it all or just eating your own produce'. Most of them, who are of no formal education, paying their children's school fees and striving for some economic independence, need the income from wage labour to maintain their livelihoods. They, therefore, are compelled to divide their time and energy between farm and farm workdepending on their household demographics and family labour dynamics. In this process, it is the legumes and vegetables that are mostly sacrificed in place of the staple starches. Constant struggles and food trade-offs food characterise the affected people.
Ama's story below, exemplifies the struggles women go through to maintain their basic food security and livelihoods while working on plantations.
Ama, a 52-year-old woman without formal education, worked for seven years at SGSOG-Volta red, received an award for her dedication, but remains a casual worker. She is in a polygamous marriage. She migrated from another district with her husband. 'We have no option but to follow our husbands, and then we labour'. She has been assisting her husband on their family farm and also sharecrops maize, cassava and cocoyam of about 6 ropes 8 . However, in 2018, she could not farm her 6 ropes because she was worn out from the plantation work, and does not get ample family labour support because her kids have migrated or are in school. She has been sharecropping since 2015 and now relies a lot on wage labour. In June, she requires time on her own farm to plant corn and cocoyam, but because it clashes with the palm season, she hires labour. 'I only go there on Sundays and sometimes Saturdays'. When there is a good harvest, she sells some of the cassava to support her children's education, but she subsists more than sells. When there is a poor market for her husband's yam produce, she takes responsibility in the payment of school fees. 'We help our husbands, but when there is money from yields, you have to 'fight' to get some share; 'as wives, we pay for the soup, mill the corn, and cook the food, and he has used his income to build a (mud) house in our hometown. I will not stop the work now unless I do not have the strength to work anymore'.
In the light of the challenges from farm size reductions, food purchases, hiring of wage labour and others, it is expedient that women can earn liveable wages as well as decent labour conditions In reality, these 47 beneficiaries are expected to share the rents with their large nuclear families and other household members, but this is at their discretion, and the rent is often controlled by those whose names appear on the list.)Source: Fieldwork, 2018. of work in general. Systemic and structural reasons make this impossible for women. As earlier mentioned, the difference in task opportunities produces income inequalities. With women being taken away for and women spending more time for less income in the plantations than on their farms in Cameroon, the situation of basic foodstuff in most markets across the country are on the deficit. This has encouraged food price spike and the importation of food items like tomatoes, wheat, vegetable (Table 3), all previously grown locally in abundance at the expense of high interest foreign loans. Table 3 shows that national food production fall short of demand, thus creating high deficits. Cameroon obtained a loan of US$19.2 million in 2010 from IFAD to improve food security in rural areas by supporting the production of rice, onion or reduce import (IFAD, 2010).
In Ghana, during peak season, harvesters (men) can earn more than four times what women can earn in that period as is the case with the men who constitute the majority in better-paid plantation jobs in Cameroon. It is worth noting that these wages are almost 50 per cent higher than the nation's minimum wage but calculated on a piece-rate basis, which is quite complicated. Compared to the local smallholder wage labour rates, the wages are relatively minimal, but given the limited options, workers are attracted to the regularity of income. When women are made to do slashing, for instance, it is often difficult for them to meet their targets for a full daily wage. Even the men slashers are also among the least earners for the same reasons, and both management and the men are not in favour of women slashers' demands for gendered considerations. The women must keep working in these plantations since their low education, skills, and capital limit access to stable offfarm livelihood diversification alternatives. The low income is associated with borrowing and buying of groceries on creditsmany shop owners corroborated these claims. The debt rate is 40 per cent among the plantation workers in Ghana, and for many who desire permanent contracts; the main reason is to allow them to access from banks through the company-regardless of the extremely high interest rates (60 per cent) to be paid within 18 months.
Work on the plantations is time and energy demanding. In both countries, farmworkers are transported in buckets of trucks as early as 5:30hrs and arrive as late as 18:00hrs. A woman in Ghana lamented, ''when we close at this time [17:00hs], when will we get home and fetch water? What time will we cook'? This affects their own eating habits, as well as their ability to prepare nutritious meals for their households, especially children. These struggles can be summed in words of Epey, a Lipenja village woman in Cameroon who stated ''. . .plantation companies have taken our land, and this has led to a reduction in our income. . ..women used to work in farms, produced enough food for the home and to sell,. . .Nevertheless, now they work in plantation nurseries with insignificant pay. . .they work for long hours in the nurseries and do not have time to work on their own farms. . .. They have left their farms for nurseries. . .in return for hunger, poverty and hopelessness."

Conclusions
In this paper, we sought to examine the food security implications of the large-scale land acquisitions and the impacts of the promotion of plantation agriculture on women in Ghana and Cameroon. In highlighting the importance of the agricultural productivity and incomes of small-scale food producers in achieving zero hunger (SDG-2), the paper underscored the need to integrate the gender dimension (SDG-5) in current plantation-led LSLAs by ensuring that women's interests are mainstreamed into important decisions that take away their land and farms which have been the sole source of family livelihood for ages. Thus, gender equality cannot continue to serve as window dressing for the pursuit of corporate interest in Africa. It behoves on agro-companies to provide decent working conditions; ensure gender equity in labour opportunities; and invest in the local food economy, if the goal is indeed to achieve zero global hunger for all.
Plantation investments can achieve local food security when it is intentionally designed to benefit the rural working poor, but this hardly happens in Africa. Following the evidence from Ghana and Cameroon, the land acquisition of smallholder farmlands and forest for plantation has led to displacement and relegation to marginal lands, reduced food production and exploitative wage labour conditions, thus low incomes (see also Yaro, Teye, & Torvikey, 2017). As a result, Cameroon, which was hitherto the 'food basket' of Central Africa, is now a food importing country; creating a situation that does not only drain its limited liquidity but also compromise food security particularly for the rural poor. The findings also reveal dynamic and context-specific food security trade-offs among different groups of rural women, taking into consideration their class, generational and household composition, and social statuses. For example, in the Ghana case study, reduced food production is further worsened by the fact that the vast majorities of the population are settlers and thus produce under sharecropping arrangements. The food security situation in Ghana is not akin to an extreme or chronic hunger situation, because they still strive to continue their own farming, no matter how small.
Nonetheless, for women farmers and farmworkers who lost their land, their limited land access means lesser activities in intercropping nutritious legumes. Even for the very few who work as casual labourers, the intensity of their job also influences their choice of crops on their own farms-often unable to effectively cultivate labour demanding but highly nutritious foods. The absence of gender supportive land and agricultural policies in Cameroon has left poor landless women who previously depended on the commons to grow food and harvest NTFPs for family consumption and income, food insecure and disillusioned. Unfortunately, faced with rising food prices and poor plantation working conditions in a society where women still maintain the burden of feeding the family, women exodus into cities has further complicated the national food situation, poverty and food importation. The dynamics of the food insecurity situation in both countries is a reflection of two main systemic issues. First, is the incoherence of the presumed local development logic that underpins many such investments. Capitalist development survives on accumulation and expanded reproduction, thus what appears rather logical is that investors would hardly compromise on their profits, especially in such contexts of fragile governance. As investors push for formalised and secure land access to secure land for their own production, such processes exacerbate existing inequality within the existing customary tenure systems, enclose land resources including water, and demand cheap labour in remote agrarian communities where livelihood struggles make plantation jobs attractive to both men and women. For example, in Dzansi, Lagakos, Otoo, Telli, & Zindam (2018) paper purported to be a guide book to investors interested in the recent 'one district factory' policy, instead of problematising the employment situation in Ghana, they list the availability of cheap skilled and unskilled as advantages to be harnessed. The narratives of the struggles of Ama and Epey and others highlighted in this paper provide an excellent critique to simplistic livelihood diversification discourses that promote casual wage labour in plantations.
Coupled with the challenges above, are the policy, institutional and governance inadequacies. Global and national level policy debates continue to highlight the need for better regulatory frameworks to check and mitigate the potential negativities of largescale agricultural land acquisitions (see . Examples include the several voluntary guidelines on land investmentsmany of which remain on shelves in government offices, premised on equal power relations between investors and communities, and the fact they are already too late and not binding. In both countries, the labour laws are designed for industries and not for agricultural workers, and the minimum wage is not a liveable wage. Labour and investment policies are woefully inadequate to address the needs of agricultural workers, let alone the particularities of women; even the ones in place are hardly implemented. Although existing legal provisioning support labour unions, casual workers in rural areas face numerous organisational and capacity challenges. Resonating with Peters (2013, p. 548), perhaps it is only appropriate that women and farmers in Africa are provided with the kinds of support that enable them to make more productive use of their own land, water and labour. Rural women in Ghana and Cameroon cannot fight hunger and poverty without land no matter the here-and -now -opportunities that plantation agriculture may create for them. A majority of women/small-scale farmers in Cameroon did not have any legal claims to land despite their long years of access. It underscores the fact that achieving SDG 2 and SDG-5 requires gender-inclusive land and agricultural policies that recognise, protect and empower small farmers (especially women) and create opportunities for sustainable agriculture. It is therefore crucial that national food self-sufficiency take precedence over capitalist-driven export-oriented schemes which often siphon off local resources without significant trickle-down benefits to them.
Further quantitative and longitudinal studies should explore and assess the implications of women' land rights and gendersensitive practices in plantation agriculture on food security and gender equality in Africa. Such research is needed to provide broad perspectives and deepen our understanding of the dynamics of impacts when capitalist investments increase in rural and peasant economies.

Data statement
Due to the sensitive nature of the study in some of the localities studied, those who participated in the survey were assured raw data would remain confidential and would not be shared.