4.1. Socio-economic characteristics and cocoa agroforestry adoption
The findings of the current work revealed that 93% of farmers in the “Espace Taï” are migrants from other parts of the country as well as from other countries. This large proportion of non-local population is explained by the massive migration of new farmers attracted by the cocoa boom and prior cocoa farmers who moved from the eastern part of the country to the centre-west and then to the south-west parts searching for fertile endogenous soils in the primary forests. This massive movement which took place initially in the 1980s was termed by Ruf and Varlet (2017) as the cocoa pioneer front displacement. A second vague of cocoa farmers' migration occurred during the socio-political crisis between 2002 and 2011 (Ruf and Varlet 2017; Ongolo et al. 2018). The development of cocoa farmers' settlements in the vicinity of the TNP results in many instances in the encroachment on gazetted forests and protected areas, such as in the Mount Péko National Park and the Goin-Debe forest reserve (Ousmane et al. 2020; Kouassi et al. 2021a).
The likelihood of adopting cocoa agroforestry in relation to the origin and ethnicity of the farmers could be due to the fact that most of the migrants had prior experience with cocoa farming, including the benefits of keeping other tree species (shade, food, fuel and medicinal uses) in their cocoa orchards (Herzog 1994; Adou Yao et al. 2016).
A typical farmer is an adult, married, and illiterate and this has already been observed by Assiri et al. (2009). It has been shown that the education level positively improved the propensity to replant and adopt agroforestry (Agkpo et al. 2002).
4.2. Current cocoa farms characteristics and cropping practices
The average cocoa farms size was of 5.2 ha, which is lower than the 6.3 ha observed by Assiri et al. (2009) in the same area. The shrinkage of the cocoa farm size over time could be explained by the reduction of forest areas, and the farm inheritance between several heirs. Also, because of the different challenges encountered in cocoa farming and for diversification purposes, some farmers convert part of their cocoa orchards into alternative crops such as rubber and oil palm (Aguilar et al. 2005).
Most farmers used unimproved planting material and other ineffective cropping practices leading to low average yields as already documented by Assiri et al. (2009) and Balineau et al. (2016), which is also due to climate change, and pest and disease pressure in the region. Good agricultural practices such as pruning, suckering, cutting, and sanitary harvest are critical to boosting cocoa productivity. Our findings showed that these practices are not timely and effectively implemented by most farmers, although doing so could reduce pest and disease impacts and increase productivity. The cocoa yield of 376.3 kg/ha found in the landscape is lower than the productivity of 2,100–2,400 kg/ha identified by research stations with hybrid plant materials (Kébé et al. 2009). The estimate of yield found in the study is close to the average of 400 kg/ha found a decade ago in the same region (Aguilar et al. 2005; Assiri et al. 2009). Low yields may be explained by the high use of non-selected plant materials, non-compliance with technical outreach, and the lack of farm maintenance as well as climate change, pest and disease impacts in the region. Cocoa pods are harvested monthly with the highest levels from September to December. Similar findings were observed in Ghana with high production occurring from October to February (Schroth and Ruf 2014).
Maintenance practices including pruning, suckering, cutting, and sanitary harvest are critical to boosting cocoa productivity. Our findings showed that farmers do not carry out timely and proper maintenance practices, which negatively affects cocoa beans' quality. These practices reduce pest and disease impacts and increase productivity. Pruning has a positive effect on yields (Balasimha, 2009; Riedel et al. 2019) and can lower chemical fertilizer and pesticide costs. The low adoption of phytosanitary harvests (harvesting of infested and mummified pods, mistletoe, and water shoot) could reduce the productivity and the yield of cocoa plantations. Ndoumbe-Nkeng et al. (2004) highlighted that pod removal and phytosanitary harvests practices could enhance by 50% cherelles' production and the number of ripe-healthy pods. Sanitation pruning reduces the contagion of black pod disease caused by Phytophthora spp. (Ndoumbe-Nkeng et al. 2004). GAP recommend using pesticides and fertilizers to improve productivity and sustain yield (Kébé et al. 2009; Koko 2014; Ruf and Zadi, 1998; Siapo et al. 2018). The non-adoption of fertilizers can be explained by the high price of chemical fertilizers. In contrast, some low-income farmers use chicken droppings and compost which are costless as alternative options to boost the growth and the development of cocoa farms (Ruf and Allagba 2016). However, one of the challenges cocoa farmers faced is the financial cost of these maintenance practices which remain time-consuming. The low yield cannot allow farmers to drive more investments in soil inputs and other maintenance practices. With the scarcity of casual workers and funding coupled with the relatively huge mean size (5.2 ha) of cocoa farms, it would be challenging for farmers to perform solely these maintenance activities.
Although farmers are used to weeding manually their plantations, half adopted the use of herbicides in order to be overwhelmed by the abundance of weeds, the emergence of new and more aggressive species, and get accounted to labor unavailability (Ruf and Allagba 2016). Herbicide usage by smallholder farmers is led by the lack of workers and the low yield which cannot afford to pay anyone to help them. Farmers use herbicides to reduce labor costs that affect their revenues. Konlan et al. (2019) highlighted that glyphosate application significantly reduced weed management costs and increased the yield of three-year-old cocoa compared to manual weeding. However, regular usage of herbicides and pesticides may reduce the quality of cocoa beans by increasing pesticide residues and other harmful substances in beans (Siow et al. 2022).
In cocoa orchards, mirid and black pod attacks targeting trees and pods are common and cause significant loss of productivity estimated at around 30–40% and 35%, respectively (Kéli et al. 2005; Ploetz 2016). In the same way, the prevalence of CSSVD can lead to complete orchards’ destruction (Kéli et al. 2005; Ploetz 2016). These pest and disease attacks could explain the low yield identified in the region.
4.3. Drivers of cocoa agroforestry adoption by farmers
Our study revealed that the adoption of agroforestry is driven by gender, the length of residency, and the number of cash crops grown by the farmers. In a previous study, Owusu and Frimpong (2014) found that cocoa agroforestry adoption depends on age, gender, and household size and increased cocoa yield and thus household incomes. The gender differentiation could be explained by the fact the large majority of cocoa farmers are male who have more income to reinvest in the additional workload for farm diversification i.e. agroforestry practices. Indeed, the limited land rights access of women leads to smaller land sizes and thus low income generated from these farmers as compared to the men. Similar gender differentiation has been observed by different authors in Nigeria, Ghana and Cameroon (Owusu and Frimpong 2014; Schroth and Ruf 2014; Adetoye et al. 2018)
Although farmers value cocoa agroforestry in the region, most orchards are intercropped with tree species other than cocoa at an average low density (< 20 trees per ha) confirming the full sun monocropping system common in the study areas (Smith Dumont et al. 2014). This low density is dominated by edible fruit species corroborating prior results that fruit and other food tree species were the dominant tree species in cocoa landscapes (Koko et al. 2013; Smith Dumont et al. 2014; Kouassi et al. 2021b). These low diverse agroforestry systems are comparable to the Cabruca systems in Brazil (Schroth et al. 2011), include also species that can contribute to climate adaptation and mitigation, soil fertility improvement, and reduction of disease pressure (Wartenberg et al. 2017; Blaser et al. 2018). These species can help cocoa farmers in diversifying their revenues and ensuring food security (Mbow et al. 2014), and to boost cocoa yield (Koko et al. 2013). The difference in cocoa yields due to cocoa agroforestry adoption might be attributed to farmers age, soil fertility enhancement and soil moisture in cocoa agroforestry plantations. Despite the importance of cocoa agroforestry, more than 50% of the farmers indicated that they removed tree due from their farms to avoid resources competition with cocoa trees and the spread of Cocoa Swollen Shoot Disease (CSSVD) following advice from extension services (Smith Dumont et al. 2014). As an example, species such as Psidium guajava and Cola nitida are commonly removed by farmers due to nutrient competition for the former, and host of CSSVD for the latter species (Smith Dumont et al. 2014; Sanial and Ruf 2018), although no well-documented proof exists to confirm these. This tree removal could be led by continuous pressures and damages caused by loggers which are not always sanctioned by the forestry administration as reported by Kouassi et al. (2021b).
More than 50% of the farmers indicated that they would agree to plant trees or adopt cocoa agroforestry using tree species such as legume, food, fruit, and shade trees. These findings are consistent with previous studies in West Africa as the diversity of shade tree diversity increases soil fertility and reduces the damage caused by cocoa plantation pests (Bisseleua et al. 2013; Wartenberg et al. 2017). Additionally, cocoa agroforestry systems can help mitigate CSSVD severity (Andres et al. 2018). The adoption of cocoa agroforestry systems with a minimum of 18 trees per hectare from 3–5 species can benefit from the certification schemes resulting in premium payment to the farmers (Lescuyer and Bassanaga 2021). Similarly, the REDD + and biodiversity programs in the region have also contributed to the adoption of cocoa agroforestry (Sanial 2019).
Despite the above benefits, about 45% of the farmers indicated that they do not intend to keep and plant more trees in their cocoa orchards due to the tree tenure. Indeed, most farmers are not yet informed that the new forest code has transferred the tree ownership from the state to the farmers. In addition, the new code has also put an emphasis on the essential role of cocoa agroforestry in the restoration of the degraded forest and cocoa landscapes. This observation suggests that more active communication on this new forest code is essential to improve the adoption of cocoa agroforestry in the region.
Similarly, gender, the presence and the preference of tree species on cocoa farmland are one of the factors influencing agroforestry adoption, and female groups are less likely to participate than their male counterparts in Nigeria (Adetoye et al. 2018)d te d’Ivoire (Kouassi et al. 2021b). Also, length of residency significantly affects agroforestry adoption. Farmers with a longer residency may own larger properties with better environmental features (Pokorny et al. 2021) as a result of the attendance of several training and sensitization programs on tree adoption, which may enhance their willingness to adopt cocoa agroforestry.