The FAO provided the code for R (R Core Team, 2023) to be used in RStudio (Posit Team, 2023) for the analysis of the TAPE results, which can be accessed via GitHub (FAO Agroecology Team, 2023).

A focus was placed on the Step 2 category of “Health and Nutrition” with the indicators Minimum Dietary Diversity for Women (MDD-W) and the Food Insecurity Experience Scale (FIES) score, which were then visualized and further analyzed in MS Excel (Microsoft Corporation, 2024). Detailed information on the methodology of the different indices is assessable via FAO’s TAPE application guidelines and additional sources (FAO, 2019; FAO and FAO, 2021; FAO and FHI 360, 2016; International Food Policy Research Institute, 2012; FAO, 2025).

In this study, the Spearman’s rank correlation coefficient (ρ) was employed to assess the strength and direction of associations between the agroecological scores derived from TAPE. Spearman’s correlation is a non-parametric method that does not assume normality of the data and is appropriate for ordinal or non-normally distributed continuous variables and small datasets. The Spearman’s rank correlation coefficient was calculated in MS Excel using the following function (DATAtab Team, 2024):

While calculating the Spearman’s rank correlation coefficient, it became evident that the dataset encompassed a considerable number of tied ranks. To mitigate the impact of tied ranks, a corrective method was employed, namely the averaging of tied ranks (DATAtab Team, 2024; Forthofer et al., 2007; ScienceDirect Topics, 2024).

The Spearman’s rank correlation coefficient ranges from −1 to 1. A value of 1 indicates a strong positive correlation, 0 indicates no correlation, and −1 indicates a negative correlation. The results were interpreted in accordance with the methodology proposed by DATAtab Team (2024) and Kuckartz (2013), assuming a significance level of p < 0.05. For visualization, a color code was applied, ranging from light to dark colors, to symbolize the strength of the correlation (see Table 3).

Subsequently, the resulting Spearman’s rank correlation coefficients were subjected to a linear regression analysis. A scatter plot was constructed, and the visual observation, the coefficient of determination (R²), and the function were included in the conclusions regarding the statistical relevance of the regression between the two variables in question.

The mean household size in the target area is approximately six to seven individuals (Table 4). The majority of men in the sampled households (94%) are engaged in agricultural work. The proportion of women involved in agricultural activities is slightly lower, at 76%. It is uncommon for young people (15–35 years) to be engaged in agricultural activities, with a percentage ranging from 10% for male youth to 22% for female.

The average number of different crop species cultivated per household is 6.3, with subsistence food crops such as plantain, cassava and cocoyam being grown together in one field and cash crops such as cocoa being cultivated separately in another field. A total of 19 different crops were reported by participants, with cassava, cocoa, maize, plantain, yam and cocoyam identified as the most important crops. The main livestock species raised in the target area are goats and chickens, with only one farmer rearing grass cutters and two having a pig farm. In this region, livestock are typically kept near the household and allowed to graze freely.

The characterization of agroecological transition (CAET) scores the analyzed farming households on a scale from 0 to 100 for the 10 elements of agroecology (FAO, 2019). The mean CAET score of the targeted farming households is 48.82%, indicating that the system is currently non-agroecological but has considerable potential for transitioning toward greater agroecological sustainability. Figures 2A,B illustrate that none of the assessed households are classified as advanced agroecological systems, with values greater than CAET 70%. Nevertheless, 43% of the households are currently undergoing a transition to agroecological practices, while the remaining majority of 57% are not engaged in adopting agroecological techniques.

A review of Figure 2B reveals a clear distinction between the CAET results of farmers associated with A Rocha activities and those not participating in A Rocha training programs: All farming households undergoing a transition to agroecology belong to A Rocha-trained farmers.

The CAET scores indicate varying levels of agroecological transition across the different elements. The highest score was observed in the element of culture and food traditions (64%) (Figure 3), which is indicative for strong local food traditions and the utilization of local varieties; however, indicators such as appropriate diet and nutritional awareness were found to be lacking. A comparatively high score for the element of circular economy (60%) in the TAPE assessment reflects two distinct trends among the surveyed farmers. On the one hand, farmers who primarily cultivate cocoa as a cash crop and who seem to strongly rely on external markets. Their income therewith is closely tied to government-regulated cocoa prices and global market trends, and they depend on selling their produce via COCOBOD (governmental body responsible for cocoa farming) intermediaries. This often results in limited local value creation and increased vulnerability to external price fluctuations.

On the other hand, some farmers cultivate few or no cash crops, often making them less dependent of global markets. Instead, such farmers often rely on local markets for exchange of products. Despite these differences, both groups grow food for subsistence and supplement their income by selling surplus produce at local markets. Most inputs, such as seeds and fertilizers, are sourced locally or exchanged between farmers, except for the origin of the cocoa seeds. The TAPE assessment values both facts and show potential for agroecological transition. However, income from food crop sales remains minimal, due to low prices and little value addition. The study further found a deficiency in producer networks and support for developing marketing capacity. These divergent patterns of market engagement and self-sufficiency contribute to an overall moderate score for circular economy.

Early signs of an agroecological transition were evident in the areas of co-creation and knowledge sharing (55%) and responsible governance (54%). Organizations, that provide platforms for knowledge sharing, and that are often linked to the biosphere reserve, such as the CREMA, A Rocha Ghana, MOFA, and COCOBOD meetings, are in place; however, these are not being utilized to the full extent for the sharing of agroecological knowledge. While the participation rate in these forums is satisfactory, the majority of interviewed farmers seem to lack knowledge of agroecological practices. Those engaged with A Rocha demonstrated slightly higher knowledge levels, presumably due to training in climate-adapted practices such as composting and biological pest control. The element responsible governance indicates a moderate empowerment of the producer and acknowledges the CREMA structures as a means of participation in the decision-making processes of farmers, although the mechanism functions incompletely. Three individuals interviewed were part of the CREMA, and two of them exhibited notable higher scores, as their perception of empowerment in decision-making is high.

The elements of efficiency and diversity both received scores that were slightly below 50%. Efficiency indicates that a range of inputs, including energy, fuel, fertilizers, seeds, young animals and the workforce, are produced on the farm. With regard to soil fertility and pest management, the target groups exhibit a distinction between farmers who regularly utilize synthetic fertilizers and chemical pesticides, and those who employ these methods on a limited number of crops and integrate biological substances into their management practices. With respect to the element diversity, crops cultivated demonstrate a notable degree of diversity, although the introduction and diversity of animals and trees is relatively limited. Additionally, the income is derived from a constricted range of activities, being mainly the selling of fish products and crops.

The human and social values received a score of 42%, which reflects challenges such as a lack of interest among young people in pursuing farming as a career option and poor working conditions. Despite the integration of women in decision-making processes, limitations persist. Women typically allocate a greater proportion of their time to paid and unpaid work. This is often due to their traditional role in household maintenance and food preparation. Furthermore, some women have expressed concerns regarding unequal decision-making powers within their households, particularly with regard to the decision on household items and the decision on the use of generated income.

The synergies and recycling (both 41%) scores were also found to be low. The integration of livestock and crop production is limited on the interviewed farms, and the recycling of manure and biomass is uncommon. While many farms exhibit favorable soil cover, the integration of trees and the overall connectivity between the agroecosystem and the landscape, is notably low. The low score for recycling is attributed to the little reuse of manure and biomass, as well as the lack of water saving techniques. However, the positive aspects of recycling include the exchange of seeds and the minimal use of non-renewable energy.

The lowest score was recorded for resilience (31%). This encompasses the ecosystem’s ability to recover from climatic shocks, the resilience of households to financial and social challenges, and the resilience of agricultural land. The area’s vulnerability to droughts and flooding, along with a lack of social support mechanisms such as credit or insurance, contributed to the low resilience score.

This part of our study focused on the health and nutrition dimension, including the FIES (Food Insecurity Experience Scale) score and the MDD-W (Minimum Dietary Diversity for Women). The TAPE assessment indirectly addresses several sub-indicators of SDG 2, notably 2.1.1, “Prevalence of undernourishment.” TAPE provides data on the minimum dietary diversity for women, which is a key factor in addressing malnutrition (Inter-Agency and Expert Group on SDG Indicators (IAEG-SDGs), 2017). The sub-indicator 2.1.2, defined as the “prevalence of moderate or severe food insecurity in the population based on the Food Insecurity Experience Scale (FIES),” is informed directly by the TAPE assessment (Inter-Agency and Expert Group on SDG Indicators (IAEG-SDGs), 2017).

The mean FIES score for the entire target group is 55.6%, which indicates moderate levels of food insecurity. As the value approaches 100%, the level of food security increases. Figure 6 illustrates that, based on our interviews, Adwafo households have the highest FIES score, followed by Atafram, which are both notably higher than the Esaase scores. There is also a rather obvious differentiation between A Rocha and non-A Rocha farmers (65.1 versus 49.0), with the A Rocha farmers tending to have higher food security scores.

The mean score for the MDD-W for the interviewed households around Lake Bosomtwe is 4.2 (42%), indicating that, on average, out of the 10 available food groups that would constitute a balanced diet, the target group consumes only four on a daily basis. The predominant food groups consumed are white roots and plantains (plantain, cassava, and yam), fish (occasionally chicken), other vegetables (primarily tomatoes and onions), and sporadically dark green leafy vegetables such as cocoyam leaves. Legumes, nuts, dairy products, eggs, and notably other vegetables and fruits are absent. Consequently, the MDD-W indicates an unsustainable and imbalanced diet. Only two individuals (out of 14) obtained scores that met the criteria for acceptability. The highest score of 60% was achieved by a woman of high social status in Atafram, who presumably has greater financial resources and opportunities to enhance her diet.

A review of the statistical correlation between CAET scores and the food dimension (FIES; Table 6) indicates generally weak to moderate correlation: The overall correlation between the FIES score and CAET is low and not statistically relevant (ρ = 0.36, p = 0.31), which may be attributable to the relatively low CAET scores and the small sample size. Consequently, the scatter plots exhibit high variability, with no clear patterns emerging, and the findings can hardly be generalized beyond this sample.

Among the CAET elements, responsible governance showed the highest correlation with FIES (ρ = 0.58, p = 0.08), suggesting at least a trend in which households more engaged in decision-making and community organizations tend to have higher FIES scores. This may reflect that socially empowered individuals often have greater financial means to supplement their diets. Additionally, many farmers practicing agroecological methods seem to be more active in community meetings and hold respected positions, such as CREMA representatives, which could support dietary awareness. Other CAET elements, did not demonstrate statistically significant correlations with the FIES score.

Overall, these results indicate that none of the CAET elements show strong statistical relevance for food insecurity in this sample. As well, due to the small sample size, the findings should be interpreted with caution and cannot be generalized to the wider population.

The TAPE workshops (Step 3) enabled a participatory exchange of perspectives on agroecological elements, fostering dialog both with and among the farmers. A prioritization activity (see Figure 7) revealed resilience as a central concern across all three villages. Farmers consistently highlighted the lack of access to credit, loans, and insurance as a critical barrier to improving their livelihoods. The absence of financial safety nets and limited community-based support in the face of climatic or economic shocks were identified as major vulnerabilities. The restricted access to credit was frequently mentioned as a key obstacle to diversifying income sources, particularly with regard to expanding small-scale animal husbandry (e.g., pig farming) or investing in processing techniques that would enable better storage and value addition to agricultural products.

The participants expressed a strong interest in the element of circular economy, with a focus on the local processing and value addition of their crops. Overall, the participants demonstrated a considerable interest in training programs concerning biological agricultural practices and livelihood diversification, including pig farming.

The selection of participants for the TAPE methodology was based on purposive sampling, aiming at finding typical rural households in the region and building upon farmers’ willingness to participate, which may have introduced a certain bias as respondents were often engaged and interested already. The relatively limited sample size serves to further reduce the representativeness of the results. In further studies it would be advisable to randomly select farmers and expand the sample size for each village of the LBBR.

In specific cases, a notable discrepancy was observed between the Spearman’s rank correlation coefficient and the linear regression analysis. Such discrepancies can be attributed to the presence of non-linear relationships between the variables. In the event that the relationship between the variables is monotonic but not linear, the Spearman’s rank correlation coefficient may demonstrate a high degree of correlation that is not reflected in the R² value of the regression model (Lund Research Ltd, 2018). Additionally, it is important to consider the strong influence of outliers on the R². The Spearman’s rank correlation coefficient can effectively handle small sample sizes and is less sensitive to outliers, which may explain the observed discrepancy (Leon, 1998; Rahman and Al-Amri, 2011). To mitigate any potential misinterpretation of the Spearman, all results underwent further testing using linear regression analysis and were interpreted through the analysis of scatter plots, which visualized outliers.

The study shows that TAPE could be useful for NGOs, researchers, and the management of the Biosphere Reserve for assessing the impact of activities and monitoring the status of the food system. As well as assessing the status of agricultural fields, the tool generates robust data that can be compared globally, e.g., within the world network of biosphere reserves, in areas such as women’s empowerment and dietary diversity monitoring. TAPE has also demonstrated its usefulness in comparing different groups of farmers, as evidenced by the clear differences revealed by spiderweb diagrams between groups involved in A Rocha activities. The potential of TAPE for intergroup comparisons and tracking changes over time should be further assessed.

Furthermore, TAPE fosters collaboration and knowledge sharing among farmers and stakeholders, as demonstrated by their active participation in evaluation workshops. This enhances awareness and understanding of agroecological principles.

Further study is needed to determine whether TAPE has the potential to empower farmers by enabling them to identify challenges and needs, make informed decisions, and improve their practices with the support of scientifically grounded evidence when used as a self-assessment tool. Self-assessment processes such as these should be tested, supported, and guided by NGOs or other relevant organizations.

In line with earlier findings of A Rocha Ghana (2021) the pilot study indicates that the majority of the interviewed households employ mixed intercropping systems, combining at least two crops, with food crops predominantly intercropped compared to monoculture cash crops like cocoa. The farms demonstrate an overall satisfactory soil coverage with crop remnants. While formal engagement in agroforestry practices was limited, participants expressed interest in exploring tree integration, highlighting potential opportunities for agroforestry adoption in the LBBR (Obiri et al., 2013; A Rocha Ghana, 2021).

Structures for capacity building and agroecological knowledge sharing are already in place and are further supported through the key actors of the LBBR (e.g., CREMA meetings, A Rocha, MOFA, COCOBOD), yet these structures are not fully operational. Consistent with prior studies from Obiri et al. and A Rocha Ghana (Obiri et al., 2013; A Rocha Ghana, 2021) our results indicate a scarcity in information services on sustainable agriculture and demonstrate the need for awareness on climate-adapted agriculture practices such as composting, integrated livestock management, and the use of neem-based biopesticides. Institutional support for access to information and inputs should be strengthened, as also emphasized by the Measuring Agroecology and its Performance (MAP) project in Benin, in order to support income-generating activities (Chabi et al., 2025). Farmer-to-farmer exchanges and Farmer Field Schools (FFS) have been widely recommended and implemented globally to support climate-adapted agriculture and environmental education, with positive outcomes and seem to be a viable solution in the LBBR as well (Chocholata et al., 2015; Gonzálvez et al., 2015; Kiptot and Franzel, 2015; Tavares et al., 2018; Van Mourik, 2012). Studies further show the potential of farmer-to-farmer extension to promote gender balance (Tchuwa et al., 2015).

The CAET results from our pilot study indicate limited financial and social resilience among farming households within the LBBR, particularly in relation to access to credit, insurance, and organized community support. While these findings cannot be generalized, they provide important exploratory insights into constraints that may hinder the agroecological transition in the area.

Formal agricultural insurance remains largely inaccessible to smallholders, consistent with findings across Ghana where limited institutional coverage and high premiums restrict participation (Hazell et al., 2021). Weather-index insurance has shown promise in managing climatic risks and stabilizing livelihoods in parts of Sub-Saharan Africa (Tadesse et al., 2015; Waithaka, 2024), but its applicability in the context of the biosphere reserve Lake Bosomtwe would require adaptation to local financial capacities. Group-based insurance or insurance-for-work schemes (Tadesse et al., 2015) may offer more feasible alternatives.

Access to credit was also reported as insufficient, limiting the capacity of farmers to invest in equipment, training, and sustainable practices. Although causality cannot be inferred, qualitative evidence suggests that improved financial inclusion could support the adoption of agroecological practices. Microcredit systems, particularly those organized at the community level, have proven effective when paired with training, supervision, and repayment terms aligned with agricultural cycles (Siyoum et al., 2012; Omega, 2010). Group-based repayment models (Thomsen et al., n.d.) and Village Savings and Loans Associations (VSLAs) promoted by the Forest and Farm Facility (FFF) (Macqueen, 2022) could provide context-appropriate models for the farmers in the biosphere reserve. The hypothesis that increased access to credit and loans would result in a more significant agroecological transition could not be substantiated by the statistical analysis. It is advisable that an assessment be conducted to determine whether an augmentation in access to funds would indeed precipitate the desired enhancements in agroecology.

Low resilience scores further reflect exposure to climate variability, as documented by A Rocha Ghana (2021). Farmers identified erratic rainfall, high temperatures, and pest outbreaks as major challenges, conditions that align with patterns observed in Tanzania and Lesotho (Antwi, 2024; Chabi et al., 2025). Enhancing adaptive capacity will therefore require training in climate-resilient farming, soil conservation, and water management, alongside support for livelihood diversification (Yamba et al., 2017). Finally, diversification of income sources emerged as a key strategy for adaptation. Consistent with earlier research (Yamba et al., 2017), several respondents of our study indicated an interest in non-farm activities to stabilize household income under changing climatic conditions. Supporting such livelihood diversification through targeted training and market access, facilitated by institutions such as MOFA and CREMA, could significantly strengthen resilience of their farms.

Although rather indicative, our results highlight the potential of small-scale, community-based financial mechanisms and capacity-building initiatives to strengthen resilience and support the agroecological transition in the LBBR.

In our study, the element of synergies also received low CAET scores, indicating for the farms in the LBBR the need for better integration of crop and livestock systems, recycling of organic matter, and incorporation of natural structures such as trees and ponds. While these findings cannot be generalized, they suggest a need to enhance awareness of ecological interactions within farming systems in the LBBR. The limited incorporation of trees in agricultural landscapes and the lack of knowledge about their environmental and economic benefits point to a gap in agroforestry understanding.

Inadequate waste management, particularly in proximity to lakeshores, and over-reliance on agrochemicals for cocoa cultivation, impede progress toward sustainable farming (Amankwaa et al., 2021; Amankwaa et al., 2020). The reliance of the farms participating in our study on subsidized inputs and government spraying programs limits their transition to biological farming systems. Although no causal inferences can be made, promoting organic cocoa production and peer exchanges with experienced farmers may support a gradual shift toward more sustainable practices. A study by Bandanaa et al. shows that organic farming outperforms conventional methods “in terms of land degradation, greenhouse gasses, profitability and gender equity” (Bandanaa et al., 2021).

Furthermore, the utilization of co-creation and knowledge sharing practices appeared limited across existent structures. Although the A Rocha meetings offered valuable training in composting, biopesticides, and intercropping (A Rocha Ghana, n.d.), other platforms, such as those from COCOBOD and MOFA, mainly focus on conventional cocoa management. This discrepancy reflects a broader challenge in agricultural extension services beyond our study, which often prioritize pest control over climate adaptation and water management (Obiri et al., 2013; A Rocha Ghana, 2021). Although the small sample precludes generalization, these observations highlight the potential need for extension services in the biosphere reserve to broaden the scope of their training to include agroecological practices. Also, a literature review conducted by Mockshell et al. revealed that “Agricultural extension, technical assistance, and information sharing were found to be important incentives or drivers for the adoption of agroecological practices (Mockshell et al., 2023).”

The survey results indicate limited dietary diversity and seasonal food insecurity among our sample group, which may pose challenges to the development of a sustainable food system. Most of the women surveyed reported consuming only four out of the 10 food groups daily, a practice that carries risks for malnutrition. This finding may reflect either limited knowledge of dietary diversity or restricted access to the necessary resources. In addition, a considerable proportion of women indicated experiencing occasional hunger and insufficient food at home during specific seasons. Some women of our study also highlighted that food insecurity is particularly prevalent beginning in April and continuing through July and August, a period when harvests are minimal and food preservation practices appear limited.

These findings are in line with a review of studies that suggests that the implementation of agroecological practices results in positive changes to food security and nutrition (Bezner Kerr et al., 2021; Kansanga et al., 2021). It is therefore plausible that improvements in agroecological transition could enhance dietary diversity and food security in the LBBR, although this hypothesis remains to be tested in larger-scale studies.

As a result of our study, it can be said that it is necessary to empower farmers and enhance their capacity to engage in more effective governance of agroecological systems. The CREMA framework, designed to empower local decision-making, appears limited in effectiveness due to violations of regulations, including illegal farming and improper waste disposal near the lake (CREMA, n.d.). We therefore conclude that the enhancement of CREMA’s implementation through the facilitation of community education and the establishment of independent reporting mechanisms has the potential to enhance its effectiveness. Several farmers expressed a desire for greater involvement in local planning, highlighting an opportunity to enhance responsiveness and collective ownership of agroecological initiatives.

The implementation of some of the discussed measures necessitates policy discussions and the adaptation of management plans. A Rocha has highlighted that climate-adapted agriculture has been inadequately incorporated into the medium-term plans for the Bosomtwe and Bosome Freho districts (A Rocha Ghana, 2021). The CREMA management plan emphasizes the enhancement of farming methodologies; however, it is noted that climate-adapted practices are lacking, and there is promotion of waste management practices that are questionable, such as the burial of non-biodegradable materials (MAB National Committee, 2014). Targeted interventions at the Biosphere Reserve level could therefore focus on locally appropriate waste management practices and participatory training for farmers.

The food system in the region combines traditional, circular practices with increasing reliance on global markets due to cocoa cultivation. All households grow subsistence food crops manually, using family labor, with minimal input from regional or national markets. Cocoa production, however, generally creates strong financial dependence on global and national markets and most cocoa farmers rely on subsidized or free inputs from COCOBOD and MOFA, limiting opportunities for local value addition. Strengthening local input supply systems, promoting diversified crop management, and linking cocoa production to locally appropriate agroecological practices could reduce dependency and enhance resilience.

Overall, our findings suggest that locally grounded strategies, including improved governance through CREMA, participatory training, and support for diversified, climate-adapted farming, could strengthen both sustainability and farmer agency within the LBBR.