Conference Agenda

Agribusinesses currently operate in a volatile, uncertain, complex, and ambiguous environment, severely impacted by factors such as globalisation, climate change, geopolitical tensions, and systemic shocks. As agribusiness value chains continue to increase in complexity as interconnectedness advances, so too does the potential of risk contagion throughout the value chain. As a result, agribusinesses and by extension the broader food system are subject to cascading, non-linear impacts that pose an existential threat to achieving their goals. Despite the growing concern expressed for the impact of these challenges, a systematic and comprehensive framework for analysing agribusiness’s fragility, risk exposure, and interconnectedness in the South African context is yet to be completed.

The aim of this research is to address this gap by developing a comprehensive framework to characterise, assess, and measure the fragility, risk exposure, interconnectedness, and contagion within South African agribusinesses and their value chains. Grounded in the concepts of antifragility and the action domain in the institutional environment, the conceptual foundation positions agribusinesses as nested, goal-oriented entities whose vulnerabilities are amplified by the dynamic interdependencies of modern value chains.

The study will employ a combination of heuristic stress testing and a dynamic risk assessment to assess the vulnerability of the system. Primary data will be collected by surveys across agribusinesses in various agricultural sectors, industries and regions in South Africa to ensure a representative sample is obtained.

The study will to contribute to literature by 1) Quantifying and mapping the risks faced by South African agribusinesses; 2) Analysing the causes of the fragility of South African agribusinesses while determining the level of interconnectedness and contagion of the risks identified; and 3) Developing recommendations and strategies to improve the resilience and robustness of South African agribusinesses.

Indigenous and neglected crops (INC) have been studied in several contexts for their potential in the transition to more resilient food systems. Existing research highlights the potential of INC in climate change adaptation based on their abiotic stress tolerance characteristics. Also, INC are considered a key element of nutrition-sensitive food systems. Some INC have shown higher levels of fiber, essential amino acids, and micronutrients compared to the most common commercial crops. In the case of South Africa, research highlights the potential of INC to provide food and nutrition security, especially in disadvantaged areas. Nevertheless, existing research focuses on INC potential only at specific levels. Therefore, the objective of this study is to map the existing evidence on INC in South Africa using a food systems perspective. Our aim is to provide an integrative view of what has been researched at production, marketing and utilisation levels with environmental, economic, social and health implications. For this systematic map, we followed the Guidelines and Standards for Evidence synthesis in Environmental Management developed by the Collaboration for Environmental Evidence (CEE). We developed a comprehensive overview of research topics, description of the most studied and under-researched areas, distribution of research across the three South African provinces of the Eastern Cape, Mpumalanga and KwaZulu-Natal. We identify research gaps and analyse their implications for the promotion of INC. Our preliminary results suggest that most research is concentrated at the production level and that there is limited integrative food systems evidence to inform policy.

Hidden hunger remains a persistent and often overlooked challenge in Southern Africa. Deficiencies in iron, zinc and vitamin A coexist with adequate or even excessive caloric intake. This paradox reflects the region’s double burden of malnutrition, marked by rising obesity alongside insufficient micronutrient consumption. Indigenous and underutilised crops such as sorghum, bambara groundnut and amaranth present a promising yet underexploited solution. They are nutrient-dense, resilient to drought and marginal agroecological conditions and are deeply embedded in local food cultures.

However, these nutrient-dense crops remain marginalised within formal markets, research agendas and policy frameworks. This poster asks: To what extent can indigenous crops contribute to reducing hidden hunger while strengthening climate-resilient and sustainable food systems? The objectives are to (1) assess the micronutrient density of selected indigenous crops relative to dominant staples, (2) examine their role in enhancing dietary diversity and nutrition security and (3) analyse value chain and institutional constraints limiting their wider utilisation.

The poster will adopt a conceptual synthesis approach from existing knowledge and draw on regional food composition tables, peer-reviewed nutrition studies and literature on neglected and underutilised species (NUS) in Southern Africa. Comparative nutritional evidence will be used to illustrate how crops such as sorghum, Bambara groundnuts and leafy vegetables like amaranth outperform refined maize and wheat in key micronutrients linked to hidden hunger. A food systems and value chain lens will then be applied to illustrate how weak market development, limited processing infrastructure and negative consumer perceptions restrict their integration into mainstream diets.

By bridging nutrition science with value chain analysis, the poster will help advance the argument that diversifying beyond a narrow staple-based food system is essential for climate-resilient nutrition security. This poster contributes a conceptual framework linking indigenous crop promotion to nutrition-sensitive value chain development.

Food sovereignty is a resilient strategy to global food crises in Indigenous communities. Indigenous crops are integral part of the consumed food of indigenous communities. However, food sovereignty is criticised and needs locally based solutions for an equitable transition. Based on the food sovereignty framework, this paper provides a data-driven understanding of culture-based principles of food sovereignty.

We examine the extent to which Ovazemba's resilient mechanisms promote nutrition and food security amid overlapping global crises. 60 key informants knowledgeable in Indigenous food systems from 8 Ovazemba communities in rural northwest Namibia were interviewed. A semi-structured questionnaire guided face-to-face interviews conducted following snowball sampling. A convergent parallel design enabled the analysis of frequencies of culturally informed practices and food consumption patterns.

Results suggest that, despite external stressors, including climate change, Ovazemba demonstrated a strong capacity to adapt and achieve nutrition and food security. Their adaptive capacity is centred on five culture-based principles of food sovereignty. These principles, rooted in ‘humanism’ and coexistence with nature, include Indigenous knowledge and ancestral wisdom, farming gender roles, manual and organic food production practices, food sharing initiatives, and biodiversity conservation.

The generalisability of the results is limited due to the small sample size. However, the results are reliable, valid, and provide a unique contribution to scientific discourse and policy reform to achieve food security and sovereignty within Indigenous communities. Food sovereignty can be achieved through locally based, culture-led principles, suggesting that Ovazemba’s adaptive capacity can be replicated as a potential solution for global populations facing food insecurity and diet-related deficiencies.

The effects of climate change put agricultural production systems increasingly under stress and amplify concerns about shrinking agrobiodiversity. In this research, we investigated how two indigenous opportunity crops, cowpea and sorghum, could strengthen resilience, and thus, contribute to a more diverse, sustainable and inclusive food system in Zambia

We applied a mixed-method research approach, starting with a participatory food system mapping workshop and concluding with a validation workshop. We conducted 24 individual semi-structured expert interviews with downstream value chain actors and stakeholders and captured the perceptions of smallholder producers in 18 gender differentiated focus group discussions. The interviews were analysed using a reflective thematic analysis framework.

Our results point to two policy spheres at the government’s disposal to unlock the potential of opportunity crops like sorghum and cowpea. Conserving the agro-biodiversity of opportunity crops is currently largely left to informal farmer-managed seed systems. However, these systems need to be effectively supported through participatory breeding programmes and complemented by ex-situ conservation institutions to harness their potential. Also, diversification efforts need to be supported and backed by seed regulations that recognise the smallholder community as an equal partner. Second, government procurement of agricultural produce remains strongly biased towards maize. Including sorghum and cowpea in the portfolio of the Food Reserve Agency will send a strong market signal to smallholder producers by providing a stable and reliable market through their network of rural satellite depots.

Hence, we advocate for a policy shift that firmly integrates opportunity crops into the national food system by devising a holistic framework that sets shared objectives, establishes coordination mechanisms, and provides cross-cutting enablers. Within this framework, sorghum and cowpea specific value chain development strategies need to be developed that are aligned to their distinct production and processing requirements, and target identified end-markets, including school and hospital feeding programmes.

Livelihood diversification strategies (LDS), which include a combination of on-farm strategies involving mixed cropping and the cultivation of indigenous and underutilized crops, as well as off-farm strategies, are often said to yield synergistic benefits for improving food security and establishing resilient rural households. However, the adoption rate of these strategies remains low, especially among rural households in many emerging nations, and the relationship between food security and LDS remains unexplored in many parts of South Africa. Thus, this study focuses on assessing the livelihood diversification strategies employed by rural households, examining the socio-economic determinants of these strategies, and investigating the relationship between these strategies and rural households' food security status in Nkomazi Local Municipality, South Africa. A cross-sectional survey research design was used to collect data from 111 rural households selected through a two-stage random sampling procedure, using a structured questionnaire administered by trained enumerators. Descriptive statistics, binary logistic regression, and multiple linear regression were employed for data analysis. The findings revealed that only one-third (33.3%) of the rural households practiced livelihood diversification. The findings also revealed that the most dominant livelihood strategy among the rural households was off-farm activities (78.4%). Despite a medium dietary diversity score (63.1%) in most households, more than half (54.1%) were food insecure. The binary logistic regression identified farming experience (p<0.05), access to credit (p<0.05), and marital status (p<0.10) as significant socio-economic determinants of diversification. Furthermore, the multiple linear regression showed that livelihood diversification positively influenced rural household food security. The study recommends that promoting livelihood diversification, particularly through improved access to credit and leveraging the farming experience of rural household heads, is essential for reducing reliance on government grants and improving food security. Furthermore, targeted extension-led training programs that promote and support both

on-farm (conventional and underutilized crop cultivation) and off-farm livelihood diversification strategies are recommended for implementation in the area.

Millet is a nutrient-dense, climate-resilient crop vital for food security amid prolonged droughts. Yet research on barriers to millet commercialization and its implications for household welfare remains limited. This study examines how institutional factors, such as market information, extension services, group membership, and access to credit, influence commercialization and its effects on household food security.

Using cross-sectional data from 417 millet-farming households in Singida and Dodoma, Tanzania, we applied fractional logit regression. Results show that credit or market information alone does not significantly affect commercialization, but their combined availability does (p < 0.001). Integrated access encourages greater farmer participation. Group membership also has a positive effect (p < 0.05), likely reflecting gathered social capital, trust, and shared knowledge. By contrast, extension services alone show no significant effect. When combined with credit and market information, however, the interaction is negatively associated with commercialization (p < 0.05). This may reflect the perception that millet is a subsistence crop, limiting extension’s focus on market development.

To assess dietary diversity, we used Inverse Probability Weighted Regression Adjustment (IPWRA). Commercialized households reported a 0.55-point higher Household Dietary Diversity Score (HDDS) and 9.67 points higher Food Consumption Scores (FCS) than non-commercialized ones, after controlling for covariates. Landholding, market access, and off-farm income consistently enhanced food security in both groups.

Overall, millet commercialization is associated with improved household food security. Policy should strengthen institutional support, especially coordinated provision of credit, market information, and targeted extension, recognizing the interdependence of bundled interventions.

South Africa has a substantial body of research on dietary intake and nutritional status;

however, much of this evidence remains fragmented across regions, populations, and

time periods. While the National Department of Health & DSI-NRF Centre of Excellence

in Food Security Desktop Review (1997–2020) has compiled existing studies, the

evidence remains largely static and difficult to interrogate spatially. This study asks: How

can integrated geospatial analysis transform fragmented nutrition evidence into a

dynamic tool that supports research prioritization and targeted food system

interventions?

The primary objective is to develop an interactive geospatial platform that consolidates

findings from the Desktop Review into a structured, visual map of nutritional indicators

across South Africa. The platform enables identification of geographic research gaps and

highlights areas where limited data exists, thereby guiding future research efforts.

To demonstrate applied value, the study includes a focused adolescent obesity case study

using data from the National Dietary Intake Survey 2022 (NDIS2022). For adolescents

aged 10–19 years, anthropometric measures (e.g., height, weight, mid-upper arm

circumference) are integrated with contextual variables such as household food security

and local food environments. Spatial analysis is used to examine how proximity to

different food outlets may influence dietary behaviors and nutritional outcomes.

The expected outcome is a functional interactive map that allows researchers to visualize

existing nutritional evidence, identify under-researched regions, and prioritize further

investigation. For policymakers, the tool provides spatially grounded insights to support

targeted, nutrition-sensitive interventions in vulnerable communities.

By transforming static evidence into an integrated and policy-relevant analytical

platform, this study strengthens food security analysis and contributes to more sustainable

and climate-responsive food system strategies in Southern Africa.

Post-harvest handling and storage losses in the fruit sector can reach as high as 44%. While general environmental monitoring exists, it often fails to capture the precise micro-climatic conditions and mechanical stressors experienced by individual produce items during transit. This research asks: How can an "apple-mimicking" IoT sensor combined with Artificial Intelligence provide actionable recommendations to extend and predict produce shelf-life? The primary objective is to utilize real-time data from a specialized sensor—monitoring temperature, ethylene, and acceleration—to build predictive models that offer proactive management strategies for the cold chain.

The research adopts a Digital Twin methodology powered by an Actor Model framework to ensure high fault tolerance and responsiveness. A physical sensor, designed to mimic the dimensions, weight and thermal behaviour of a real apple, travels within produce pallets to experience the "physical world" environment. The data serves as the input for a processing pipeline where machine learning algorithms analyse environmental fluctuations to predict metabolic deterioration and shelf-life.

By monitoring ethylene concentrations to control ripening and acceleration to identify mechanical injuries, the system aims to move beyond simple data collection toward intelligent recommendation. The expected outcome is a scalable, AI-driven framework that enables agro-businesses to prioritize distribution based on predicted freshness rather than just "first-in, first-out". This integration of IoT and Data Science offers a cost-effective path for small and medium-sized enterprises to minimize food waste, ensure regulatory compliance, and enhance the sustainability of the global food chain.

This research directly addresses food security by transforming the supply chain from a reactive to a predictive system. By using AI to provide real-time recommendations—such as adjusting storage temperatures—the system minimizes the volume of nutrient-dense produce lost to spoilage. Ensuring that high-quality, un-bruised fruit reaches the consumer.

Dust soiling is a persistent operational factor that reduces photovoltaic (PV) output by attenuating incident irradiance and shifting the electrical operating point under load. In agri-photovoltaic (Agri-PV) systems, where on-farm energy supports climate-resilient production and value-chain activities for indigenous and underutilized crops (including irrigation, cooling, and small-scale processing), quantifying soiling effects is important for establishing evidence-based maintenance and cleaning criteria. This study investigates PV performance under controlled laboratory conditions using three dust types: (i) very fine sand, (ii) medium sand, and (iii) coarse sand applied in incremental masses (0–5000 〖g/m〗^2) to a PV module operated with a fixed electrical load. The disparity in load voltages between the clean and soiled states was quantified and used to establish a power ratio metric. Subsequently, a semi-empirical predictive model was formulated by fitting the data for each dust category. To estimate the power output under soiled conditions, the soiling model was integrated with a parametric curve PV I-V and a numerical load line intersection technique, allowing the operating point to be determined. For an untested particle size, the associated model parameters can be derived through interpolation between the calibrated sand fractions, enabling the assessment of user-defined scenarios. The measurement indicates clear dependence, with fine sand producing a stronger power reduction per unit mass than medium and coarse fractions and a nonlinear response with an apparent threshold region at higher deposition. By linking soiling to predictable power losses and conditions-based cleaning decisions, this work supports climate adaptive Agri-PV management aimed at maintaining energy reliability while reducing unnecessary water consumption, labor and downtime. In the future, camera-based dust assessment will be integrated with these measurements to train a deep learning model for automated cleaning triggers.

Maize-based food systems in sub-Saharan Africa are threatened by destructive pests, including Fall Armyworm (Spodoptera frugiperda), African Armyworm (Spodoptera exempta), and Maize Stalk Borer (Busseola fusca), with serious implications for yield stability and food security. This study asks: how can early multi-pest identification be improved while maintaining interpretability for agricultural decision-making? The objective is to develop an explainable hybrid deep learning–fuzzy logic framework that distinguishes these visually similar pests using field-style imagery. Secondary image data sourced from publicly available datasets used in previous studies were compiled and preprocessed using augmentation to reflect real-world variability (e.g., illumination and orientation changes). A VGG16-based feature extractor is combined with fuzzy logic membership functions to enhance class separability and provide interpretable inference alongside model predictions. Expected results are improved discrimination among pest classes compared to conventional deep learning baselines. The study aligns with symposium themes by supporting timely pest management and strengthening maize-based food security and sustainable value chains.