OUR MONTHLY (MAY) RESEARCHER (Dr. FILIMON A. MGANDU)
What is the topic of your research? Why is it important to study the topic?
This study aimed to investigate the transmission dynamics and control of aflatoxin contamination in crops and its associated health risks in livestock, and humans. Aflatoxin contamination poses a significant challenge to food safety and security, as it affects both the health of consumers and the entire supply chain. Doses of aflatoxins beyond acceptable levels are dangerous and may lead to poisoning, also called aflatoxicosis, a life-threatening illness. The study identified the most sensitive parameters in aflatoxin dynamics and provides the best strategy for intervention. The study has also developed a prediction model, a tool for farmers and decision-makers to forecast the likelihood of aflatoxin contamination in the growing season so that they can apply appropriate control measures. In addition, the results from this study contribute to the attainment of the first three Sustainable Development Goals (SDGs) by 2030. No poverty, zero hunger, and good health and well-being are directly related to the production of food and feeds. Lastly, the study contributes to the scientific community’s understanding and acts as the basis for further research on using mathematical modeling techniques in fighting against aflatoxin contamination.
What are the key findings or observations of your research?
The following are the key findings from the research: Firstly, the most sensitive parameters in aflatoxin contamination dynamics are the crop contamination rate, aflatoxin fungi shading rate, and aflatoxin fungi death rate. Secondly, decreasing crop contamination and shading rates as well as increasing the death rate of aflatoxin fungi in soil by 50%, reduces the aflatoxin contamination by above 92%. Thirdly, incorporating good farming practices, biological control, and public education and awareness campaigns is a cost-effective strategy for controlling the contamination. Lastly, the prediction model predicts the likelihood of maize and groundnuts grown in a given area having aflatoxin levels above the established standard based on weather data with an accuracy of 82% in groundnuts and 68% in maize.
How can the results of your research be utilized in practice?
Policy-makers, researchers, crop processors, transporters, farmers, and the general public can benefit from this study in various ways. First, the study recommends the adoption of a control strategy involving good farming practices, biological control, and public education and awareness campaigns. In this strategy, good farming practices and public education and awareness campaigns should be fully implemented in ten and eight years respectively while biological control can be implemented at 55% at the beginning and gradually decreased to zero for the remaining years. Second, it is also important to control relative humidity in the transport and storage facilities of crops to prevent aflatoxin contamination. Relative humidity should be kept below 80% to limit the production of aflatoxin. Transport and storage facilities should have appropriate air conditioning for ventilation and humidity sensors giving real-time data for informed decisions. Third, as an alternative to reducing water stress, measures such as irrigation should be used. This is because water stress caused by inadequate rainfall has also had a great influence on crop contamination.
What are the key research methods and materials used in your research?
The study used ordinary differential equations (ODE) to study the dynamics of aflatoxin contamination in crops and its associated health impacts on livestock and humans. For sensitivity analysis, the study employed global sensitivity analysis using Latin Hypercube Sampling (LHS) and Partial Rank Correlation Coefficients (PRCC). The study then adopted Pontryagin’s Maximum Principle (PMP) to determine the best (cost-effective) strategy for controlling aflatoxin contamination. For the prediction model, the study used machine learning algorithms: Gaussian Process Classification (GPC), Support Vector Machine (SVM), Random Forest Classifier (RFC), and K Nearest Neighbor (KNN). Data for the prediction model were obtained from five districts of Tanzania: Bahi, Kilosa, Kongwa, Meru, and Masasi. The data were aflatoxin content in maize and groundnuts grown in the 2017/2018 season.
Links to the research outputs: Paper 1: https://doi.org/10.1016/j.sciaf.2023.e01980
Paper 2: https://doi.org/10.1016/j.rico.2023.100313
Prediction model: https://filimon2499.pythonanywhere.com/
Contact: Filimon Abel Mgandu
Email: filimonabel2499@gmail.com or filimon.mgandu@cbe.ac.tz
Mobile: +255 759 784 638