Soil parameters associated with Fusarium oxysporum f. sp. cubense incidence and their integration into an IoT-based monitoring system for banana plantations

Fusarium wilt, caused by Fusarium oxysporum f. sp. cubense (Foc) , is a major soil-borne threat to banana production and food security in economies such as Rwanda. Although soil physicochemical conditions are known to influence pathogen persistence, the precise relationships between specific soil parameters and disease severity along with the potential for real-time field monitoring to replace laboratory testing remain insufficiently studied. This study collected soil samples from healthy and Fusarium-contaminated plots at three banana plantation sites in Rwanda namely Huye-Rusatira, Kamonyi-Gacurabwenge, and Ruhango-Byimana), analyzing nitrogen, phosphorus, potassium, pH, and electrical conductivity in the laboratory. Independent-samples t-tests and Pearson correlation analysis compared healthy and contaminated soils and quantified parameter–severity associations, while a paired t-test evaluated agreement between laboratory and IoT-based real-time measurements. Results showed that contaminated soils were consistently more acidic, nitrogen-depleted, and lower in EC than healthy soils ( p < 0.001 at nearly all sites). pH ( r = − 0.922), nitrogen ( r = − 0.777), and EC ( r = − 0.536) exhibited the strongest negative correlations with disease severity, whereas potassium and phosphorus showed weak associations. These findings identify pH, nitrogen, and EC as reliable indicators of Fusarium wilt risk, supporting their integration into IoT-enabled monitoring systems, with future research aimed at developing AIoT-based predictive models combining soil and climatic data for proactive disease forecasting.