Maize Lethal Necrotic Disease (MLND) is a new disease in East Africa, first reported in Kenya in 2011 and then spread to Tanzania, Uganda and Rwanda. The disease is caused by Maize Chlorotic Mottle Virus (MCMV) in combination with viruses of genus Potyvirus, mostly Sugarcane Chlorotic Mottle Virus (SCMV). The co-infection is the one that results in intensive to complete yield loss. Diagnosis of MLND based on symptoms is reported ineffective because symptoms like stunting and chlorosis resembles nutrient deficiencies or maize mosaic. Detection and characterization of MLND causing viruses have been done by techniques such as enzyme-linked immuno-sorbent assay (ELISA), polymerase chain reaction (PCR) and next generation sequencing. Relatively little work has been done to characterize MLND causing viruses in Tanzania prior to those techniques. The disease can be managed through the use of certified seeds, sanitation, quarantine, crop rotation, the use of resistant/tolerant maize varieties and other cultural practices. The use of resistant maize varieties is considered the most reliable, eco-friendly, effective and economical way of managing MLND. Key words: Enzyme-linked immuno-sorbent assay (ELISA), etiology, Maize Chlorotic Mottle Virus, Maize lethal necrotic disease, nucleic acid based methods, resistant maize varieties, Sugarcane Chlorotic Mottle Virus.

Sustainable control of plant diseases requires a good understanding of the epidemiological aspects such as the biology of the causal pathogens. In the current study, we used RT-PCR and Next Generation Sequencing (NGS) to contribute to the characterization of maize lethal necrotic (MLN) viruses and to identify other possible viruses that could represent a future threat in maize production in Tanzania. RT-PCR screening for Maize Chlorotic Mottle Virus (MCMV) detected the virus in the majority (97%) of the samples (n = 223). Analysis of a subset (n = 48) of the samples using NGS-Illumina Miseq detected MCMV and Sugarcane Mosaic Virus (SCMV) at a co-infection of 62%. The analysis further detected Maize streak virus with an 8% incidence in samples where MCMV and SCMV were also detected. In addition, signatures of Maize dwarf mosaic virus, Sorghum mosaic virus, Maize yellow dwarf virus-RMV and Barley yellow dwarf virus were detected with low coverage. Phylogenetic analysis of the viral coat protein showed that isolates of MCMV and SCMV were similar to those previously reported in East Africa and Hebei, China. Besides characterization, we used farmers’ interviews and direct field observations to give insights into MLN status in different agro-ecological zones (AEZs) in Kilimanjaro, Mayara, and Arusha. Through the survey, we showed that the prevalence of MLN differed across regions (P = 0.0012) and villages (P < 0.0001) but not across AEZs (P > 0.05). The study shows changing MLN dynamics in Tanzania and emphasizes the need for regional scientists to utilize farmers’ awareness in managing the disease.

Banana and plantain ( Musa spp. L.) are fundamental to food security and rural livelihoods across sub‐Saharan Africa, yet production is severely constrained by multiple diseases, with Banana bunchy top virus (BBTV) representing the most devastating viral threat. Inadequate diagnostic infrastructure limits effective management, particularly for asymptomatic infections disseminated through the informal exchange of planting material. This study presents an integrated diagnostic framework combining loop‐mediated isothermal amplification (LAMP) molecular diagnostics with deep learning‐based computer vision for rapid, scalable disease detection under field conditions. A four‐primer LAMP assay targeting the BBTV coat protein gene was developed using conserved sequences from diverse African isolates and validated with a simplified alkaline extraction protocol that eliminates conventional nucleic acid purification. The assay achieved 100% specificity and concordant detection with polymerase chain reaction (PCR) and real‐time PCR, while reducing total diagnostic time from 4–6 h to 60 min. In‐house production of the recombinant protein Geobacillus stearothermophilus deoxyribonucleic acid polymerase, large fragment demonstrated comparable enzymatic performance to commercial alternatives, with projected per‐reaction cost reductions of 70%–80%. Concurrently, a Single Shot MultiBox Detector Lite MobileNetV2 object detection model was developed through 17 iterative training cycles on 17,703 field‐collected images spanning 22 disease and physiological stress classes. The final model achieved per class accuracies of 92.5% for BBTV, 91.0% for banana xanthomonas wilt (BXW), and 98.1% for healthy leaf classification, with deployment via the PlantVillage mobile application enabling real‐time offline diagnostics. A Quick Response (QR) code‐based metadata system links each artificial intelligence (AI) phenotypic assessment to its corresponding molecular confirmation result, enabling georeferenced surveillance that tracks both symptomatic and asymptomatic infections. Together, these complementary tools broad‐scale AI screening for rapid field survey and LAMP molecular confirmation for pre‐symptomatic detection providing an accessible, cost‐effective diagnostic framework for safeguarding banana production across sub‐Saharan Africa.