Spatio-Temporal Data Fusion for Early Detection of Phytophthora cinnamomi in Macadamia Orchards using UAV-LiDAR and Multispectral Imagery

Phytophthora cinnamomi (Pc) causes macadamia root rot, resulting in annual yield losses of 25-40% and economic impacts exceeding USD 500 million across the global macadamia industry. A major challenge in disease management is the prolonged latent phase of infection, during which root colonisation progresses without visible canopy symptoms, limiting the effectiveness of conventional field scouting. This study presents a spatio-temporal data fusion framework for the early detection of P. cinnamomi in macadamia orchards using multi-source UAV-derived data and machine learning techniques. An 18-month longitudinal field experiment was conducted across three commercial production sites in Zimbabwe using a DJI Matrice 300 RTK platform equipped with LiDAR and optical imaging sensors. Six vegetation indices and eight LiDAR-derived structural metrics were extracted for 120 individually labelled trees and integrated into 14-dimensional monthly spatio-temporal tensors. Disease status was established through TaqMan qPCR analysis. A convolutional neural network (CNN) trained on 12,300 labelled canopy images achieved 96.16% classification accuracy (weighted F1 = 0.9618). A CNN-LSTM hybrid model achieved an F1-score of 0.96 and an AUC-ROC of 0.99, providing a mean pre-symptomatic detection lead time of 41 ± 6 days across 84 confirmed infection events. This performance exceeded the critical 30-day intervention window commonly associated with phosphonate-based disease management. Economic analysis indicated a benefit-to-cost ratio of approximately 10:1 for orchards experiencing annual infection incidences above 15%. These results demonstrate the potential of spatio-temporal fusion of UAV-derived spectral and structural information for operational early disease detection in macadamia production systems, providing a foundation for improved disease management, yield protection, and enhanced livelihood resilience across Sub-Saharan Africa.