Abstract: Trees form a critical part of the ecosystem. The Baobab, Leadwood, Marula, and Shepherd trees are economically and culturally significant to the northern Limpopo region. However, climate change, agriculture, and land developments have harmed their natural habitats. Monitoring tree populations allows conservationists to take corrective measures to preserve the species. However, using field data to do so can be expensive and time-consuming. Remote-sensing data from satellite images provides a cost-effective way to monitor biodiversity in the area. This project uses field data and satellite images to develop models that will assist the Department of Geography, Geoinformatics, and Meteorology in distinguishing between the four species of interest. We evaluated a two-stage Support Vector Machine and a Convolutional Neural Network. While the SVM is more performant, it requires frequent field data collection. On the other hand, our CNN can accurately detect 66% of the trees overall and performs well in identifying the Baobab. We propose using the CNN for future applications as it reduces dependence on field data. However, the model requires further hyper-tuning. We also create a web-based application allowing users to visualise and classify the data.