<![CDATA[Preservation effort to prevent tree collections loss even on aged trees (> 100 years old) is one of important missions in Bogor Botanical Garden since its establishment in 1817. Abiotic factors such as global warming and biotic factors from pests and diseases can threaten the survival of aged tree collections. Their survival is also influenced by plant health’s deterioration as they age. As the BBG has many functions not only for conservation but also for human ecological activities, fallen tree accidents are becoming primary concern to prevent biodiversity loss and people’s lives. We examined 154 trees health to determine a falling probability of 1106 aged trees based on several factors that caused to fall in the past and to make model prediction generated by nine supervised machine learning algorithms. We also classify susceptibility of tree families prone to fall from the highest accuracy of algorithm prediction. Inverse Distance Weighted interpolation method was used to depict zone map of trees prone to fall. The prediction showed that Random Forest model had the highest accuracy and low false negative (FN) value which were important to minimize error calculation on aged trees was not prone to fall but it turns out to be prone to fall. It predicted 885 trees prone to fall which 358 had high probability to fall. Fabaceae, Lauraceae, Moraceae, Meliaceae, Dipterocarpaceae, Sapindaceae, Rubiaceae, Myrtaceae, Araucariaceae, Malvaceae, and Anacardiaceae were tree families that were highly predicted to fall.]]>
Copyrights © 2023
