Using Machine Learning to Model Future Distributions of Babandotan Ageratum conyzoides L. Under Climate Change Scenarios (CMIP 5: RCP 2.6 and RCP 8.5) until 2070 in Bandung Areas

Ageratum conyzoides L., locally known as Babandotan, is an important plant in particular in West Java, including in Bandung, due to its medicinal uses. Currently, climate change is known to influence the distribution of organisms by altering climates and making habitats suitable or not suitable. Then, this present study is aiming to use machine learning to model future distributions of A. conyzoides under climate change scenarios CMIP 5 RCP 8.5 until 2070 in Bandung areas. The A. conyzoides occurrences were sampled from nine locations in Bandung and its surrounding areas. Machine learning using the R platform and MaxEnt algorithm was used to develop species distribution modeling (SDM). The model was then simulated using RCP 2.6 and 8.5 scenarios for the years 2050 and 2070. The quality of the model was assessed using AUC values. The current SDM model shows suitable habitats for A. conyzoides are sizing 1250 km2, mostly located in Bandung (56%), Kota Bandung (24%), and Sumedang (16%). The AUC value was 0.964, showing that the resulting model is good. Climate change will affect A. conyzoides in the future. Based on the RCP model, suitable habitats for A. conyzoides will be shifted northward, eliminating the suitable habitats in the south of Bandung, as can be seen in 2070.