By Tombayu Amadeo Hidayat
As one of the countries with the largest forest cover in the world, Indonesia is facing a severe problem of deforestation. The enormous land use change in the country has serious impact in the global greenhouse emission, making REDD+ a significant initiative for Indonesia. To ensure effective REDD+ intervention measures, identifying and analysing drivers of deforestation in the country are of a great importance. In this study, we conduct spatial analysis of drivers of deforestation to assess the link between the direct and indirect drivers of deforestation. Random forest algorithm was employed to identify the major indirect drivers of deforestation in the country. Utilizing a number of direct and potential indirect driver data in 139 sample units, we found that the majority of the deforestation in the country is related to palm oil and is greatly influenced by their distance to palm oil mills and roads. Smallholder agriculture-driven deforestations tend to occur near roads and rivers. While biophysical properties of the area can influence the deforestation pattern to a certain extent, it is deemed as insignificant determinant of deforestation, alongside the socioeconomic variables. We conclude that different direct driver has specific underlying driver linked to it. Its effect can be studied by firstly distinguishing the proximate cause of the deforestation rather than analysing the deforestation as a whole. This implies the need of comprehensive direct driver data as a prequisite. This study demonstrated a way to link the direct and indirect drivers, and can possibly be extended to greater scale and detail to produce detailed information regarding drivers of deforestation. This knowledge can further contribute to countries in setting up effective and accurate REDD+ strategies.
Keywords: deforestation; direct driver; indirect driver; palm oil; REDD+; random forest; Indonesia