Development on of an Unmanned Aerial Vehicle-Based Orangutan Population Assessment and Monitoring Method for the Multifunctional Landscape of East Kalimantan, Indonesia

Muhammad Sugihono Hanggito, University of Texas at El Paso


Deforestation, habitat degradation, and other forms of land conversion are threatening the existence of orangutans (Pongo spp.), the critically endangered great apes that only live on the two large Sunda-shelf islands of Sumatra and Borneo. Currently, orangutan populations persist not only within conservation or protected areas but also in other functional landscapes such as forest/acacia plantations, oil palm plantations, and mining concessions. The presence of orangutan populations in this recently modified multifunctional landscape has the potential to exacerbate human-orangutan conflict, which could further threaten orangutan populations. Lack of information about the distribution and size of orangutan populations hampers long term conservation efforts on local to regional scales. Habitat-specific orangutan population data are crucial for effective conservation planning as such information can be used to more adequately assess population-level threats, set conservation priorities, and establish and/or maintain monitoring. Traditionally, orangutan distribution and density are assessed by conducting ground-based nest surveys, which are expensive, time-consuming, require an experienced survey team, and generally have a limited sampling area compared to the home of orangutan. This study focused on evaluating the utility of Unmanned Aerial Vehicle (UAV)-based image analysis for detecting orangutan nests in a range of multifunctional landscapes throughout East Kalimantan. Specifically, the study compared nest data derived from UAV and ground-based surveys conducted in the multifunctional landscape of East Kalimantan, Indonesia, assess what factors limit nest detectability in UAV imagery, and developed models to correct UAV-based methods to ground-based surveys. From this research, UAV flight protocols for orangutan nest detection were developed for the multifunctional landscapes inhabited by orangutans in East Kalimantan. Summary total of 15, 250 to 600-meter-long coupled ground/UAV transect surveys were conducted at different localities in three multifunctional landscape units (6,800 m surveys in total). We detected a total of 205 nests from the ground surveys and 45.37% of these nests were detected in UAV images: 82.50% in timber plantations, 45.83% in the post-mining rehabilitation areas, and 32.48% in secondary forests. UAV-based surveys failed to detect nests that were not detected in ground-based surveys, highlighting the high accuracy of ground-based surveys. Canopy openness and nest site location were key determinants of nest detectability in UAV imagery. We tested three different interactions for predictive models, which showed that models predicting ground-based nest counts from UAV imagery were strongest when a two-way interaction with average transect UAV-derived crown spread was accounted for. Although fewer nests were detected in UAV imagery compared to ground-based surveys, UAV surveys required significantly less time for a smaller field team to execute. Given that UAV-derived attributes of forest structure could be used in a single model to effectively approximate ground-based survey results, this study concludes that UAV-based survey methods are an effective complement to ground-based survey methods that could enhance orangutan population surveys of the multifunctional landscape of East Kalimantan, and therefore, the protection and conservation management of orangutan.

Subject Area

Environmental science|Behavioral Sciences|Ecology

Recommended Citation

Hanggito, Muhammad Sugihono, "Development on of an Unmanned Aerial Vehicle-Based Orangutan Population Assessment and Monitoring Method for the Multifunctional Landscape of East Kalimantan, Indonesia" (2020). ETD Collection for University of Texas, El Paso. AAI28262518.