University of Twente Student Theses

Login

Using Machine Learning to Predict the Risk of Human-Elephant Conflict in the Nepal-India Transboundary Region

Khanal, Binita (2022) Using Machine Learning to Predict the Risk of Human-Elephant Conflict in the Nepal-India Transboundary Region.

[img] PDF
2MB
Abstract:Human-elephant conflict (HEC) is a common form of human-wildlife conflict in African and Asian countries where wild elephants are present. The cross-border regions between Nepal and India are the natural habitats for Asian elephants. However, this region has experienced dramatic land cover and land use changes due to human pressure and infrastructure development over the last several decades. Habitat loss and fragmentation drive elephants closer to human settlements and cause more frequent human-elephant conflicts. The study of this phenomenon does not just concern the conservation of wildlife but also human security. This study aims to predict the risk of human-elephant conflict along the transboundary landscape of Nepal-India using machine learning algorithms. To do so, I first modelled the habitat suitability of elephants using an ensemble species distribution modelling approach and identified key factors determining habitat suitability. Then I predicted the risk of human-elephant conflict using a random forest algorithm and identified major factors contributing to the risk. The results of my study show that 26,679 km2, approximately one-third of the total transboundary landscape area, is predicted to be suitable for Asian elephants. Only twenty per cent of the predicted suitable habitat is located within the protected areas. Elevation, precipitation of the driest month and wettest month, and temperature of the warmest month are the key variables determining the habitat suitability for elephants in this region. The result of the predicted human-elephant conflict indicated high human interference in the remaining suitable habitats of Asian elephants. Human settlements and agricultural fields near protected areas experienced a high risk of conflict. The human disturbances and the expansion of settlements in the migratory route of elephants are expected to intensify human-elephant conflict. This is the first study that attempts to use state-of-the-art machine learning algorithms to predict the risk of human-elephant conflict along the cross-border landscape of Nepal-India. The suitable elephant habitat and the human-elephant risk areas identified by this study are important, which could serve as a basis for developing transboundary conservation of elephants as well as strategies for mitigating man-elephant conflicts. The study recommends that the transboundary conservation efforts need to be strengthened, and special attention should be paid to human colonisation around the protected area while implementing measures to mitigate the risks of conflict between humans and elephants. Keywords: Asian elephant, species distribution modelling, ensemble model, habitat suitability, human-wildlife conflict, cross border
Item Type:Essay (Master)
Faculty:ITC: Faculty of Geo-information Science and Earth Observation
Subject:02 science and culture in general, 43 environmental science, 54 computer science
Programme:Geoinformation Science and Earth Observation MSc (75014)
Link to this item:https://purl.utwente.nl/essays/92243
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page