Likelihood of elephant death risk applying kernel density estimation model along the railway track within biodiversity hotspot of Bhutan–Bengal Himalayan Foothill

Abstract

The seasonal and occasional movement of wild animals including elephants from one terrestrial habitat to another habitat are very common natural as well as biological habits for the survival of their lives. Around 163 km railway track runs from Siliguri Junction (SGUJ) to Alipurduar Junction (APDJ) within biodiversity hotspot region. According to the report of Forest Department, around 14 prominent and 59 minor elephant corridors have been crisscrossed over the railway track and in between the years 1974 and 2019, nearly 71 elephants are brutally killed due to collision with running trains. The kernel density estimation model is applied to evaluate the spatio-temporal death risk of elephants. To predict the spatial vulnerability or risk of accidental death of elephants, Kernel density estimation analytical algorithms like Bandwidth selection, Intensity, Grid Resolution, Kernel function have been considered. This model considers past accidental deaths data along with secondary data like train speed, fragmentation of elephant habitats, distance of croplands from railway track, human encroachment with electric fencing etc. The anthropogenic activities as well as land use and land cover change have been analysed using Landsat 8 Operational Land Imager—OLI (2019) satellite images and digitally processed by EDRAS IMAGINE 16.5(v16.5.0.852) software. Similarly digital thematic maps have also been prepared on ArcGIS 10.3 platform. Kernel Density Estimation (KDE) intensity (risk) is computed considering the Elephant Death Risk Score (EDRS) as it takes into account of the first and second order properties. Fixed bandwidth is selected for the whole distribution and it is computed by the nearest neighbour distance mean method. On the basis of calculated $$\tau$$ value, (the mean distance between centroids and nearest neighbour index) Kernel Density map is prepared to show the elephant death risk in term of first order properties and second order properties. The result shows very high and high KD (risk) values (0.9799–10.0709) are confined in the eastern and western segments of the SGUJ to APDJ railway line. Therefore, this model helps to take sustainable holistic environmental management of wild animals as well as wild natural habitats in different corners of the globe.