Conservation of designated source sites is a fundamental strategy for global tiger recovery. Reliable estimates of tiger Panthera tigris habitat use within these source sites are crucial for informing effective management strategies. In this study, we assessed tiger habitat use within Bardia-Banke Complex, one of the 42 global source sites, situated in the western Terai Arc Landscape (TAL) of Nepal. We conducted a grid-based detection and non-detection camera trap survey across 719 grid cells, each measuring 2×2 km². To assess tiger habitat use while accounting for imperfect detectability, we applied a single-season occupancy model. We analyzed nine covariates that have potential to influence tiger habitat use in the Complex, including terrain, co-predators, prey, water availability and disturbance. We found that fine scale (2x2 km²) tiger habitat use in the Complex was 0.43 (SE ± 0.0085, 95% CI: 0.414 - 0.448). Our analysis demonstrated that tigers used habitats unevenly across the Bardia-Banke Complex. Our results showed that terrain ruggedness index, prey index, and proximity to waterholes were key determinants of tiger habitat use. Tiger habitat use was positively associated with prey abundance and negatively associated with terrain ruggedness, and distance to waterholes. We emphasize the importance of influencing habitat covariates that determine the probability of habitat use for taking appropriate habitat-management decisions for tiger conservation in the TAL. We highlight the importance of periodic assessment of tiger habitat use in this globally significant source site to monitor changes in spatial habitat use patterns, serving as a measure of the effectiveness of wildlife management interventions.

To assess fine-scale habitat associations, we used non-invasive camera traps to survey the entire Bardia and Banke National Park along with adjoining forest area following the standard tiger monitoring protocol from 16 December 2021 to 12 March 2022. We used grid-based sampling approach and divided the study area into 719 grid cells of size 2x2 km² for systematic detection and non-detection camera trap survey. Cuddeback (C1), Panthera (V5 and V6) and Reconyx automated cameras with white flash were used to capture photos of tigers and other wild animals. Camera traps were mounted on standing trees or pegged wooden poles 45 – 60 cm above the ground, perpendicular to, and 6 m – 10 m apart on either side of game trails, forest roads, and riverbeds.  Each survey occasion consisted of a 15-16 day camera-trapping session, with cameras monitored twice a week in accordance with Nepal's tiger and prey monitoring protocol (DNPWC, 2017). Tiger detection (photographs) and non-detection (absence of photo captures) were systematically recorded and compiled for analysis.