Population Density and Habitat Suitability of Gaur (Bos gaurus H. Smith) in the Phu Fa Non-Hunting Area, Nan Province

Abstract

Background and Objectives: Spatial ecology of wildlife provides a fundamental basis for conservation planning and sustainable natural resource management. This is particularly important for gaur (Bos gaurus H. Smith),           the largest extant wild bovine, which requires extensive home ranges, exhibits low reproductive rates, and is highly sensitive to environmental changes. Currently, gaur populations are facing increasing pressure from habitat loss and fragmentation driven by human activities, as well as limitations in forest connectivity. These factors contribute to population decline and elevate the risk of habitat isolation in the long term. Although protected areas in Thailand have been progressively expanded, information on population density and spatial use patterns of gaur remains limited, particularly in mountainous forest ecosystems in northern Thailand, where terrain complexity and inaccessibility constrain field studies. Phu Fa Non-Hunting Area represents an ecologically significant forest landscape for wildlife, especially gaur, comprising relatively intact montane evergreen and dry evergreen forests. The area also has high potential as an ecological corridor facilitating movement of large mammals between Thailand and the Lao People’s Democratic Republic. However, current knowledge regarding population status and environmental determinants of habitat suitability in this area remains insufficient to support effective spatial management. Therefore, this study aims to estimate the population density of gaur and to analyze environmental factors influencing habitat selection. The findings are expected to provide baseline information for conservation planning, habitat restoration, and long-term population monitoring.

Material and method: Field data collection was conducted from November 2024 to October 2025. A total of 10 camera trap stations were systematically deployed across the study area to record the occurrence of gaur. Camera locations were selected based on areas with a high likelihood of gaur presence, including wildlife trails, water sources, and natural salt licks. In addition, signs of wildlife were recorded during patrol surveys, including direct sightings, vocalizations, footprints, dung, and feeding evidence, along with their geographic coordinates. All collected data were processed for statistical and spatial analyses. Population abundance was assessed using the Relative Abundance Index (RAI), calculated as the detection rate per number of trap nights. Population density was estimated using the Random Encounter Model (REM), which does not require individual identification and is particularly suitable for large ungulates lacking distinctive individual markings. Habitat suitability was evaluated using the Maximum Entropy (MaxEnt) model, which has been widely recognized for its high performance in predicting species distribution from presence-only data. A total of seven environmental variables were included in the analysis, comprising both physical and biological factors: elevation above mean sea level, slope, land use, distance to water sources, distance to roads, distance to villages, and distance to ranger stations. Model performance was assessed using the Area Under the Receiver Operating Characteristic Curve (AUC) to evaluate the model’s ability to discriminate between suitable and unsuitable habitats.

Results and discussion: A total sampling effort of 3,650 trap nights resulted in 27 independent detections of gaur, yielding a Relative Abundance Index (RAI) of 0.74%, indicating a relatively low detection rate. However, population density estimated using the REM was 19.6 individuals/km², suggesting that the study area still has the capacity to support a viable gaur population. The spatial distribution of gaur was uneven, with individuals tending to utilize specific areas associated with key resources, such as water sources, foraging grounds, and refuge sites. Habitat suitability analysis using the MaxEnt model demonstrated excellent model performance (AUC = 0.987). Approximately 35 km², representing 29.41% of the total study area (~119 km²), was classified as suitable habitat for gaur. Highly suitable areas were predominantly distributed within montane evergreen and dry evergreen forests, which provide favorable conditions for gaur survival. Permutation importance analysis indicated that distance to water sources, distance to ranger stations, and distance to villages were the most influential variables determining gaur occurrence. In contrast, percent contribution values highlighted the importance of both physical and biological factors, particularly elevation, land use (dry evergreen and montane evergreen forests), and distance to water sources, in shaping habitat suitability. Highly suitable habitats were primarily concentrated in high-elevation mountainous areas (1,200–1,600 m above mean sea level), characterized by a mosaic of ridge-top grasslands interspersed within montane and dry evergreen forests. These habitat features likely enhance resource availability, provide suitable shelter, and facilitate avoidance of potential threats. Although anthropogenic disturbance factors, such as roads and human settlements, had relatively lower contributions, they still played a role in constraining the spatial distribution of gaur within the study area.

Conclusion: The results indicate that although the population density of gaur in the study area is relatively low, the presence of suitable habitat covering approximately one-third of the total area reflects the potential of the forest landscape to support population persistence and future expansion. These findings provide important implications for management interventions, including the improvement of water sources, enhancement of forage availability, and maintenance of habitat connectivity to reduce the risk of population isolation. Furthermore, the establishment of baseline data on population density and habitat suitability serves as a critical tool for long-term monitoring of population dynamics and for evaluating the effectiveness of conservation measures. With appropriate management, the study area has the potential to function as a core habitat and a source population, contributing to regional population stability. This, in turn, would strengthen forest ecosystem integrity and support the sustainable conservation of wildlife in northern Thailand over the long term.