Abstract ID: 3.5108 | Accepted as Poster | Talk/Oral | TBA | TBA

Betula utilis, a critical treeline species in the Indian Himalayas, plays a vital role in maintaining ecological balance and supports local communities by providing non-timber forest products (NTFPs). This study models the potential distribution and assesses the climate vulnerability of Betula utilis using the Random Forest (RF) model, a robust machine learning method suited for analyzing complex ecological datasets. Occurrence data were collected along trekking paths in Uttarakhand and Himachal Pradesh, while climatic variables from the CHELSA dataset—such as temperature and precipitation—were integrated with topographic factors, including elevation and aspect. Climate projections were based on the IPCC AR6 scenarios: SSP1-2.6 (low emissions), SSP3-7.0 (intermediate emissions), and SSP5-8.5 (high emissions). The RF model demonstrated high predictive accuracy, with an Area Under the Curve (AUC) of 0.98 and an overall accuracy of 94.6%, underscoring its suitability for habitat modeling. Results revealed that elevation and precipitation were the most influential factors shaping the distribution of Betula utilis. Under the high-emission scenario (SSP5-8.5), significant habitat loss and fragmentation were projected, posing a severe threat to Betula utilis and the livelihoods dependent on its ecosystem services. In contrast, the low-emission scenario (SSP1-2.6) suggested minimal habitat loss, emphasizing the potential of sustainable practices to preserve this species and enhance community resilience. These findings underscore the need for targeted climate mitigation and conservation strategies to protect both Betula utilis and the local communities that rely on it.