<![CDATA[Abstract. Nonthakhot P, Podong C. 2025. Carbon and biodiversity responses to fire and timber harvesting in dry dipterocarp community forest in Thailand. Biodiversitas 26: 5081-5088. Understanding how burning and harvesting systems in tropical dry forests influence both carbon storage and biodiversity is crucial for sustainable management. This study presents the first landscape-scale assessment of above- and below-ground carbon pools and biodiversity in dry dipterocarp forest managed by community in Thailand under four management regimes: community-based prescribed fire (Ching-Pao), selective harvesting (Ching-Kep), unmanaged natural forest, and wildfire-affected areas. Forty-eight stratified plots (i.e., 12 plots per management regime, each plot measuring 20 × 20 m) were established to measure aboveground biomass, litter biomass, Soil Organic Carbon (SOC), and tree species composition. Prescribed fire plots exhibited the highest aboveground carbon (36.74 ± 2.13 t C/ha) and Shannon diversity index (H? = 3.21), significantly exceeding wildfire plots (24.11 ± 1.84 t C/ha; H? = 2.48). SOC was also greatest in prescribed fire plots (38.21 ± 2.87 t C/ha), followed by selective harvesting, unmanaged forest, and wildfire treatments. Beta diversity analysis (Bray–Curtis) revealed distinct community assemblages among treatments, with prescribed fire fostering fire-tolerant species such as Shorea obtusa and Dipterocarpus tuberculatus. These differences indicate that low-intensity, community-managed burns can enhance carbon sequestration and sustain higher functional and taxonomic diversity compared to unmanaged or severely burned areas. The integration of functional traits—such as thick bark, resprouting ability, and high wood density—appears to underpin the resilience and productivity of fire-managed forests. Overall, the findings provide empirical evidence supporting “fire-smart” community-based forest management as a strategy to align biodiversity conservation with climate mitigation goals in tropical dry forest landscapes.]]>
