<![CDATA[Abstract. Kaewbanlao W, Sunthornhao P, Lumyai P. 2025. Tree diversity and carbon sequestration assessment using multiple methods in a dry evergreen forest restoration site, Nakhon Ratchasima, Thailand. Biodiversitas 26: 3886-3902. This study assessed the forest structure, tree diversity, and carbon sequestration in a Dry Evergreen Forest (DEF) restoration area (Forest Plantation Target 49: FPT 49) in Nakhon Ratchasima Province, Thailand. The site contained 4,538 individual trees across 96 species, with a high tree density (1,512 stems ha-¹), basal area (130.84 m² ha-¹), and diversity (Simpson’s index: 0.90; Shannon-Wiener index: 3.02). Microcos tomentosa, Nephelium hypoleucum, and Dialium cochinchinense were the dominant species, with Fabaceae being the most species-rich family. Biomass and carbon sequestration were estimated using three methods. Wood volume-based estimation method used the Wood specific Density (WD) and Biomass Expansion Factors (BEFs) and resulted in the highest estimates (152.36 Mg ha-¹ of biomass; 71.61 Mg C ha-¹; 262.59 Mg CO? ha-¹), likely due to the inclusion of Biomass Expansion Factors (BEFs) and species-specific wood density. Method I yielded intermediate estimates, while Method II produced the lowest. Statistical analysis revealed that the DBH size class had the strongest effect on carbon sequestration (?²: 0.14), followed by its interaction with the method used (?²: 0.17), with significant differences observed across DBH classes and methods. Bland-Altman and boxplot analyses showed a strong agreement between the allometric method and the direct carbon method, but highlighted variability and potential overestimation by Method III, especially for trees with large DBH. Although carbon content estimates were lower than those found in old-growth forests, the results indicated substantial recovery in tree diversity and carbon stock 28 years post restoration. The study emphasizes the importance of method selection in estimating carbon sequestration and recommends integrating structural and ecological factors to enhance the accuracy of forest carbon accounting.]]>
