Optimal plot size and shape for sampling growing stocks and tree species diversity in tropical forests: Results from a forest inventory in Hazarikhil Wildlife Sanctuary of Bangladesh

Abstract

Tropical Forests play a significant role in providing ecosystem services such as biomass carbon, timber, and tree species diversity. However, variations in sampling layouts such as plot sizes and shapes in forest inventory can cause discrepancies in the estimation of tree biomass carbon, thus hindering planning, sustainable forest management, and conservation of tropical forests. We estimated tree growing stocks (tree density, seedling density, basal area, stand volume, and above-ground biomass carbon) and tree species diversity (Shannon-Wiener Index and Jaccard Index) using different sampling layout in Hazarikhil Wildlife Sanctuary (HWS) of Southeastern Bangladesh. The results obtained from different plot sizes and shapes were compared to those derived from a complete enumeration of 1 ha (10,000 m²) of the same forest land. We also estimated coefficient of variation (CV%) and evaluated optimal plot size and shape in terms of estimation accuracy and time efficiency. The survey was carried out in 450 concentric plots (50 points with three sample plot sizes of 400, 900, and 1134 m² under circular, rectangular, and square plot shapes) where all trees and bamboo with a DBH of ≥ 5 cm were measured. Moreover, three subplots with a size of 4, 9, and 20 m² were placed to count the number of seedlings in three different shapes, thus making a total of 450 subplots in 50 points. Our results revealed that tree density, basal area, stand volume and AGBC were 4652, 38 m², 526 m³, and 98 Mg per hectare, respectively, in the complete enumeration. Tree growing stocks and Shannon-Wiener index estimated in the large (1134 m²) circular plots were close to those in the complete enumeration. The seedling density in small circular subplots was highest. The lower CV% values observed for all these attributes in large circular plots indicated higher accuracy in estimation. We recommend, large (1134 m²) circular plots to estimate growing stocks and tree species diversity while large (19 m²) circular subplots to estimate seedling density in hill forests.