Costs and precisions of alternative plot configurations for estimating above-ground tree biomass in tropical forests

Abstract We used field measurements and sampling simulations with existing tree-level data to assess the precision of the estimates and data measurement costs for ground-based sampling inventory for above-ground tree biomass (AGB) and basal area estimation. Data were obtained from primary wet tropical forests using field measurements in Costa Rica and, for simulations, published tree-level data on Barro Colorado Island in Panama. Plot configurations and costs were compared using criteria commonly used by forest carbon offset project developers to meet precision of +/−10% at 90% confidence interval for AGB estimation. Detached plots were tested using circular fixed-area, nested, and several types of angle-count configurations. A plot divided into a cluster of sub-plots was further tested with angle-count plots. The basal area factor (BAF) used for angle-count plots varied between 2 and 20. Key findings include (i) angle-count plots emerged as a highly efficient plot configuration for forest inventories in tropical regions with high tree density and size variability with 33%–80% cost savings without precision penalties compared to the nested and fixed-area plots respectively; (ii) angle-count plots with large BAFs (3–8) are cost-efficient and can achieve low sampling errors (⁓10%) with fewer plots and trees than the fixed-area and nested plots; (iii) a cluster of angle-count subplots with a large BAF (e.g. BAF = 12, 16, 20) is an efficient configuration for further investigation; (iv) small fixed-area plots for small trees with angle-count sampling for large trees is also an efficient configuration for further investigation. Among the alternative plot configurations compared, angle-count plots with their variants emerged as a highly cost-efficient and versatile alternative in complex tropical forests with visual occlusion, and are worth studying further.