Predicting the below-ground biomass of trees in young mixed-species stands in Brazilian Amazon
摘要
Below-ground biomass (BGB) is a key carbon pool. However, its determination is laborious and extremely costly. Hence, indirect methods are needed to predict the below-ground biomass of trees and forests. In this study, we evaluated the performance of different approaches when predicting the below-ground biomass of trees in young forest restoration stands in Rondônia, southwestern Brazilian Amazon. A total of 50 trees from 34 tree species were felled and bucked for direct biomass determination (both above and below-ground) at age 10 years. Different approaches based on root-to-shoot ratios proposed in the literature, regression models, and root-to-shoot ratios fitted with our data were compared. In general, literature approaches and the mean root-to-shoot ratio (R = 0.24) presented biased estimates, while the regression equation using tree diameter at 1.30 m above-ground, total height, and wood density presented the best performance. Different approaches are discussed when forest inventory data and destructive sampling data are available, suggesting a framework to be followed in particular cases. This study provides useful information to understand below-ground biomass assessments and reduce their uncertainties and improve the Brazilian National Greenhouse Gas Inventory.