A generalized allometric model for predicting aboveground biomass across various bamboo species

This study aimed to develop a generalized allometric model to predict the aboveground biomass (AGB) of various bamboo species. Four crucial bamboo species, namely, makino bamboo (Phyllostachys makinoi), moso bamboo (Phyllostachys pubescens), thorny bamboo (Bambusa stenostachya), and ma bamboo (Dendrocalamus latiflorus), were used. This study collected data from previous studies with sample sizes of 90, Diameter at breast height (DBH), culm height (H), and age (A) were used as independent variables to develop three models for predicting AGB, where Model I used only DBH, Model II used DBH and H, and Model III used DBH, H, and A as independent variables. The results showed that Models II and III were superior to Model I based on the root mean square error (RMSE). Moreover, Model III only slightly improved the AGB prediction compared with Model II. This study also evaluated the predictive ability of the models for bamboo species based on the bias and RMSE values. The bias value helps to assess whether the AGB of a single bamboo species is under- or overestimated when that value appears positive or negative in the models. The RMSE values were higher for makino and ma bamboo in Model III when comparing the RMSE values for each bamboo species between Models II and III. However, it was higher in moso and thorny bamboo in Model II. Therefore, it was difficult to infer which model was better for all species because the ability to predict AGB was inconsistent among species for these two models.