Developing the Additive Systems of Stand Basal Area Model for Broad-Leaved Mixed Forests

Abstract

Stand basal area (SBA) is an important variable in the prediction of forest growth and harvest yield. However, achieving the additivity of SBA models for multiple tree species in the complex structure of broad-leaved mixed forests is an urgent scientific issue in the study of accurately predicting the SBA of mixed forests. This study used data from 58 sample plots (30 m × 30 m) for Populus davidiana × Betula platyphylla broad-leaved mixed forests to construct the SBA basic model based on nonlinear least squares regression (NLS). Adjustment in proportion (AP) and nonlinear seemingly unrelated regression (NSUR) were used to construct a multi-species additive basal area prediction model. The results identified the Richards model (M6) and Korf model (M1) as optimal for predicting the SBA of P. davidiana and B. platyphylla, respectively. The SBA models incorporate site quality, stand density index, and age at 1.3 m above ground level, which improves the prediction accuracy of basal area. Compared to AP, NSUR is an effective method for addressing the additivity of basal area in multi-species mixed forests. The results of this study can provide a scientific basis for optimizing stand structure and accurately predicting SBA in multi-species mixed forests.