关键词: 栎类天然林, 最粗优势木, 胸径生长, 混合效应, 立地质量评价模型

Abstract:

Objective: This study was carried out to analyze the effects of site factors on the DBH growth of the thickest dominant tree in the stand, construct a growth model of the dominant tree in Hunan Quercus natural forests with the mixed effects of site types and derive a site quality evaluation model based on the DBH of the thickest dominant tree as an indicator. The site quality evaluation equation was expected to propose a new method for the site quality evaluation of Quercus natural forests. Method: This paper took Hunan Quercus natural forests as the research objects. Based on the measured data of 51 plots, the site factors those having significant impacts on the growth of the dominant woods at breast height were selected by the quantitative method Ⅰ, and the site factors were classified and combined according to the standard. The initial site type, through the R language, the DBH-age of the dominant tree in the natural Quercus forests in Hunan was fitted, the optimal basic model was screened, and the initial site type was added as a random effect to the basic model to construct a mixed effect model. The k-means was used to gather the class clusters the initial site types with similar effects to form a site type group, and it was added as a random effect to the optimal basic model to construct a mixed-effect model containing the site type group. By derivative, the site quality evaluation model was obtained, and the analysis of variance was used to verify the significant relationships between forest stand area and site index. Result: The site factors those having significant effects on the growth of the thickest dominant tree diameter at breast height were altitude, slope gradient, slope position, and slope aspect. The order of significance was altitude>slope gradient>slope aspect>slope position. Four theoretical growth equations were selected for fitting, and the coefficient of determination of the model was about 0.7. The optimal basic model was the Richards model, and the coefficient of determination R² was 0.731 8, the mean absolute error (MAE) was 5.442 6, the root mean square error (RMSE) was 6.879 1, and the expression equation was D=a×[1-exp(-c×AGE)]^b.By combining the selected four significant site factors to form the initial site type and adding it to the basic model to construct a mixed effect model with site types, the determination coefficient (R²) was increased to 0.9016. The k-means clustering was used to cluster the initial site types into 6 site type groups, which were added to the basic model as a random effect, the optimal model expression equation was D_j=a_j×[1-exp(-c×AGE)]^b+ε, the model's determination coefficient (R²) was increased to 0.924 4, which was 26.7% higher than that of the basic model, whereas both AIC and BIC were reduced. By plotting a polymorphic site index curve based on a mixed effect model containing site type groups, the site quality evaluation equation of Hunan Quercus natural forests, based on site classification with the dominant wood diameter at breast height as the evaluation index, was derived as SQEI_M-DBH=a_j×[1-exp(-0.03×AGE₀)]^$\left\{\frac{\ln D_j}{\ln a_j \times[1-\exp (-0.03 \times \mathrm{AGE})]}\right\}$+ε, and through the stand-off area verification, the site index SQEI_M-DBH obtained in this study was significantly related to the stand-off area. Conclusion: Site factors might have a significant impact on the growth of the diameter at breast height of the Quercus in Hunan Province. It could be theoretically feasible to evaluate the site quality of natural Quercus forests and to predict the productivity of the natural forests based on the diameter of the thickest dominant wood as an indicator, which was expected to provide a way for the evaluation of site quality of natural forests.

Key words: Quercus natural forest, the thickest dominant tree, DBH growth, mixed effect, site quality evaluation model