| Site productivity remains a fundamental concern in forestry and is s a major driver of forest resource availability. Study of site productivity provides a theoretical basis for scientific forest management, for sustainable development of forests, for restoration and utilization of forest resources, and for economic, ecological and social benefits of forests. Currently, many foresters have paid much more attention to the estimation of site productivity mainly through site index in pure even-aged stands. To our knowledge, there are only a few researches on the estimation of site productivity in natural mixed uneven-aged stands. Therefore, the objectives of this study are(1) to explore a determinate method of stand top height in mixed Quercus broadleaved natural forest(2) to model the relationship between stand top height and mean height and(3) to model site productivity base model through the relationship between stand top height and dominant diameter at breast height, the relationship between stand mean height and mean age, and the relationship between stand mean height and mean diameter at breast height. It can provides a theoretical basis for mixed Quercus broadleaved natural forest through these objectives. The specific content of this study is as follows:(1) Eight methods for determining dominant height, including arithmetic mean dominant height for regardless of tree species, for one tree species, for two tree species, for three tree species, and for four tree species(hereafter referred to as H0, H1, H2, H3 and H4, respectively), and basal area weighted mean dominant height for two tree species, for three tree species, and for four tree species(hereafter referred to as H2 g, H3 g and H4 g, respectively), were compared through paired t test using one time permanent sample plot data i.e. 12 one hectare mixed Quercus mongolica broadleaved natural stands(hereafter referred to as one time data) located on the Tazigou state-owned forest farm of Wangqing Forest Bureau in Jilin. The results showed that the differences among H0, H1, H2, H3 and H4 were lager than those among H2 g, H3 g and H4 g. H0 has the highest value, while H1 has the lowest value of all eight methods. To facilitate the theoretical analysis, H0 was selected as the stand top height for its highest value.(2) Simultaneous autoregressive(SAR) models, including spatial lag model(SLM), spatial Durbin model(SDM) and spatial error model(SEM), within nine spatial weight matrices, including Rook, Queen, 1/d, 1/d2, 1/d5, Exp, Gaus, Spher and LSM, were utilized to model the stand top and stand mean height relationship in one time data, using ordinary least squares(OLS) as a benchmark model. The results showed that there was a high linear relationship between stand top and stand mean height and that there was a positive spatial autocorrelation pattern in model residuals of OLS for Moran’s I of OLS residuals have a significant difference with expectation I0(p<0.01). Moreover, SEM and SDM performed better than OLS in terms of reducing the spatial dependence of model residuals and model fitting, regardless of which spatial weight matrix was used. SEM was better than SDM. SLM scarcely reduced the spatial autocorrelation of model residuals. Among nine spatial matrices in SEM, Rook performed best in model fitting, followed by 1/d2 and LSM. Consider the distribution of repeated sample plots in mixed Quercus broadleaved natural forest(hereafter referred to as repeated data), 1/d2 was the optimum spatial weight matrix. Therefore, SEM and 1/d2 were selected to model stand top height and mean height relationship.(3) The stand top height of repeated data was estimated through the(2) model of stand top height and mean height relationship. Three algebraic difference approaches were constructed based on the relationship between stand top height and dominant diameter at breast height, between stand mean height and mean age, and between stand mean height and mean diameter at breast height, respectively. At each algebraic difference approach, Richards model was selected as a base algebraic difference approach from 21 alternative nonlinear modes using five criteria, including adjusted coefficient of determination(Ra2), root mean square error(RMSE), mean absolute error(MAE), relative mean absolute error(RMAE) and Akaike information criterion(AIC). The four different data structures, including nonoverlapping and nondescending(typeC), nonoverlapping(typeD), all possible nondescending(typeE) and all possible(typeF), were applied and comparative analyzed in base algebraic difference approach. The results showed that typeF was the best data structure in base algebraic difference approach of the relationship between stand top height and dominant diameter at breast height, while both in base algebraic difference approach of the relationship between stand mean height and mean age and between stand mean height and mean diameter at breast height, typeD was the optimum data structure. Five different diversity index of diameter and four different diversity index of tree species were applied and comparative analyzed in base algebraic difference approach of the relationship between stand top height and dominant diameter at breast height using typeF as a data structure. The results showed that the algebraic difference approach which incorporating the diameter diversity index of Gini coefficient(GinI) and tree species diversity index of McIntosh evenness index(MceI) has the best mode performance. Therefore, the algebraic difference approach into which considerating GinI and MceI was regarded as a site productivity base model based on the relationship between stand top height and dominant diameter at breast height. Five different stand density index, five different diversity index of diameter and four different diversity index of tree species were applied and comparative analyzed in two base algebraic difference approaches of the relationships between stand mean height and mean age, and between stand mean height and mean diameter at breast height, using typeD as a data structure. The results showed that stand density index and tree species diversity index both had no significance in the algebraic difference approach based on the relationship between stand mean height and mean age, while, diameter diversity index had significance in it. Moreover, GinI was the best diameter diversity index. Therefore, the algebraic difference approach into which considerating GinI was regarded as a site productivity base model based on the relationship between stand mean height and mean age. However, stand density index, diameter diversity index and tree species diversity index had significance in the algebraic difference approach based on the relationship between stand mean height and mean diameter at breast height. The algebraic difference approach which incorporating the stand density of Reineke(1933) stand density index(SDIr), the diameter diversity index of Shannon evenness index(ShaI) and tree species diversity index of Berger-Parker index(BerI) was the best mode performance. Therefore, the algebraic difference approach into which considerating SDIr, ShaI and BerI was regarded as a site productivity base model based on the relationship between stand mean height and mean diameter at breast height. |
PDF Full Text Request