论文部分内容阅读
通过4hm2样地调查的数据资料,采用随机分布多度模型和聚集分布多度模型,在对山西灵空山海拔1500～1800m的油松—辽东栎林物种多度及其水平空间分布分析的基础上,运用估计优度评价了2个模型预测分布多度的适宜性。结果表明:①在不同像元的30种木本植物中有20种的多度依次增加且所占的水平空间也依次扩展,有10种不表现为上述关系。②随着像元面积的扩大,遇到多度序列中面积小于上一个种时,多度—面积曲线呈现较大波动;剔除波动节点的物种时,多度—面积曲线的波动趋于平缓。对于同一个物种来说,像元面积越大,其物种所占面积也越大。③估计优度评价结果显示聚集分布多度模型用于预测多度—面积关系优于随机分布多度模型。④无论是随机分布多度模型还是聚集分布多度模型均依赖于m的取值,即物种在固定像元下所占像元数。对于分散程度较高的物种,采用2种模型进行预测时结果较精确,反之预测结果误差越大;在样地总面积一定时,像元面积越小,预测结果越精确。
Based on the data from the 4hm2 sample survey, based on the analysis of the species abundance and spatial distribution of Pinus tabulaeformis and Quercus liaotungensis forest at an altitude of 1500-1800 m on Lingqingshan in Shanxi Province using random distribution degree of abundance model and aggregated distribution degree of abundance model , And evaluated the suitability of the two models in predicting the distribution degree of multipleness by using the goodness of estimate. The results showed that: (1) 20 out of 30 species of woody plants in different pixels increased in turn and their horizontal space expanded in turn, while 10 species did not show the above relationship. (2) With the enlargement of pixel area, the multi-degree-area curve shows greater fluctuation when the area of the multi-degree sequence is smaller than the previous one. When the species of the fluctuating node is excluded, the fluctuation of the multi-degree-area curve tends to be gentle. For the same species, the larger the cell size, the greater the area of its species. (3) The estimation of goodness degree evaluation shows that the aggregation degree of abundance model is better than the stochastic distribution degree degree model for predicting the degree-area relationship. (4) Whether the model of multi-degree distribution or the degree of aggregation distribution depend on the value of m, that is, the number of species in a fixed pixel. For species with higher degree of dispersion, the results of the two models are more accurate, while the error of prediction results is larger. When the total sample area is constant, the smaller the pixel area, the more accurate the forecast result.