泰森多边形与传统四株树法的天然红树林空间结构比较研究

Comparative study on natural mangrove forest stand spatial structure by Tyson polygon and traditional four trees method

【目的】以漳江口红树林为研究对象,利用泰森多边形和四株树法划分天然红树林林木平面空间分布结构,确定对象木的最邻近竞争木,计算林分空间结构指数,对比分析两种方法的异同,探讨泰森多边形法应用红树林林分空间结构量化的适用性。【方法】设置20 m×20 m样地,每木调查时用钢尺测量横纵坐标,在坐标纸上标记调查木的位置和编号,同时在调查表上记录树种、树高、胸径等林分因子。利用ArcGIS软件,将外业调查的坐标纸上的样木位置图矢量化形成样木点图层,将林分因子调查表导入样木属性表,并由调查木的点位置生成泰森多边形。用VBA编程来计算林分空间结构指数:混交度、大小比数、角尺度。【结果】在福建漳江口红树林国家自然保护区选择样地进行试验,对比分析基于泰森多边形法与传统的四株树法的计算结果,结果表明:1)基于泰森多边形法确定的林分空间结构单元由1株对象木和3~12株竞争木构成,平均值为6株。2)两种方法计算的混交度的相关系数为0.828,大小比数的相关系数为0.881,角尺度的相关系数为0.225。这两种方法计算的混交度和大小比数具有较高的相关性和一致性,而角尺度,两者差异较大。3)基于泰森多边形法与四株树法的不同树种的混交度和大小比数平均值非常接近,混交度最大差0.05,大小比数最大差0.02;角尺度差异较大,最大差值达0.11;泰森多边形法计算的混交度比四株树法的略大一些,而大小比数和角尺度则略小一些。【结论】无论从单株还是从整个调查林分上看,基于泰森多边形法计算的混交度和大小比数与传统四株树法的计算结果差异小,两种方法有较高的相关性和一致性,但其中泰森多边形法与四株树法相比,混交度略大,大小比数略小;而对于角尺度,两种方法的计算结果有较大的差异,并且泰森多边形法的计算结果值小于四株树法。这种差异的根源在于泰森多边形法确定的竞争木数量波动。泰森多边形法通过最邻近原则将平面空间进行划分,不重不漏,对于量化表达林木在空间上的竞争关系具有合理性和有效性,同时外业调查时不需要判断对象木的最邻近竞争木以及距离量算和角度测量,大幅减轻外业调查工作量,适用于天然红树林林分空间结构研究。

【Objective】Taking natural mangrove forest in Zhangjiang river Estuary as research object,the possibility was evaluated to calculate the indexes of mangrove forest stand spatial structure by Tyson polygon and traditional four trees method which divided spatial distribution structure of natural mangrove trees.Comparative analysis the results of the similarities and differences between the two methods and determine the target’s nearest neighbor competitive trees.【Method】The sample plot was 20 m×20 m.In the field survey,the horizontal and vertical coordinates of each tree were measured with a steel ruler,the locations and sequence number of the investigated trees were marked on the coordinate paper.At the same time,forest stand factors such as species,height and DBH were recorded on the questionnaire.The investigated tree locations marked on the coordinate paper were digitized to form points layer of sample trees by ArcGIS software,the forest stand factor questionnaires were imported into the attribute table of sample trees,and the Tyson polygons were generated from the points layer of survey trees.The spatial structure indexes of forest stand such as mingling,neighborhood comparison and uniform angle index were calculated by VBA program.【Result】With testing in the plot in Fujian Zhangjiangkou mangrove nation natural reserve,the paper gives a contrastive analysis of the outcomes of calculating the spatial structure indexes based on Tyson polygon method and the outcomes of traditional four trees algorithms.The result,showed:1)The forest stand spatial structure unit determined by Tyson polygon method is made up of 1 target tree and 3-12 competitive trees,with an average of 6.2)The correlation coefficient of mingling calculated based on the two methods is 0.828,0.881 for neighborhood comparison,and 0.225 for uniform angle index.The outcomes of mingling and neighborhood comparison have significantly correlation and agreement between the two methods,but the outcomes exist remarkably difference when calculating uniform angle index.3)The calculation of forest stand spatial structure index of different species are very close based on Tyson polygon method and traditional four trees algorithms.The maximum difference of mingling is 0.05,neighborhood comparison’s is 0.02,while it’s an obvious maximum difference at uniform angle index with 0.11.The results of mingling calculated by Tyson polygon method is slightly larger than those calculated by four trees algorithms,but neighborhood comparison and uniform angle index are slightly smaller.【Conclusion】No matter from the individual plant or the whole investigation sample trees,there is a little difference and high correlation and consistency in the calculation of mingling and neighborhood comparison based on Tyson polygon method or traditional four trees algorithms,but quite an obvious difference at uniform angle index.Compared with the four tree method,there is slightly larger at mingling and slightly smaller at neighborhood comparison based on Tyson polygon method.The difference comes from the fluctuation of competitive wood quantity determined by Tyson polygon method.Tyson polygon method divides planar space by principle of nearest neighbor,so it is reasonable and effective for quantitatively describing spatial competition relationship of trees.At the same time,it does not need to judge the nearest competing tree of the target tree,distance measurement and angle measurement in field investigation,which greatly reduces the workloads of field investigation,and is suitable for forest stand spatial structure research of natural mangrove forest.