天然混交林林分发育阶段划分方法的比较  

Comparison of methods for dividing developmental stages of natural mixed forest stands

作　　者：徐德鹏 许芳泽 孙海龙 陈美晴 向玮 Xu Depeng;Xu Fangze;Sun Hailong;Chen Meiqing;Xiang Wei(Key Laboratory of Forest Resources and Environmental Management of National Forestry and Grassland Administration,Beijing Forestry University,Beijing 100083,China;Wangqing Forestry Bureau,Yanji 133200,Jilin,China)

机构地区：[1]北京林业大学森林资源和环境管理国家林草局重点实验室,北京100083 [2]吉林省汪清林业局,吉林延吉133200

出　　处：《北京林业大学学报》2024年第7期139-152,共14页Journal of Beijing Forestry University

基　　金：“十四五”国家重点研发计划(2022YFD2200502-04)。

摘　　要：【目的】以长白山云冷杉针叶混交林为例,比较林分不同发育阶段划分的方法,确定最适宜长白山云冷杉针叶混交林发育阶段划分的方法,以便有针对性地制定不同发育阶段的森林经营策略,为天然林发育阶段的科学划分提供方法依据。【方法】以吉林省汪清林业局金沟岭林场天然云冷杉针叶混交林为研究对象,共收集2589个样地样本,采用种间联结−最优分割法(A)、林相特征判别法(B)、TWINSPAN双向指示物种分析法(C)、MRT多元回归树法(D)和基于TWINSPAN的判别分析法(E)5种方法,对云冷杉针叶混交林发育阶段进行划分。通过多重比较检验,验证不同发育阶段间是否存在显著差异,并使用吻合系数来评估不同方法划分结果间的一致性。最后,从树种组成和林分特征(包括样地的平均胸径、平均树高、公顷蓄积量、公顷株数、公顷生物量、公顷断面积、针阔比、物种多样性和径级大小多样性9个指标)方面,对5种方法的划分结果进行比较分析。【结果】多重比较的结果显示,用于描述林分特征的9个指标在不同发育阶段间基本存在显著差异,这表示结果较为可靠。方法D和方法E的吻合系数较高,表示这两种方法的划分结果较为一致;而方法B与其他方法的吻合性均较低。从林分的发育趋势来看,5种方法的划分结果大致相同。随着林分的逐渐发育,林分特征指标普遍增大,多样性指标降低,树种组成趋于简单。主要区别体现在公顷株数和树种组成指标,并且在不同方法的划分结果中,各个阶段之间指标的差异程度和每个阶段的样本量也有所不同。方法E的划分结果更符合林分的生长发育规律。具体来说,第1阶段云冷杉针叶混交林形成期,林分迅速生长;第2阶段林木间竞争加强期,个体显著分化;第3阶段林分发展为近自然林状态,即达到森林经营的目标状态。【结论】基于TWINSPAN的判�[Objective]Taking the Changbai Mountain spruce coniferous mixed forest as an example,this study compared the methods for dividing different stages of forest development and determined the most suitable method for dividing the development stages of Changbai Mountain spruce coniferous mixed forest,aiming to develop targeted forest management strategies for different stages of development and provide a methodological basis for the scientific division of natural forest development stages.[Method]A total of 2589 plot samples were divided by interspecies linkage-optimal segmentation method(A),forest facies characteristic discrimination method(B),TWINSPAN bidirectional indicator species analysis method(C),MRT multiple regression tree method(D),and TWINSPAN-based discriminant analysis method(E)of Jingouling Forest Farm of Wangqing Forestry Bureau of Jilin Province,northeastern China.Multiple comparisons were used to verify whether there were significant differences between different stages of stand development,and the anastomosis coefficient was used to compare the anastomosis between the classification results of different methods,and finally the classification results of five methods were compared from the aspects of tree species composition and stand characteristics.[Result]The results of multiple comparisons showed that the nine numerical indicators used to describe the characteristics of forest stands in this paper basically had significant differences between different stages of stand development,indicating that the results were more reliable.Method D and E had the highest anastomosis coefficient,indicating that the division results were more consistent,while method B had a low agreement with other methods.From the perspective of development trend of the stand,the results of each method are roughly the same.As the forest gradually developed,the forest characteristic indicators increased,the diversity indicators decreased,and the tree species composition tended to be simpler.The differences were mainly reflected in

关 键 词：发育阶段划分 云冷杉针叶混交林 最优分割法 判别分析 TWINSPAN MRT 

分 类 号：S750[农业科学—森林经理学]