祁连山青海云杉林动态监测样地群落特征

依据中国森林生态系统动态监测样地设置方法，参照国际森林生态学大样地建设技术规范，于2010年和2011年在祁连山自然保护区青海云杉（Picea crassifolia kom.）林内建立了10.2 hm2动态监测样地，定位调查了样地内19 927株DBH≥1 cm活立木木本植物，完成了第一次群落特征调查和分析。调查结果表明：①青海云杉群落成层现象明显，可划分为乔木层、灌木层、草本层和苔藓层4个层次。乔木层是群落的最主要层，整体而言其垂直高度结构复杂性要大于灌木层和草本层，苔藓层较为发达。②青海云杉径级结构呈明显的倒“J”形，个体集中在径级1~5 cm至21~25 cm，其占到总个体数目的91.42%，青海云杉更新良好。③青海云杉DBH≥1 cm对应的树高结构呈“单峰”形，高度主要集中在小于6 m，占到总个体数目的60.00%，高度偏小，小树较多，中树占有一定的比例，大树较少。同时，树高和胸径二者之间显著符合二次函数关系（p＜0.05），反映了青海云杉群落生物学特征。④从空间分布格局来看，青海云杉表现出明显的聚集性分布格局。对不同年龄段青海云杉小树、中树和大树点格局分析表明，随着龄级的增大，种群的聚集程度减小，即由聚集分布变为随机分布，表现出明显的扩散趋势。青海云杉个体分布没有明显的空间异质性，而且2 534株大树随机分布在该样地，表明该群落未受大范围的人为干扰。

Community Structure of a Dynamical Plot of Picea crassifolia Forest in Qilian Mountains,China

According to Chinese forest ecosystem dynamic monitoring plot setting method and international forest ecology large plot construction technical standards, a 10.2 hm2 Picea crassifolia forest dynamic monitoring plot was established in the Qilian Mountains Nature Reserve in 2010 and 2011. The 19 927 woody plants with DBH (diameter at breast height) greater than 1 cm at the plot were surveyed based on positioning survey method, and the community characteristic was analyzed. (1) The layer structure difference is obvious for the P. crassifolia community, and it can be divided into tree layer, shrub layer, herb layer and bryophyta layer. The tree layer is the main layer of the community, and its vertical structure is more complex than the shrub layer and herb layer, and the bryophyta layer is also well developed. (2) The diameter structure of the P. crassifolia community is obviously an "inverted-J" form, the DBH of the individuals is concentrated in 1-5 cm and 21-25 cm, which account for 91.42% of the total individuals, so regeneration of the P. crassifolia population is good. (3) The height structure of the P. crassifolia individuals with DBH greater than 1 cm is single-peak form, and the height of the main individuals is less than 6 m, which account for more than 60.00% of the total individuals. So the height of the P. crassifolia individuals is relatively low, and young trees are much, mid trees hold a certain proportion, and big trees are little. There is a significantly quadratic function relation between the tree height and the DBH of the P. crassifolia (P<0.05). (4) The spatial distribution of the P. crassifolia individuals displays obviously a clustering distribution pattern. The point pattern analysis of small trees, mid trees and big trees of the P. crassifolia community shows that the clump intensity of the P. crassifolia population reduces with the increasing of the population age, and the congregation distribution turns into random distribution, which displays obviously a diffusion trend. There is no obvious spatial heterogeneity for spatial distribution of the P. crassifolia individuals, and 2 534 big trees of the P. crassifolia community are randomly distributed in the large plot, showing no significant human interference.