%0 Journal Article
%T Trend surface analysis on community structure of a grassland in Bailihuang, Yichang county<br>宜昌百里荒草山草坡群落物种分布的空间趋势分析
%A XIN Xiao-Ping
%A <br>辛晓平
%J 生态学报
%D 2003
%I 
%X Spatial structure influenced the organization of community and ecosystem as a functional variable, other than the background in which biological and environmental factors act on community and ecosystem. This is why present-day ecologists and bio-geologists are interested in detecting the spatial arrangement of population and community. A large set of quantitative ecological methods related with spatial heterogeneity, spatial autocorrelation, spatial scales were developed in recent decades. Spatial trend surface analysis is one of the quantitative ecological methods that study the relation between spatial structure and species abundance distribution in community. In Canonical Correspondence Analysis (CCA), environmental variables can be instead by the spatial coordinates (x,y) of data points. In such case, an ordination of the species data can be obtained that will be constrained to be consistent with the spatial distribution of sampling localities. A high-degree polynomial of the x, y, x2, y2, xy and possible higher powers of basic coordinates can be used to fit to the species data in the manner of trend surface regression. A biplot of species and spatial coordinates of data points should indicate what species have the most important spatial structures. This paper studied the relation between community structure and spatial variability of a grassland in subtropical mid-mountainous region using trend surface analysis based on CCA ordination. In this paper a canonical ordination analysis on the species abundance data constrained by the spatial position of sampling localities was conducted. The first two eigenvalues are 0.116 and 0.056 respectively, they measure the species data that is explained by the first and the second canonical axes and, hence, by the spatial position of sampling localities. The first two canonical axes together account for 55.3% of the variance of 27-species-spatial localities relation, and for 18%of the total variance of species data. The community structure surface is obtained by kriging the sample scores which are weighted averages of species, while the data values for the spatial trend surface maps is based on trend surface regression function. For the canonical axis 1, the first community structure surface is pretty well approximated (correction = 87.6%) by the first spatial trend surface map. In other words, the trend surface regression function predicts the community structure from the simple knowledge of the sampling localities. The same goes for axis 2, the second community structure surface is well fitted (correction = 82.4%) with the second spatial trend surface map. This result suggests a community organization mechanism which is strongly spatial structured. Correlation between spatial trend of community structure and main environmental factors was also studied. Soil effective phosphor is significantly positively related with the first community structure surface and the first spatial trend surface, and significan
%K CCA ordination
%K trend surface analysis
%K community structure<br>CCA排序
%K 趋势面分析
%K 群落结构
%K 空间插值
%K 草场
%U http://www.alljournals.cn/get_abstract_url.aspx?pcid=90BA3D13E7F3BC869AC96FB3DA594E3FE34FBF7B8BC0E591&jid=FE163E5DB2274E5937319DE98913EC37&aid=AB14C131E4703FE1&yid=D43C4A19B2EE3C0A&vid=EA389574707BDED3&iid=5D311CA918CA9A03&sid=DEEC1AC3B6D3EB96&eid=D5FE4D327C08DA1B&journal_id=1000-0933&journal_name=生态学报&referenced_num=0&reference_num=30