QQ客服 官方微博 反馈留言

题名	新疆塔里木河中下游天然植物群落及其环境相关性研究
作者	李涛
学位类别	硕士
答辩日期	2003
授予单位	中国科学院．中国科学院新疆生态与地理研究所．植物学．
导师	尹林克，中国科学院新疆生态与地理研究所
关键词	塔里木河
其他题名	Study on The Natural Plant Communities and The effect of Environment in Tarim River, Xinjiang
中文摘要	以塔里木河中下游流域尉犁至农二师35团为研究区，应用相关分析、回归分析、主成分分析和逐步判别分析研究了该地区天然植物群落的物种多样性与环境因子的关系，应用数量分类（TWINSPAN）和排序（CCA）方法，对塔里木河中下游地区天然植物群落类型进行了划分，并探讨了决定该地区天然植物群落类型的主要环境因子和群落类型及其与环境的相关性，应用Fisher精确检验研究种间联结性，应用Pearson相关分析、sPearman秩相关分析和Kendall秩相关分析研究了种间相关性。研究结果表明：1）该地区天然植物群落物种多样性普遍较低，D=0- 0.82，H＝o-1.68，群落植物种类组成较单一，一般为3-8种，有的仅由1-2种植物组成。2）物种多样性指数H’R4与土壤最高含盐量（S2SOLT）呈线性负相关，但相关性程度不高，线性回归方程为：H'R4＝-1.934E-02（S2SOLT）+1.119。多样性指数H'R4与土壤最高酸碱度（SZPH）具有二次曲线关系：H’R4＝190.32-44.25（S2ZPH）＋2.58（S2PH）2.3）对8个环境因子应用主成分方法分析，多样性指数（H·R4）与经旋转后的第一主成分（FACI-2）存在二次曲线关系。4）逐步判别分析总的判断正确率为76．2％，错判的样方中大部分是将属于多样性水平低的划为多样性水平高的，表明在理论上该环境条件可以容纳更多的种类，因为其他因素使实际多样性水平低于理论上的多样性水平。综合相关分析、回归分析和逐步判别分析，表明所涉及到的环境因子仅部分影响物种多样性，物种多样性还与其他的因素有关。5）制约该地区天然植物群落类型的主导环境因子为地下水位、地下水矿化度、地下水酸碱度。6）通过CCA二维排序图将16种植物对干旱、盐碱的适应性划分5类型。21个样地在cCA二维排序图上可聚集成9个植物群落类群，即胡杨（PopuluseuPhratica）群落·铃挡刺(Halimodendron halodendron）群落·库尔勒沙拐枣（Calligonum Kuerlese）群落、多枝怪柳（Tamarix ramosissima）群落、黑果枸祀（Lycium ruthenicum）群落、盐穗木(Halostashys caspica）群落·花花柴（Karelinia casPica）群落、疏叶骆驼刺（Alhagi sparsolia）群落、罗布麻（Apocynum veoetum）群落，CCA连续排序与TWINSPAN分类结果吻合较好。9种植物群落类型中，能耐受最大盐胁迫的为盐穗木群落，能耐受最大干旱胁迫的为铃挡刺群落，能耐受最大地下水碱胁迫的为黑果枸祀群落。7）不同种群之间发生作用的空间范围是不同的；并且随着样方面积的增大，各自有不同的种间联结变化规律，将其概括为4种类型，种间联结随空间变化的曲线表现为坡形、u形、波形和水平。8）不同取样面积对种间联结性分析的有效性是有影响的，仅在某一种样方面积下研究物种的种间联结，得到的结果是片面的，不完全的，认为并不存在对样地内所有物种均有效的最适样方面积，在分析种间联结时应综合参考各种样方面积级别的分析结果。在该地区研究天然植物群落中的种间联结性至少要在中等样方面积以上，结果才能有效。9）该地区主要的乔木树种胡杨与大部分其他物种是呈独立分布的，灌木层的物种间也是呈独立分布，具有种间联结和相关的种对主要是灌木与草本之间和草本层内。Based on the data obtained from natural plant communities in the middle and lower reaches of Tarim River valley, correlation analysis, regression analysis, principal components analysis (PCA) and stepwise discriminant analysis were used to study the relationship between species diversity and environmental variables. Quantitative classification (TWINSPAN) and ordination methods (CCA,DCA and DCCA) were used. The natural plant communities are classified and the major environmental variables that correlated with the natural plant community types are discussed. Fisher's exactly test was used to study interspecific association. Pearson's correlation coefficient, Spearman's rank correlation coefficient and Kendall's rank correlation coefficient were used to study interspecific correlation. From these analysises, the conclusions suggest: The species diversity indices of the natural plant communities in this region are low: D=0~0.82, H'=0~ 1.68 . The plant category is dearth. The species diversity index (H'R4) is negatively correlated with the highest total salt content in the soil (S2SOLT), but the correlation coefficient is low. The liner regression equation is H'R4 = -1.934E-02(S2SOLT) + 1.119. The relationship between the species diversity index (H'R4) and the highest pH in the soil is best fitted in form of quadratic equation:H'R4=190.32-44.25(S2pH)+2.58(S2pH)2. The other 6 environmental variables show there is no direct correlation with the species diversity index. 8 environment variables were studied by principal components analysis (PCA). The relationship between the species diversity index (H'R4) and the first principal component (FAC1-2) is best fitted in form of quadratic equation. 76.2% of original grouped cases can be correctly classified by stepwise discriminant analysis. The cases of lower diversity level are classified into the cases of higher diversity level in the most wrong classified cases. The results of all above analysis show that the environmental variables only partly affect the species diversity and there are some other factors that affect it. The results of CCA ordination reflect sufficiently the relationship between vegetation and ground water depth, ground water pH and ground water degree of mineralization. Five groups are identified from 16 plant species by their compatibility to endure drought and salt-alkali. The gathered groups of 21 plant plots in CCA ordination are consistent with the result of quantitative classification (TWINSPAN). The results of the two analysis methods were in good agreement. Nine types of plant communities are Populus euphratica forest, Halimodendron halodendron shrubland, Calligonum Kuerlese forest, Tamarix ramosissima shrubland, Lycium ruthenicum shrubland, Halostashys caspica shrubland, Karelinia caspica community, Alhagi sparsifolia community, and Apocynum venetum community. Among nine types of plant communities, Halostashys caspica shrubland can endure the highest salt stress, Halimodendron halodendron shrubland can endure the highest drought stress and Lycium ruthenicum shrubland can endure the highest alkali stress of the groundwater. The impact between different populations occurred in different space scope and had different transformation characters with the plot's area enlarging. It can be included 4 curve types: Aslant curve, U curve, N curve, horizontal line.The result of association correlate with the different plot areas. The result should be unilateral and incomplete, if we research association only by investigating one certain plot area. There can't be one fittest plot area that can be used to research all the plant species in one community. The results of interspecific association under a series of plot area should be considered synthetically. The moderate plot areas or more could exactly display the characters of association. Populus euphratica and other plant species, species in the shrub layer show no association and correlation. Shrub and herbs, species in the herb layer show association and correlation.
语种	中文
学科主题	植物生态学
公开日期	2010-11-12
页码	共50页
内容类型	学位论文
源URL	[http://ir.xjlas.org/handle/365004/8154]
专题	新疆生态与地理研究所_中国科学院新疆生态与地理研究所(2010年以前数据)
推荐引用方式 GB/T 7714	李涛　. 新疆塔里木河中下游天然植物群落及其环境相关性研究[D]. 中国科学院．中国科学院新疆生态与地理研究所．植物学．. 2003.

个性服务

查看访问统计

相关权益政策

暂无数据

收藏/分享

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。