【作者】 任学敏；

【导师】 杨改河； 王得祥；

【作者基本信息】 西北农林科技大学 ， 植物资源学， 2012， 博士

【摘要】 了解植被、物种组成和多样性的分布格局及其与环境变量的关系对于发展和优化物种多样性保护策略、可持续管理和利用植物资源具有重要的指导意义。本研究基于野外调查和实验数据，采用种－面积曲线拟合法分析了太白山8种典型植物群落的最小取样面积；用聚类分析、DCA排序及指示种分析法对太白山南坡和北坡植被分别进行了分类，并对各群落物种组成进行了描述；分别用DCCA和RDA排序法分析了南坡和北坡整体物种组成和单个群落物种组成与环境变量的关系，偏DCCA和偏RDA评估了各个环境变量的相对重要程度，GLM和GAM拟合了物种丰富度对各个环境变量的响应，主要结果如下：（1）太白山8种主要植物群落乔木、灌木及草本层最小取样面积均随研究精度增高而增大，乔木、灌木和草本层分别取125–228m~2、18–170m~2、7–91m~2时，可满足精度80%的研究要求；分别取203–508m~2、62–296m~2、29–150m~2时，可满足精度90%的研究要求。本研究中的所有植物群落的取样面积均达到了精度90%的研究要求。（2）聚类分析和DCA排序把太白山南坡和北坡植被各自区分为8种群落类型，这些群落沿海拔呈带状分布，但在中部海拔（约2250–3350m），不同群落所占据的海拔区间比两端有更多的重叠。不同群落物种组成差异巨大，这种差异在南、北两坡的同一种群落中依然存在。（3）发现了3种尚未被报道过的植物群落，分别是太白深灰槭群落、刺叶高山栎群落和大毛状薹草群落。（4）南坡和北坡的总物种组成分别受13个和11个环境变量的显著影响。南坡，海拔的影响强度最大，其次为年均温，郁闭度、土壤厚度、岩石盖度、土壤有机质含量、全氮含量、碳氮比、全钾含量、坡向、土壤pH值、凋落物厚度和碱解氮含量的影响依次降低；北坡，环境变量影响强度大小次序为年均温、海拔、郁闭度、坡向、全钾含量、有效钾含量、土壤厚度、岩石盖度、土壤pH值、全磷含量和凋落物厚度。南坡和北坡乔木、灌木和草本物种组成与环境变量关系呈现出与各自总物种相似的特征。（5）能够显著影响太白山南坡和北坡总物种丰富度的环境变量及总物种丰富度对各环境变量的响应曲线存在一定不同。南坡，总物种丰富度与海拔、年均温、土壤有机质含量、全氮含量、碱解氮含量和郁闭度呈显著的单峰关系，而与凋落物厚度、土壤碳氮比和有效钾含量呈显著的线性关系；北坡，总物种丰富度与海拔、年均温和土壤厚度呈显著单峰关系，与土壤全氮含量、碱解氮含量和全钾含量呈显著倒单峰关系，与坡向和土壤pH值呈显著线性关系，而与土壤有效磷含量和郁闭度的关系较为复杂，均表现为倒“S”形的变化格局。不同生活型物种丰富度与环境变量的关系与总物种丰富度存在相似特征，即南坡和北坡能够显著解释同一生活型物种丰富度的环境变量不同，其响应曲线亦不同，但总体上可概括为六种关系：线性增加、线性减小、单峰、倒单峰、倒“S”形和无明显关系。（6）分别对5种植物群落物种组成和丰富度与环境变量关系的分析表明，锐齿槲栎群落的物种组成主要受海拔、坡度、年均温和郁闭度影响，其影响强度为年均温>海拔>郁闭度>坡度。物种丰富度主要受海拔、坡向、年均温、土壤厚度、土壤pH值和全钾含量的影响，物种丰富度与海拔存在显著倒单峰关系，随坡向由冷湿向暖干方向线性增加，随年均温和土壤厚度增加线性减小，而随土壤pH值和全钾含量的变化较为复杂。影响红桦群落物种组成的主要是年均温、土壤厚度、碳氮比、全钾含量和郁闭度，其影响强度为土壤碳氮比>土壤厚度>郁闭度>全钾含量>年均温。没有发现显著影响红桦群落物种丰富度的环境变量。糙皮桦群落的物种组成主要受土壤pH值和郁闭度影响，影响强度为郁闭度>土壤pH值。物种丰富度主要受土壤碳氮比和全磷含量影响，物种丰富度与前者存在显著倒单峰关系，但随后者增加显著地线性增加。影响巴山冷杉群落物种组成的最主要环境因子是坡向，其次为岩石盖度，土壤有效磷含量、海拔、土壤有效钾含量、凋落物厚度和郁闭度的影响依次降低。巴山冷杉群落物种丰富度与海拔呈显著倒单峰关系，与凋落物厚度和土壤全氮含量呈显著单峰关系，与土壤有机质含量和碳氮比呈显著线性关系，随二者减小而减小，随郁闭度增加而增加，但郁闭度达到60%后，物种丰富度基本不再变化。头花杜鹃-杯腺柳-大毛状薹草群落物种组成主要受坡向、土壤有机质含量和郁闭度影响，影响强度为郁闭度>坡向>土壤有机质含量。其物种丰富度主要受海拔、年均温、凋落物厚度、土壤有机质含量、全氮含量、土壤碳氮比、碱解氮含量、全磷含量和郁闭度影响。物种丰富度随海拔、土壤全氮含量、碱解氮含量和全磷含量的增加线性降低，随年均温和郁闭度的增加线性增加，而与凋落物厚度、土壤有机质含量和土壤碳氮比存在显著倒单峰关系。太白山植被、物种组成及物种丰富度是多种环境变量共同作用的结果。但分别从太白山南坡和北坡的整体考虑，地形和气候变量影响最大，对于单个植物群落，土壤变量则更为重要，而植被变量（郁闭度）无论对整体还是单个群落都有重要影响。更多还原

【Abstract】 Understanding the distribution patterns of vegetation, species composition and diversityand the relationships between them and environmental variables has important guidingsignificance for developing optimal strategies for conservation of species diversity,sustainable managing and utilizing plant resources. Based on field investigation andlaboratory experimental data, the present study analysed the minimum sampling areas ofseveral typical plant communities in Taibai Mountain using the method of fitting species-areacurves; the vegetation on the south slope and the north slope of Taibai Mountain wereseparately classified using cluster analysis, detrended correspondence analysis (DCA) andindicator species analysis (ISA), and species composition of each plant community was givena quantitative description; the relationships between the overall species composition of eachslope of Taibai Mountain and species composition of each single plant community andenvironmental variables were analysed with detrended canonical correspondence analysis(DCCA) and redundancy analysis (RDA), respectively, and partial detrended canonicalcorrespondence analysis (pDCCA) and partial redundancy analysis (pRDA) were carried outto evaluated the relative importance of each environmental variables on species composition;generalized linear model (GLM) and generalized additive model (GAM) were conducted to fitresponse of species diversity to various environmental variables. The major findings are asfollows:(1) The minimum sampling areas for tree, shrub and herb layers of several types of plantcommunities in Taibai Mountain all increased with the increase of the study accuracy level.When the sampling areas for tree, shrub and herb layers of these communities were separatelyset to be125-228m~2,18-170m~2and7-91m~2, it could meet below80%study accuracy leveland when they were set to be203-508m~2,62-296m~2and29-150m~2, it could meet below90%study accuracy level. All the plant communities related in this paper met below90%study accuracy level.(2) Eight plant communities were clearly identifiable on each slope and these communities presented zonational distributions along elevational gradients, however, theelevational ranges occupied by different communities had more overlaps at middle elevations(c.2250-3350m a.s.l.) than at two extremes. The species composition of differentcommunities existed great difference, and this kind of difference still existed even in the samecommunities occurred on between the south slope and the north slope.(3) Three communities which were not reported in previous literature were found andthey were Quences spinosa community, Acer caesium subsp. giraldii community and Carexcapilliformis var. major community.(4) There were respectively13and11environmental variables significantly explainingall species composition on the south slope and the north slope. On the south slope, elevationhad the strongest explanatory power, followed by annual mean temperature, and theexplanatory powers of crown density, soil thickness, stone cover, soil organic matter content,total nitrogen content, carbon: nitrogen ratio, total potassium content, aspect, soil pH value,litter thickness and alkali-hydrolyzable nitrogen content decreased successively. On the northslope, the order of the explanatory powers of environmental variables on species compositionsuccessively were annual mean temperature, elevation, crown density, aspect, soil totalpotassium content, available potassium content, soil thickness, stone cover, soil pH value,total phosphorus content and litter thickness. The relationships between species compositionof tree, shrub and herb and environmental variables on both slopes all showed the similarcharacteristics with all species composition.(5) The environmental variables that could significantly explain total species richnessand the response curves of species richness to environmental variables between the southslope and the north slope existed some differences. On the south slope, there weresignificantly unimodal relationships between total species richness and elevation, annualmean temperature, soil organic matter content, total nitrogen content, alkali-hydrolyzablenitrogen content and crown density, and there were significant linear relationships betweentotal species richness and litter thickness, soil carbon: nitrogen ratio and available potassiumcontent. On the north slope, total species richness showed significantly unidomal relationshipswith elevation, annual mean temperature and soil thickness, significant reverse hump-shapedrelationships with soil total nitrogen content, alkali-hydrolyzable nitrogen content and totalpotassium content, significant linear relationships with aspect and soil pH value, while hadmore complex relationships with soil available phosphorus content and crown density whichall showed reverse S-shaped relationships. The relationships between species richness ofdifferent life-form and environmental variables presented similar features with total speciesrichness, that is the environmental variables that could significantly explain species richness of the same life-form between on the south slope and the north slope were different, and sowere the response curves. However, generally, they could summarize as four kinds ofrelationships: increase linearly, decrease linearly, unimodal, reverse hump-shaped, reverseS-shaped and no obvious relationship.(6) The relationships between the species distribution and richness of each of five plantcommunities and environmental variables were analyzed. Species composition of Quercusaliena var. Acuteserrata community was mainly effected by elevation, slope, annual meantemperature and crown density and impact intensity being annual mean temperature>elevation> crown density> slope. Species richness was mainly influenced by elevation,aspect, annual mean temperature, soil thickness, soil pH value and total potassium content,which exhibited a significant reverse hump-shaped relationship with elevation, increasedlinearly from cold-moist aspect to warm-dry aspect, decreased linearly with increase of annualmean temperature and soil thickness, while had more complex changes with soil pH value andtotal potassium content gradients.The major environmental variables impacting species composition of Betulaalbo-sinensis community were annual mean temperature, soil thickness, carbon: nitrogenratio, total potassium content and crown density and effect intensity being soil carbon:nitrogen ratio> soil thickness> crown density> total potassium content> annual meantemperature. No environmental variables significantly impacting species richness of Betulaalbo-sinensis community were detected.Species composition of Betula utilis community mainly related to soil pH value andcrown density and impact intensity being crown density> soil pH value. Species richness wasmainly associated with soil carbon: nitrogen ratio and total phosphorus, and which displayeda significant reverse hump-shaped relationship with the former and significant linear increasewith increase of the later.The most environmental variable influencing species composition of Abies fargesiicommunity was aspect, followed by stone cover, and soil available phosphorus content,elevation, available potassium content, litter thickness, and crown density decreasedsuccessively. Species richness of Abies fargesii community showed a significant reversehump-shaped relationship with elevation, significantly unidomal relationships with litterthickness and soil total nitrogen content, significant linear relationships with soil organicmatter content and carbon: nitrogen ratio, which decreased with decrease of the twoenvironmental gradients. Species richness increased with increasing crown density, but whencrown density reached to60%, basically, it did not change any more.Species composition of Rhododendron capitatum-Salix cupularis-Carex capilliformis var. major community mainly related to aspect, soil organic matter content and crown densityand impact intensity being crown density> aspect> soil organic matter content. Speciesrichness of this community mainly related to elevation, annual mean temperature, litterthickness, soil organic matter content, total nitrogen content, soil carbon: nitrogen ratio,alkali-hydrolyzable nitrogen content, total phosphorus content and crown density. Speciesrichness linearly declined with increasing elevation, soil total nitrogen content,alkali-hydrolyzable nitrogen content and total phosphorus, and linearly increased withincreasing annual mean temperature and crown density, while demonstrated significantreverse hump-shaped relationships with litter thickness, soil organic matter content andcarbon: nitrogen ratio.Vegetation, species composition and diversity in Taibai Mountain were effected by avariety of environmental variables. However, topographic and climate variables played themost important role when considering from the entire south slope and the entire north slope ofTaibai Mountain, respectively. For single community, soil variables were more important.While vegetation variable (crown density) had full impact on not only overall vegetation,species composition and diversity on the south slope or on the north slope but also singlecommunity.更多还原