【作者】 秦晓威；

【导师】 杨改河； 王得祥；

【作者基本信息】 西北农林科技大学 ， 作物栽培学与耕作学， 2012， 博士

【摘要】 生态环境退化和极端气候事件频发使物种多样性丧失，因而物种多样性保护及与此相关的多样性维持机制成为当今生态学领域亟待研究的重要科学问题。有冰期植物“避难所”之称的秦岭，是中国11个具有全球意义的陆地生物多样性关键保护区。然而，关于秦岭植物物种的分布和群落物种的共存机制研究较少。鉴于此，本研究选择秦岭主峰太白山植物群落为研究对象，通过资料收集和野外调查，从区域物种多样性、局域物种多样性及自然干扰三个角度分析：（1）太白山草本植物丰富度的海拔梯度格局及成因；（2）局域与区域种库对林隙-非林隙系统草本植物多样性的影响；（3）林隙干扰对森林植物群落草本植物多样性的影响。以期揭示局域和区域生态过程在草本植物群落构建中的相对重要性，为秦岭山地植物群落物种多样性保护提供科学依据。主要研究结果如下：1.共统计太白山草本植物1094种，隶属69科410属。草本植物丰富度的海拔梯度格局呈单峰分布，且丰富度峰值在海拔1200～1300m。通过物种种域Rhode法分析表明，草本植物平均种域沿海拔梯度呈单峰分布，不支持Stevens的“Rapoport法则”假说。空间自回归模型分析表明，反映山体垂直特征的面积、几何（边界）限制、温度和河网密度等因素能显著解释草本植物丰富度海拔梯度变异的75%。其中，面积单变量能解释46.9%，但其与几何（边界）限制的共线性问题掩盖了面积因素的相对重要性。几何（边界）限制的确对物种多样性，特别是大种域物种多样性具有显著的影响（R2=0.601），但与小种域种的分布相关性不显著（R2=0.007）。当温度和河网密度综合作用时能显著解释小种域种丰富度变化的36%，表明小种域物种更容易受局域和区域因子的综合作用影响。2.利用Bootstrap非参数统计方法对7种典型植物群落潜在区域物种丰富度与密闭林分和林隙干扰条件下的局域物种丰富度相关性分析表明：密闭林分下，1m~2局域样方草本植物平均丰富度与区域物种丰富度相互独立，呈“饱和关系”；林隙干扰下，1m~2局域样方草本植物平均丰富度随区域物种丰富度的增加而显著线性上升，呈“非饱和关系”。表明林隙干扰可以改变局域与区域物种丰富度的相关关系。林隙干扰使得森林系统向非平衡态变化，但其与密闭林分水平的整合可能促使新的复合稳定态再次出现。局域尺度上的随机过程与区域尺度上的相对稳定性互相兼容可能维持着植物群落的物种多样性。3.锐齿栎群落林隙调查记录57种草本植物，密闭林分41种。林隙内物种丰富度随林隙大小梯度显著增加（R~2=0.189, P=0.016）。但仅5种草本植物依赖于面积>130m~2的林隙，大部分物种在所调查林隙内的分布是泛性的，未呈现随林隙大小梯度连续变化的分布规律（P=0.120）。个体数≥4的51种物种与林隙环境变量的CCA梯度分析表明，仅36.3%的物种与生境显著相关。4.红桦群落林隙调查记录108种草本植物，密闭林分74种。林隙内0.25m~2和1m~2局域样方草本植物平均丰富度与林隙总物种数均显著相关（r=0.727和r=0.754），且随林隙总物种数的增加而线性增加。同时，林隙面积也能显著增加局域物种丰富度和林隙内总物种数。但控制林隙面积变量的多元回归分析表明，0.25m~2和1m~2局域物种丰富度仍与林隙内总物种数呈显著线性正相关（R~2=0.595和R~2=0.665）。方差分离表明，林隙面积仅能解释0.25m~2和1m~2局域物种丰富度变异的8.0%和1.0%，而林隙物种总丰富度能解释27.2%和39.4%，表明区域物种丰富度的变异在一定程度上决定着局域物种丰富度的组成。5.巴山冷杉群落林隙调查记录69种草本植物，密闭林分42种。ANOSIM检验表明，林隙与密闭林分物种组成差异显著(ANOSIM, R=0.509, P=0.001)，但不同年龄组(ANOSIM, R=0.035, P=0.191)和不同大小组(ANOSIM, R=0.089,P=0.057)林隙内物种组成差异并不显著。在统计上，林隙内草本植物多度分布符合中性理论零和多项式模型的预测分布(χ2=5.58, P=0.90)。其中，有15种草本植物最大多度出现在面积>120m~2的林隙，仅8种依赖于面积>146.1m~2的林隙，其余88.5%的物种广泛分布在所调查的林隙。个体数≥5的55种物种与环境变量的CCA梯度分析表明，仅27%的物种与环境变量显著相关。6.太白红杉群落林隙调查记录84种草本植物，密闭林分52种，其物种多度分布均符合中性模型的预测，且多度的观测值均落在95%置信区间。其中，仅19种物种最大多度出现在面积>140m~2的林隙，5种物种仅依赖于面积>229m~2的林隙，其余77.3%的物种广泛分布于所调查的林隙。另外，除林地早熟禾(Poanemoralis)、披针苔草(Carex lanceolata)和唐松草外(Thalictrum aquilegiifoliumvar. sibiricum)，MANOVA检验表明，12种早期演替种的物种多度组成在林隙与密闭林分差异不显著。个体数≥5的68种物种与环境变量的CCA梯度分析表明，30.9%的物种与环境变量显著相关，表明林隙对草本植物组成并未起到主要的筛选作用，在维持物种多样性上的作用偏于中性。更多还原

【Abstract】 Biodiversity conservation is an urgent matter all over the world. Ecologicaldegradation and its corollary-biodiversity loss-have become pressing global issues.The Qinling Mountains as biology refugia after the Pleistocene glaciations has beendesignated one of the11key regions terresterial “biodiversity hotspots” in China.However, elevational patterns of species richness and mechanisms of speciescoexistence in ecological communities are still not fully known in the region.Therefore, in order to assess the relative importance of local and regional processes instructing herbaceous communities across a regional of Mt. Taibai in the middle part ofQingling Mountains. Using well-documented regional flora, local species richnessand small-scale disturbances information, the purposes of the present study is to:(1)examine the elevational pattern of herbaceous species richness and related hypothesesof mechanisms in Taibai Mountain;(2) evaluate the effects of local and regionalprocesses on herbaceous plant assemblages in temperate forest canopy gaps in TaibaiMountain;(3) detect the effects of natural canopy gap disturbances on herbaceousplant diversity in different temperate forest of Taibai Mountain. The main results wereas follows:1. Herbaceous species diversity totaled1094species in410genera and69families in Mt. Taibai, Qinling Mountains. The species richness of the1094recordedspecies showed a hump-shaped pattern with peak at1200～1300m, and the speciesdistributed in high and low elevations had smaller average elevational range sizes thanthose in mid-elevations accoding to the Rhode’s methods. Thus, Rapoport’ rule wasnot backed. The best fit models, CARs identified area, geometric constraints,temperature, and drainage density variables as stronger predtctors for overall species,and explained c.75%of the variation in species richness. The area explained46.9%of the richness variance, but the area itself had very low influence when the geometricconstraints was controlled because area and geometric constraints were highlycollinear variables. While the geometric constraints model predictions were well correlated with herbaceous species richness for all species and for large-rangedspecies (R~2=0.601), the explanatory power of the geometric constraints for thesmall-ranged species was lower (R~2=0.007). For small-ranged species, which showeda distribution only strongly correlated with drainage density and temperature. Thus,herbaceous species richness peak deviated notably from geometric constraints modelmay be largely caused by plently of small-ranged species, whose distribution partlyaffected by local-scale rather than regional-scale environmental factors.2. The Non-parametric Brootstrap was used to test the strengths of therelationships between potential regional species richness and local species richness inunderstories and canopy gaps conditions of7different vegetation types in TaibaiMountain. We found that, without canopy gap disturbances, the relationship betweenlocal and regional species richness was weak or nonexistent, suggesting speciessaturation in1m~2small-scale plots; but under canopy gap disturbances, local speciesrichness was positive function of regional species richness, because1m~2small-scaleplots were coloninzed from the potential species pool. These results suggested thatnatural canopy gap disturbances could alter local-regional richness relationships inTaibai Mountain. Thus, to investigate natural canopy gap disturbances may be animportant tool to understanding local and regional processes in determining speciesdiversity.3. We found57herbaceous species in the Quercus aliena var. acuteserratacommunity canopy gaps, and41species within closed canopy sites. The speciesrichness in herbaceous layer in gaps was significantly positively correlated with gapsize (R~2=0.189, P=0.016), but only7of57species displayed gap sizes specialization(>130m~2),87.7%herbaceous species were found across all gap sizes. Nomini-succesional sequence was observed in the herbaceous distribution with the gapsize gradient (P=0.12). Additionally, canonical correspondence analysis and randompermutation tests suggested that only36.3%of herbaceous with abundance≥4wereassociated with environmental variables in canopy gaps. It is concluded thatherbaceous species composition in canopy gaps was not constrainted significantly bygap characteristics in the Quercus aliena var. acuteserrata community.4. We found108herbaceous species in the Betula albosinensis communitycanopy gaps, and74species within closed canopy sites. The mean herbaceous speciesrichness in0.25m~2and1m~2within canopy gaps was correlated with the speciesrichness of the whole gap (r=0.727and r=0.754, respectively). And the two small-scales species richness showed a positive function of gap species richness,respectively. Additionally, there was also a significance correlation between gap sizewith gap species richness and small-scale richness, which might form a spuriousunsaturation local-regional richness relationships. But the unsaturation local-regionalrichness relationships were also found after controlling the effect of gap size inmultiple regression (R2=0.595for0.25m~2, and R2=0.665for1m~2). Moreover,analysis of variance showed that gap size only explained8.0%and1.0%of variationin species richness at0.25m~2and1m~2, whereas gap species richness could explaind27.2%and39.4%, respectively. These results suggested that even relativesmall-scales, local species richness can be constrained by the size of species pool inimmediately surrounding region.5. We found69herbacous species in the Abies fargesii community canopy gaps,and42species within closed canopy sites. Although herbaceous species compositionin gaps significantly differed from closed canopy (ANOSIM, R=0.509, P=0.001),there were no significant differences among gaps of different age groups (ANOSIM,R=0.035, P=0.191) or gaps of different size groups (ANOSIM, R=0.089, P=0.057).Herbaceous species abundance distribution within canopy gaps was well fit by aneutral model (χ2=5.58, P=0.90), showing that relative abundance was qualitativelyconsistent with stochastic processes. Moreover,15of69herbaceous species foundmore commonly in larger gaps (>120m~2), and eight species displayed gap sizespecialization (>146.1m~2). However,88.5%species were observed across the gapsize gradient. Additionally, the CCA ordination suggested that only27%of the55herbaceous species with abundance≥5was associated with environmental variables inthe Abies fargesii community gaps.6. We found84herbaceous species in the Larix chinensis community canopygaps, and52species within closed canopy sites. Herbaceous species-abundancedistribution within canopy gaps and closed canopy sites was well fit by zero-summultinomial, respectively, and no of the species in two of these communities fell outof the95%confidence envelope of the best fit neutral model distribution ofrank-abundance curves. And of the84herbaceous species recorded in the canopygaps,77.3%species were found across all gap sizes, and only5herbaceous speciesdisplayed gap size specialization. An additional14species were more common inlarger gaps but there was no gradient in composition of herbaceous species across gapsize. Additionally, except species of Poa nemoralis, Carex lanceolata, and Thalictrum aquilegiifolium var. sibiricum, the abundances of12early successional species werenot significant differed between canopy gaps and non-gaps. The CCA oridinationshowed that30.9%of68species with abundance≥5was correlated withenvironmental variables in the Larix chinensis community canopy gaps. In summary,the species compostion in gaps was not constrainted significnantly by gap traits;rather species were distributed stochastically, likely through by random dispersal andrecruitment limitation of species from the surrounding available species pool.更多还原