黄土高原样带土壤水分空间变异与动态模拟
文献类型：学位论文
| 作者 | 王聪 |
| 答辩日期 | 2017-05 |
| 文献子类 | 博士 |
| 授予单位 | 中国科学院大学 |
| 授予地点 | 北京 |
| 导师 | 傅伯杰 |
| 关键词 | 黄土高原 样带尺度 Loess Plateau 土壤水分 Transect Scale 空间变异 Soil Moisture 随机模拟 Spatial Variation Stochastic modeling |
| 学位专业 | 生态学 |
| 其他题名 | Spatial variations and temporal dynamics of soil moisture across the transect in the Loess Plateau, China |
| 英文摘要 | 黄土高原大规模植被恢复措施在控制区域水土流失、改善土壤质量等方面起 到积极作用，但不合理的人工恢复植被格局也引起一些负面生态效应，如土壤干 层、产水量减少、植被退化，出现“小老头树”等现象，合理植被格局的依据是 植被承载力，土壤水分是制约植被恢复可持续性的关键因素。 土壤水分的时空动态直接影响植被的生长、演替与分布，开展黄土高原区域 土壤水分空间变异与动态模拟研究对于正确认识植被承载力，指导生态系统恢复 与重建具有要意义。因此，本研究在黄土高原区域上设置样带，根据植被恢复类 型和降雨梯度选择了 92个样点，开展植被群落特征、土壤水分、土壤理化性质 等的调查与测定。采用经典统计、权衡分析、随机模型模拟等方法在区域样带尺 度上分析了不同土地覆盖下土壤水分空间变异、土壤水分与植被、土壤特征的交 互作用，并模拟了样带典型植被的土壤水分动态和水量平衡变化。通过系统研究 获得的主要结果和结论如下： 1.样带上林地和灌丛下土壤水分相比草地和农田体现出更大的变异性，在 年均降雨（MAP）小于 520mm的地区林地和灌丛中层和深层（1-2m，2-3m）土 壤水分已到达可利用土壤水分的下限，土地覆盖类型和降雨是影响样带土壤水分 的重要因子。在 0-3m的土壤剖面内，各层土壤水分均呈现出农田>草地>灌丛> 林地的顺序；各土地覆盖类型下剖面平均含水量与年均降雨显著线性相关 （P<0.05），草地土壤水分受降雨影响更大；协方差的分析的结果表明降雨与土 地覆盖对中层和深层土壤水分有显著性影响，但土地覆盖对表层（0-1m）土壤水 分无显著影响；在 MAP < 520mm的地区，林地和灌丛下的中层和深层土壤水分 已处于接近或者低于土壤萎蔫点的水平。 2.样带上土壤水分与植被、土壤特征之间的交互作用受到植被类型和降雨 的影响。土壤水分（SM）、物种丰富度（SR）、土壤有机碳（SOC）和全氮（TN） 基本随降雨梯度呈现线性增长的趋势，但不同类型下各指标的变化速率不同；通 过计算土壤水分与其他指标之间的权衡程度（均方根偏差，RSMD），反映土壤水 分与其它各指标随降雨梯度变化速率的不一致性，衡量它们之间的交互作用，大 部分样地中土壤水分与各指标（SOC、TN和 SR）的权衡程度均不超过 0.2，但 部分样地的权衡程度较高，达到 0.6；土壤水分与各指标的权衡程度与年均降雨 呈显著正相关（P<0.05, P<0.01），并且受到植被类型的影响；采用线性分段分位 数回归的方法探究各权衡程度对降雨梯度的响应变化，发现 SM-SOC之间的权 衡关系在年均降雨为 570mm处出现转折点，SM-TN之间的权衡关系在研究的降 雨梯度带内未发现转折点，林地和草地中 SM-SR之间的权衡关系分别在年均降 雨为 380mm和 570mm处出现转折点，土壤水分对于土壤有机碳和物种丰富度 的制约作用随降雨梯度带出现了转折变化，随着降雨的增加，土壤水分的制约作 用减缓或者消失。 3.运用随机土壤水分平衡模型，模拟样带两种典型植被（刺槐、长芒草）的 土壤水分动态和水量平衡变化，发现两种典型植被土壤水分动态和水量平衡组分 呈现显著差异。随机土壤水分平衡模型模拟的土壤水分概率密度与羊圈沟的监测 结果具有较好的一致性；刺槐土壤水分动态相较于长芒草对湿润 -干旱阶段交替 变化响应更明显；样带上，刺槐和长芒草蒸散发组分随降雨梯度出现线性变化的 趋势，刺槐水分压力下的蒸散发要显著高于长芒草，其不受水分压力影响的蒸散 发要显著低于长芒草；刺槐受到的动态水分压力基本达到最大值 1，显著高于长 芒草；根据水分压力最小同时水分利用效率最大的最优性原则，建立了植被适宜 性参数以评估植被适宜性，样带上除年均降雨为 560mm的样点外，长芒草的适 宜性参数的值都要小于刺槐，说明样带上长芒草的适宜性更高。 4.根据研究结果，提出了黄土高原植被恢复的对策建议。从植被可利用水分 可持续的角度看，在 MAP<520mm的地区，林地和灌丛不是黄土高原大规模植 被恢复的适宜类型；从恢复物种的适应性角度看，在 MAP<560mm的地区，刺槐 不是适宜的恢复物种；在植被恢复过程中应考虑植被恢复引起的生态效应，在较 干旱地区，土壤水分会对物种丰富度维持和碳固定起到制约作用; Large-scale “Grain-for-Green” project has been implemented in the Loess Plateau since 1999, the vegetation restoration has made remarkable contribution to the soil and water control and soil quality improvement. However, the irrational vegetation restoration pattern has led to negative ecological effects, such as formation of dried soil layer, decreased runoff, vegetation degradation and occurrence of “dwarf and aged” trees. The vegetation restoration should be implemented based on the vegetation carrying capacity. Soil moisture is the most important water source for plant growth in the Loess Plateau, and it’s the key factor affecting the sustainability of vegetation restoration. The spatial variation and temporal dynamics of soil moisture had direct influence on the plant growth and vegetation distribution, characterizing spatial variation and temporal dynamics of soil moisture is significant in understanding the vegetation carrying capacity and guiding the vegetation restoration. Thus, a typical transect was set across Loess Plateau, where 92 sampling points were selected according to the vegetation type and precipitation gradient. Soil sampling and field investigation were conducted at each sampling point. Standard statistical analyses, tradeoff analysis and stochastic soil water balance model were applied to analyse soil moisture spatial variations, explore the interactions between soil moisture and soil-plant variables, model soil moisture dynamics and water balance at transect scale, respectively. The major results and conclusions are: 1. The vertical profiles of soil moisture revealed larger variations in woodland and shrubland compared to grassland and farmland across the transect. Soil moisture in the medium and deep layers of woodland and shrubland has reached the threshold of the available soil water for plant growth. Land cover type and the mean annual precipitation (MAP) were the controlling factors determining the soil moisture variations across the transect. In each soil layer of the profiles (0-3m), soil moisture under each land cover type commonly revealed a consistent trend as farmland > grassland > shrubland > woodland. Significant linear correlation existed between the average soil moisture of the profile and the MAP for each land use type (P<0•05). Hereinto, the soil moisture under the grassland was affected more greatly by precipitation. Covariance analysis confirmed the effects of land cover type and MAP on the soil moisture in depth of 1- 3m, and it showed land cover did not pose significant effects in the surface layer (0- 1m). The soil moisture of woodland or shrubland at the selected points was below or approximate to the permanent wilting point in regions with MAP <520 mm. 2. The interaction between soil moisture and soil-plant variables were affected by the vegetation type and precipitation across the transect. Soil moisture (SM), soil organic carbon (SOC), total nitrogen (TN), and species richness (SR) commonly revealed a linear increasing trend along the precipitation gradient, while the increasing rate varied with the vegetation type. The tradeoff (statistical parameter root mean square deviation, RMSD) was calculated to quantify the uneven rates of same-direction changes between these variables. The relative frequency distribution showed that < 0.2 accounted the majority of these tradeoffs, while the maximum of these tradeoffs could reach 0.6. Correlation analysis showed that the SM-SOC and SM-TN tradeoffs were significantly correlated with MAP across the three vegetation types (woodland, shrubland and grassland), and SM-SR tradeoff was significantly correlated with MAP in grassland and woodland, respectively. The linear piece-wise quantile regression was applied to determine the inflection points of these tradeoffs responses to the precipitation gradient. The inflection point for the SM-SOC tradeoff was detected at MAP=570mm; No inflection point was detected for SM-TN tradeoff; SM-SR tradeoff variation trends were different in the woodland and grassland, and the inflection points were detected at MAP=380mm and MAP=570mm, respectively. Before the turning point, constraint exerted by soil moisture on SOC and SR existed in the relatively arid regions, while the constraint disappears or is lessened in the relatively humid regions in this study. 3. The modeled soil moisture dynamics and water balance of two typical vegetation types (Robinia pseudoacacia and Stipa bungeana) with the stochastic soil water balance model revealed significant differences across the transect. The modeled soil moisture probability density function (pdf) showed a good agreement with the observed soil moisture data in the Yangjuangou catchment. The soil moisture dynamics under Robinia pseudoacacia revealed more obvious responses to the wet and dry period changing. The evapotranspiration components of these two vegetation types revealed a linear changing trends with the precipitation gradient, and the proportion of evapotranspiration under water stress to the rainfall for Robinia pseudoacacia is significantly higher than the Stipa bungeana. Along the transect, Robinia pseudoacacia generally suffered greater water stress than the Stipa bungeana. Individual species fitness (F Value) is defined on an optimality trade-off hypothesis, which states that dryland vegetation patterns are constrained by maximization of water use and simultaneous minimization of water stress. According to the F Value, Stipa bungeana is more suitable compared to Robinia pseudoacacia across the transect except in the most humid site. 4. The vegetation restoration policy in the Loess Plateau should shift to the results- based approach. The woodland/shrubland is non-sustainable with regard to soil water availability and might not be the optimal solution to the current vegetation restoration in the regions with MAP < 520mm. Considering the species fitness, Robinia pseudoacacia is not quite suitable in the regions with MAP < 560mm. In addition, the ecological effects caused by the implementation of vegetation restoration should be carefully considered and assessed. Soil moisture exerted constraint on carbon sequestration or species maintaining in the relatively arid regions. |
| 源URL | [http://ir.rcees.ac.cn/handle/311016/38697]  |
| 专题 | 生态环境研究中心_城市与区域生态国家重点实验室 |
| 作者单位 | 中国科学院生态环境研究中心 |
| 推荐引用方式 GB/T 7714 | 王聪. 黄土高原样带土壤水分空间变异与动态模拟[D]. 北京. 中国科学院大学. 2017. |
入库方式： OAI收割
来源：生态环境研究中心
浏览0
下载0
收藏0
其他版本
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。