【作者】 佘冬立；

【导师】 邵明安；

【作者基本信息】 中国科学院研究生院（教育部水土保持与生态环境研究中心） ， 土壤学， 2009， 博士

【摘要】 土地利用结构/覆被变化是全球变化研究的热点问题之一。土地利用结构变化引起如土壤水分、土壤养分等物质的循环发生变化。黄土高原是我国水土流失最为严重的地区,也是生态环境建设重点实施的区域之一。近年来该区开展了大量的水土流失治理和植被建设工作,特别是1999年国家在西部10省363县开展的“退耕还林还草”工程以来,对黄土高原土地利用格局及生态环境产生了强烈影响。本文针对黄土高原地区植被恢复的水土环境效应问题,以水蚀风蚀交错带为主要研究区域,选择神木六道沟流域退化生态系统为研究对象开展了试验研究,通过小区—坡面—流域尺度研究了土地利用及其格局与土壤水分、土壤理化特性、植被群落演替特征、生态水文循环特征和径流侵蚀的相互关系,得到了以下主要结果:1、混合的土地利用结构形成了土壤含水量高低不同的斑块镶嵌坡面格局,有助于坡面径流侵蚀的控制。小流域内土壤水分呈中等变异,且具有明显的空间结构,高斯模型可以较好的反映这种结构特征。干旱条件下,土壤水分保持较高的基台值和变程,而异质比相对较低,且它们随土壤剖面分布差异明显,表明土壤含水量的不同能改变土壤水分的变异程度及分布格局。建立了土壤水分空间回归预测模型,干旱条件下土地利用和土壤属性变量综合解释了78.7%-86.5%的土壤水分空间变异;而较湿润条件下增加了地形属性变量,三者综合仅能解释58.2%-77.7%的土壤水分空间变异。模型的平均预测误差和均方根预测误差都较小,模型可用于研究区土壤水分的预测。2、通过不同土地利用空间配置成混合利用结构拦截和减少径流侵蚀,形成了土壤属性坡面斑块镶嵌格局。小流域内除土壤容重外,土壤理化性状呈中等的空间变异,且均具有中等的空间自相关,其变异性由土壤系统内部因素包括土壤质地、矿物、成土过程、地形特征和人类活动造成的外部因素包括施肥和耕作等共同控制。建立了土壤理化属性的多元回归预测模型,不同土壤属性模型的自变量不同,表明不同的土壤理化性状变异受不同的环境因子控制。分析认为合理配置土地利用形成斑块状结构和增加养分的投入可以有效地改善该地区的土壤质量。3、Kriging插值显示混合土地利用结构小区土壤表层水分和容重呈明显的斑块镶嵌格局,与土地利用结构一致,易于形成径流侵蚀的自我调控系统,其中M1小区(柠条-豆地-苜蓿)的年侵蚀模数最低,两年的减蚀率分别达到98%和94%。从侵蚀控制角度出发提出了土地管理策略:通过土地利用配置,提高土壤的空间变异性,形成水文单元镶嵌分布的径流侵蚀自我调控系统;通过农业措施可提高镶嵌结构下各斑块的土壤水分入渗能力,从而提高整个流域产流的土壤入渗能力阈值。4、种植的苜蓿7年内产量持续增加,而后逐渐降低,大约在种植10年左右后演替发育成顶级草地群落—长芒草群落。随着苜蓿生长年限的延长,土壤容重增加,苜蓿人工草地群落阶段对土壤水分消耗严重。但在演替后期,土壤水分是一个补给恢复过程。苜蓿人工草地退化一方面是由于与乡土物种之间的竞争引起,但更重要的因素是苜蓿生长对土壤水分的过渡消耗而形成的土壤水分胁迫。有效控制植物种群密度可以减少苜蓿草地的土壤水分消耗;通过改变土地利用方式,苜蓿草地与农作物、翻耕裸露进行轮作可以有效地恢复因苜蓿生长消耗的土壤水分。5、研究区柠条林冠层平均截留率为24.2%,稳定截留率约为12%。“漏斗”状结构的杏树林冠平均截留率为15.1%,稳定截留率约为10%。相对于农地小区,2007年长芒草、柠条、杏树和撂荒地的减蚀率基本都达到了80%。植被盖度和高度因子对坡面产流产沙的灰色关联度都大于0.5。土地利用类型和土壤性质对土壤水分含量具有重要的影响,各小区浅层土壤水分季节变化剧烈,随土壤深度增加土壤含水量变化减小。柠条和杏树林地剖面土壤水分的季节变化表现为“消耗-补偿”的过程,0-120cm土层储水量表现为:杏树林>沙黄土柠条>黄绵土柠条>硬黄土柠条>风沙土柠条,120-400cm土层由于受降雨的影响较小,土壤储水量未表现出明显的季节变化,其储水量表现为杏树>沙黄土柠条>硬黄土柠条>黄绵土柠条>风沙土柠条。黄绵土种植的4-6年生苜蓿土壤水分消耗大,而农地撂荒后演替发育的浅根系草本对浅层土壤水分的消耗比农作物对水分的消耗要多。6、水量平衡分析结果表明,两个生长季度平均下来,各小区土壤水分基本保持平衡,观测期末土壤储水均略有盈余,其中农地的盈余最高,为106.9mm,ET/P为71.0%。单个生长季各小区水量平衡具有一定的差异性,2008年黄绵土、风沙土柠条地和黄绵土生长的苜蓿(I)地蒸散量超过了同期降水量,ET/P分别为103.2%、102.5%和104.6%,观测期末土壤储水呈负平衡。用SWAP模型的模拟结果表明,土壤水分和储水量模拟结果与实测值具有很好的一致性;长芒草地水分收支基本平衡,苜蓿草地的水分支出是农地的1.38倍,其中苜蓿的蒸腾耗水量是农地绿豆的3.88倍,这是引起苜蓿草地群落过度消耗土壤储水而呈现负补偿的主要原因。可见,农地退耕还林还草后会增加SVAT系统水分支出,如果植被群落耗水过大很可能使土壤干化。植被恢复的水土环境效应研究从一定角度评价了水蚀风蚀交错带退耕还林(草)工程对环境的影响,同时也为进一步开展科学合理的植被恢复提供指导。本文在野外测定的基础上,对水蚀风蚀带小流域植被恢复的水土环境效应进行了研究,并主要就土地利用及格局改变下土壤水分、土壤水分物理-养分属性空间分异特性及其水土保持效应、人工建设植被的群落演替特征与土壤水分物理属性的耦合关系及不同生态系统水分循环过程和植被耗水规律等问题进行了较为深入的分析。希望本文对于不同景观尺度下土壤属性时空变异及SVAT系统水分循环的基础理论研究提供借鉴,对黄土高原以小流域为基本单元的生态环境建设具有实践指导意义。基于土地利用格局下土壤属性时空变异规律的生态过程模拟似应是这一领域今后研究的重点之一。更多还原

【Abstract】 In recent years, much attention has been focused on the land use/cover change at various scales. Land use/cover change may influence soil water and soil nutrients, as well as various material cycles. The Loess Plateau is very serious in soil erosion and water loss and is one of the key regions for eco-environmental construction. Much effort on soil erosion control and ecosystem restoration has been made in recent years, especially the project of“Conversion of Farmland to Forest and Grassland Regeneration”, which initiated in 1999 and carried out in 10 provinces including 363 counties located in mid- and western China. Extensive vegetation restoration was implemented and land use/cover has changed during the 8 years since this program started. The paper mainly considered the soil-water environmental effects of revegetation on the Plateau. The primary study area of this project was the crisscross area of wind-water erosion in the north part of the Plateau. Liu Daogou watershed was the representative small watershed for the severely degraded ecosystems in this area. Through two years field observations, the effects of land use and its patterns on soil moisture, soil water-physical and nutrient properties, successional characteristics of vegetation community, eco-hydrological cycle and runoff and erosion were studied in plot, slope, and small watershed scales. The results are as follows:1. The mixed land use structures formed the patterns of plaque mosaic with different soil moisture in the slope, which was helpful to slope runoff and erosion controlling. In the small watershed, soil moisture presented moderate variability, and had obvious spatial structures, which could be described by Gaussian models. Compared to the moist condition, the soil moisture had higher sill and range, but lower nugget-to-sill ratio under dry condition, indicating that precipitation could change the intensity of soil moisture variability and the distribution structure. We constructed the spatial regression-prediction models of soil moisture. In these models, land uses and soil attribution variables could synthetically explain 78.7%-86.5% of the soil moisture spatial variability under dry condition; but under moist condition, integrated the land uses, soil texture and topographic variables could only explain 58.2%-77.7% of the variability. The models had small mean prediction errors (MPE) and root mean square prediction errors (RMSPE), which could be used in the research region to predict soil moisture.2. Mixed land uses patterns developed by the spatial arrangement of different land uses could trap the runoff and sediments, which ultimately formed the plaque mosaic patterns of soil physcical and chemical properties on the slope. Except for soil bulk density in the small watershed, soil physical and chemical properties presented moderate variability, and had moderate spatial autocorrelation. These spatial variabilities could be controlled by intrinsic variations in soil characteristics (texture, mineralogy and soil genesis processes) and extrinsic variations (soil fertilization and cultivation practices). The study built the spatial multiple regression-prediction models of soil properties. There were different variables entered to the different soil properties prediction models, which meant that at different soil depths, there were different environmental factors controlled the spatial variabilities of different soil properties. The results suggested that creating a mosaic pattern to increasing spatial variation of areas by land use arrangement to trap soil nutrients to stay in the ecosystem and more nutrient matter input such as manure addition and crop residues return would improve the soil quality effectively on the hilly area of the Plateau.3. Kriging interpolation represented that soil surface moisture and bulk density in mixed land-use structure plot had apparent characteristics of plaque mosaic pattern, which were uniform with land-use structures and easy to form self-regulation system of runoff erosion. M1 plot (korshinsk peashrub-mung bean-alfalfa) had the smallest annual erosion modulus, and the erosion reduction rate in the research two years reached to 98% and 94%, respectively. With regards to the control of erosion, this study suggested two practices for land management, i.e., creating a mosaic pattern by land use arrangement and raising the soil infiltration capacity within the spatial mosaic pattern.4. Alfalfa yield increases for about seven years after seeding, then, it declines and alfalfa is replaced by a natural community, dominated by stipa bungeana, that begins to thrive about ten years after seeding with alfalfa. Soil bulk density increases over time and soil water is severely depleted under alfalfa. The decline in alfalfa yields is most likely related to water stress resulting from depletion of soil water at deeper depths of the soil profiles, rather than by competition from native species. Restricting the plant density could reduce water consumption by alfalfa; while soil water recharge may be facilitated by rotating alfalfa with other crops, natural vegetation or by leaving the soil bare for more than three years before replanting.5. In the study site, the average canopy interception rate of korshinsk peashrub was 24.2%, and the steady interception rate was about 12%. The average interception rate of apricots canopy with“Funnel”shape was 15.1%, at the same time the steady interception rate was about 10%. Compared to farmland, the erosion reduction rate of stipa bungeana land, korshinsk peashrub land, apricot land and fallow land were greater than 80% in 2007. Gray correlation analysis suggested that vegetation coverage and plant height were the the important factors affecting the runoff and sediment production on sloping land, whose gray correlation degrees all exceeded 0.5. The soil moisture in a profile was greatly affected by different land uses and soil texture. At top layers for every plot, soil moisture varied seasonally, but with the increase of soil depth, these variations were reduced. The soil moisture in the soil profiles showed a“consumption-compensation”seasonal process in korshinsk peashrub and apricot land. The soil water storage in 0-120cm was apricot> korshinsk peashrub in sandy loess soil> korshinsk peashrub in loess soil> korshinsk peashrub in hard loess soil> korshinsk peashrub in sandy soil. The soil water storage didn’t show obvious seasonal changes with small effect of rainfall at 120-400cm soil depth. The soil water storage in 120-400cm was apricot> korshinsk peashrub in sand loess soil> korshinsk peashrub in hard loess soil > korshinsk peashrub in loess soil> korshinsk peashrub in sand soil. 4-6 years old M.sativa used more water in 0-400cm soil layer, and the soil moisture consumption of the top layer by the shallow root herbage in fallow land was greater than in crop land by bean.6. The water balance of different land uses indicated that the soil water kept balance for the each plot over the whole study period. There was a slight surplus in soil water storage during the end of the observation period on each land use, especially farmland, which had the highest surplus, being 106.9mm, and the average evapotranspiration/precipitation (ET/P) was 71.0%. In each experimental year, the soil water balance condition had small difference among the land uses. Korshinsk peashrub in loess soil and in sand soil and the alfalfa (I) grown on loess soil in 2008 had more evapotranspiration than precipitation in the study period, and ET/P were 103.2%, 102.5% and 104.6%, respectively. In the end of the observation period, soil water storages were in negative balance. The simulation of water flow by SWAP model represented that the predicted values of soil moisture and soil water storage had obviously consistent with the measured values. The water input on stipa bungeana grassland was roughly equal to the output. The water output on alfalfa grassland was 1.38 times of that on cropland, and the evapotranspiration of alfalfa was 3.88 times of that on cropland with mung bean, which was the main reason causing soil water deficit on alfalfa grassland. So conversion of farmland to forest and grassland regeneration will increase the output of water in the SVAT system, if there is a great deal of water consumption by vegetation community, it will lead to soil desiccation.Based on a small watershed, the studies of the soil-water environmental effects of revegetation evaluated the project of“Conversion of Farmland to Forest and Grassland Regeneration”in the Wind-water Erosion Crisscross Zone, but also could be as recommendation for revegetation practice. In this paper, issues related to the soil-water environmental effects including temporal-spatial variability of soil moisture, soil-water physical and nutrient properties under the effect of land use and its patterns, successional characteristics of vegetation community and their relationships with soil-water physical properties under vegetation construction processes, the water flow of soil-vegetation-atmosphere-transport and its consumption characteristics of the representative sparse vegetation in the region are analyzed. Hopefully, the paper could be of some importance for both the basic research of soil variability, water cycle of SVAT system and practical application of vegetation restoration. In the future, more accurate modeling of ecological processes should be done by taking the characteristics of temporal-spatial soil variability based on land use patterns into consideration.更多还原