【作者】 刘德林；

【导师】 李壁成；

【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， 2010， 博士

【摘要】 黄土高原位于黄河中上游,地处我国幅员中心位置,是中华农耕文化的发祥地,在我国生态环境建设和经济发展中具有重要战略地位。小流域是黄土高原水土流失综合治理和社会经济的基本研究单元,研究黄土高原小流域尺度上的土地利用/覆被格局变化具有重要的理论和现实意义。不合理的土地利用加剧了黄土高原的水土流失,从而导致土地资源退化和群众生活贫困。因此,开展小流域土地利用/格局变化研究,深入分析其形成原因和驱动力,提出合理、持续的土地利用优化配置模式,进行土地生态功能区规划和实施,是从根本上防治水土流失、改善生态环境和发展生态农业,实现生态环境良性循环和农业可持续发展的重要途径。本文以“景观生态学”理论与方法为指导,运用RS/GIS等新的技术手段,利用黄土高原上黄试区近30年丰富的定位试验研究资料,较为全面的研究小流域尺度的土地利用格局变化、土地资源的适宜性、并对试区进行了生态功能区规划。同时,针对小流域大比例尺地理信息数据缺乏现状,利用试区1:1万土地利用数据库,探讨小流域尺度下景观指数随栅格粒度变化和土地利用分类不同的效应问题。论文主要研究结果与结论如下:⑴数据分辨率和土地利用分类对景观指数的影响基于土地利用/景观专题图(栅格数据)所计算的景观指数大多都存在粒度效应问题,景观指数的粒度效应主要是受到制图比例尺的影响,而与土地利用变化关系不大,对1:1万比例尺下的土地利用/景观图进行景观指数计算的适宜粒度范围为5-15m。同时,景观指数在一定程度上受土地利用分类系统变化的影响,分类系统过细或过粗都会对这些景观指数产生较大的影响;不同历史时期的数据资料对研究土地利用分类系统变化对景观指数的影响不大。⑵土地利用格局动态变化特征近30年来上黄试区各土地利用类型均发生了剧烈的变化。根据试区景观基质变化情况将其划分为3个研究阶段,即1982-1990(草地为景观基质)、1990-2002(耕地为景观基质)和2002-2008(灌木林地为景观基质);在整个研究时段,草地、耕地分别以6.96%/年和3.40%/年的速率递减,而林地、居民用地和果园则分别以19.45%/年、3.22%/年和7.08%/年的速率增加;上黄试区土地利用变化在不同阶段受着不同驱动力的影响,在近30年可分3个阶段:调整土地利用结构,改善生态环境(1982-1990年)阶段;“三化两提高”(1991-2002年)阶段和退耕还林还草整体推进(2002-2008年)阶段。⑶土地利用景观格局特征试区内各景观组分的面积分布极不平衡,林地面积最大,耕地次之,二者面积之和占试区总面积806.39ha的83.03%。面积加权的形状指数表明:水域、未利用地、林地和草地受人类活动影响较小,耕地和果园受人类活动的影响较大,居民点最大,符合试区的实际情况。受人类活动和国家政策的影响,上黄试区1982-2008年土地利用景观格局发生了一定的变化,具体表现为土地利用斑块类型在空间上的分布呈现均匀化、景观破碎化程度和景观异质性先增加后减少等特征。⑷土地利用坡度分析上黄试区各土地利用类型的坡度分布较为合理,基本体现了因地制宜、合理配置资源的原则。从退耕还林还草的政策执行情况看,试区仍有约6.81ha的土地需要退耕还林还草,但该面积还占不到全试区总面积的1%,基本完成国家退耕还林还草任务的要求。⑸土地资源适宜性评价地貌部位、坡度、坡向、侵蚀强度、土壤有机质和灌溉条件6个因子基本能反映试区土地资源特征,可作为该流域土地适宜性评价的最终指标体系。评价结果显示:试区高度宜农地的土地面积为115.57 ha,占到全流域土地总面积的14.43%;中度宜农地土地面积为79.78 ha,占全流域面积的9.96%;勉强宜农地面积为121.11 ha,占流域土地总面积的29%;宜林地土地面积为46.8ha,只占流域总面积的5.84%;宜牧地面积为419.04 ha,占到流域总面积的一半以上,所占比重较大;宜林牧地的面积为139.86ha,所占面积接近流域总面积的1/5。⑹土地利用功能区规划运用景观生态学空间格局优化基本原理,结合土地利用格局分析和土地资源利用适宜性评价结果,结合宁夏“六盘山生态经济圈规划”,共划分三个功能区,即:东山生态环境保护功能区、西山旱作农业功能区和平川高效生态农业功能区。东山保护区面积最大,占总面积的51.84%,地形比较破碎,面蚀和沟蚀都很强烈;西山旱作农业区面积占总面积的33%,坡面较为平缓,为中度土壤侵蚀;平川高效生态农业区面积约占总面积的15.16%,这是研究区的白菜心。土地平整,土层深厚,土壤肥力较高,是乡村聚落区。更多还原

【Abstract】 The Loess Plateau, the birthplace of Chinese farming culture, is located in the upper and middle reaches of the Yellow River, the center of China. It has an important strategic position in Chinese ecological environment construction and economic development. A small watershed is the basic research unit for comprehensive soil erosion control and socio-economic on the Loess Plateau. It has important theoretical and practical significance to study the changes of land use/cover pattern at a small watershed scale on the Loess Plateau.Irrational land use accelerated soil erosion on the Loess Plateau, and has resulted in land degradation and poverty life of farmers. Therefore, it is necessary to study land use change and their driving forces, to put forward reasonable and sustainable land use optimization program, and to carry out ecological function zoning. The above work is an important way for comprehensive soil erosion control, improving the eco-environment, developing ecological agriculture, and achieving good circulation of eco-environment and sustainable development of agriculture.In this dissertation, based on the theories and methods of landscape ecology and RS/GIS, the changes of land use pattern, suitability and optimized model of land resources, and ecological function zoning at Shanghuang Small watershed are studied by using the rich experiment data obtained from 1982 to 2008. At the same time, we explored the impact of land use classification and grid resolution on landscape metrics using the land use data base of this small watershed.The results and conclusions of this paper are as follows:⑴Impacts of grid resolution and land use classification on landscape metrics Most landscape metrics calculated from land use maps (grid date) were affected by grid resolution. The grain effect of landscape metrics was mainly affected by drawing scale rather than land use change, and the appropriate grain range for calculating landscape metrics from land use maps at 1:10 000 is from 5m to 15m. At the same time, landscape metrics aere affected by land use classification. Too detailed or too coarse classification system would generate greater impacts on these metrics, and the data in different historical periods are little impacts on the study of the response of landscape metrics to land use classification.⑵Dynamic change characteristics of land use patternAll land use types had undergone substantial changes in the Shanghuang small watershed from 1982 to 2008. According to the change of landscape matrix, land use change throughout the entire study period could be divided into three phases, i.e., grassland (1982-1990), cropland (1990-2002) and forest land (2002-2008). During the entire study period, grassland and cropland were decreased at the rate of 6.96%/yr and 3.40%/yr, respectively, and frost land, residential area and orchard were increased at the rate of 20.69%/yr, 2.12%/yr and 6.12%/yr, respectively. The driving forces of land use could be divided into three phases, i.e., adjust land use structure and improve the ecological environment (1982-1990); optimization of land use, rising of farmer’s quality, and improvement of eco-economic benefits (1990-2002); and the overall advancement of reforestation.⑶Landscape pattern characteristics of land useThe area distribution of each land use type was extremely uneven. The largest area was forest land, and the second largest area is cropland, their area occupied 83.03% of the total area (806.39 ha). The index of SHAPE_AM showed that wetlands, unused land, forest land and grassland less affected by human activities, cropland and orchard greater, and residential area greatest. Affected by the human activities and national policy during the period from 1982 to 2008, the patch types of land use distributed more even, the fragmentation and heterogeneity of landscape increased first, and then decreased.⑷Analysis of land use in different slopeThe area distribution of each land use type in different slope was reasonable, which basically embody the principles of local conditions and rational distribution of resources. Seen from the implementation situation of reforestation, about 6.81 ha of land still need to reforestation, but this area was less than 1% of the total area. Therefore, the task of reforestation in this region was basic completed.⑸Suitability evaluation of land resourcesThe six metrics, including landscape position, slope, aspect, erosion intensity, soil organic matter and irrigation conditions, could be used as the index system for suitability evaluation of land resources. They could basically reflect the characteristics of land use resources of this region. The evaluation results showed that the area highly, moderately and barely suitable for agriculture were 115.57ha, 79.78ha and 121.11ha, respectively, and their area occupied 14.43%, 9.96% and 29% of the total area, respectively. The area suitable for forest land, grassland, and both forest land and grassland were 46.8ha, 419.04ha, and 139.86ha, respectively, and their area occupied 5.84%, 52.31%% and 17.46% of the total area, respectively.⑹Ecological function zoning of land useAccording to the analysis of land use pattern and the results of suitability evaluation of land resources, three functional regions were divided based on spatial structure optimization theory of landscape ecology. They were eco-environmental protection region in east mountain district, dry farming region in west mountain district, and high efficiency ecological agriculture region in flat district, respectively. The area of eco-environmental protection region in east mountain district was largest and occupied 51.84% of the total area. The surface erosion and gully erosion were very strong in this region for its broken terrain. The area of dry farming region in west mountain district occupied 33% of the total area, and the soil erosion was moderate for its gentle slope. The area of high efficiency ecological agriculture region in flat district occupied 15.16% of total region. There were flat land, deep soil and fertile soil, so it is rural settlement region.更多还原