【作者】 曹建军；

【导师】 徐建华；

【作者基本信息】 华东师范大学 ， 地图学与地理信息系统， 2014， 博士

【摘要】 长期以来,水土流失一直是我国“头号”环境问题。植被是水土流失的关键控制因子,因而植树造林一度成为治理水土流失的首要措施。经过多年治理,我国人工林面积已接近世界三分之一,居世界首位,南方红壤区的森林覆盖率已达50%以上。但林下普遍缺少灌木或草本植被覆盖,水土流失(“林下水蚀区”)问题仍然突出,称之为“远看青山在,近看水土流”。“林下水蚀区”的存在,给水土保持工作提出了新的课题,“林下水蚀区”土壤侵蚀的发生、监测和评估研究是水土保持规划管理和学科发展的迫切需要。在水力侵蚀地区,森林通过冠层截留降水以及枯落物和根系的防蚀作用发挥其水土保持功能,故林冠、枯落物和根系构成了森林独特的水土流失防护体系。林下枯落物及灌草层由于贴近地表,在消除雨滴击溅动能、降低坡面流速、延长径流历时和增加土壤入渗方面发挥着不可忽视的作用。我国的造林工作已取得巨大成果,但主要强调林冠覆盖,而森林覆盖率指标并不能反映林地的垂直层次结构,在表征林地水平覆盖方面亦有别于水土保持常用的植被覆盖度指标。在使用最为广泛的通用水土流失方程(USLE)中,植被覆盖和管理因子C值的计算精确与否,亦主要决定于所用的植被参数。由于不同地区、不同的研究者所用植被指标的差异,同类植被覆盖与管理因子C值的计算结果可能相差10-20倍。水土保持工作中理想的植被指标应综合表征植被的水平和垂直结构信息,在测量和应用方面具有统一的规范,以便于“林下水蚀区”土壤侵蚀的发生机理研究和区域植被的恢复重建与评估,并提升研究的科学性和通用性。影响森林植被水土保持效应评估的另一个问题是植被参数遥感反演中的不确定性。应用遥感技术是目前大范围植被及生态调查的唯一途径。国内外研究表明,在植被参数的定量遥感研究中,遥感影像预处理(如辐射校正)、遥感信息提取(如各种植被指数)以及反演模型的建立和应用(包括经验模型和各种机理模型)这三个关键过程构成反演结果不确定性的重要来源。在植被结构指标与遥感植被指数的关系建模中,综合考虑上述不确定性因素,有利于充分挖掘遥感影像的信息,从而提高反演精度。我国有关植被的水土保持效应研究已取得诸多成果,但大多研究集中在黄土丘陵区,这就限制了研究成果在其它土壤类型区的应用。福建省长汀县河田镇是我国南方典型的严重水土流失地区。在20世纪80年代初,为彻底根治河田镇的水土流失,福建省政府组织八个单位联合攻关,结合多种措施开展水土保持生态恢复工作。经过二十多年的努力,该镇的水土流失程度己逐步减轻。但是作为先锋树种的马尾松林分布广泛,“林下侵蚀”问题在该镇各地不同程度地存在。本文通过新建土壤侵蚀试验小区,选取叶面积指数、植被覆盖度和植被绿量3种植被结构指标,研究马尾松及其与速生草本植被宽叶雀稗结合的水土保持效应,并根据水土保持“最优”值和遥感反演的区域植被结构,评价研究区的植被恢复程度,以期为“林下水蚀区”的研究和治理提供理论基础和技术支撑。本文基于福建省长汀县河田镇2010年初新建的12个土壤侵蚀试验小区,通过自然降雨条件下小区植被、产流和产沙连续一年共84次的观测,研究次降雨、季度和全年时间尺度下植被与产流、产沙的关系。使用叶面积指数(LAI)、植被覆盖度(VFC)和植被绿量(LVV)这3种植被结构指标表征植被的空间分布密度,并将稀疏马尾松纯林和林草结合小区的产流、产沙特征与对照小区比较,分析马尾松纯林和林草结合植被的水土保持效应；通过草地、灌木林、针叶林、针阔叶混交林和阔叶林5种植被演替群落结构指标的野外调查和多角度遥感反演,分析区域植被相对于水土保持“最优”值的恢复程度。主要结论如下：(1)林、草和二者结合植被具有不同的水土保持效应。较之对照小区的地表径流量,季度和全年马尾松纯林小区可增加8%-22%,纯草小区减少3%-28%,林草结合小区减少18%-61%。较之对照小区的土壤侵蚀量,各季和全年马尾松纯林小区减少62%-75%,纯草及林草结合小区均减少97%-99%。林草结合植被的水土保持效应并非林、草植被水土保持效应之和,且水土保持效应具有时间变化特性,这表明对林、草及二者结合植被的结构特征和水土保持效应有必要进行深入的定量表征,以期系统研究二者的客观关系。(2)在不同时间尺度下基于不同植被指标的林草植被水土保持效应计算模型精度不一。本研究表明,植被覆盖度(VFC)和绿量(LVV)可进行林地植被不同水土保持效应的表征。不同水土保持效应的计算模型或季度优于全年,或全年优于季度,或季度和全年相当,而在次降雨条件下,植被结构指标与植被水土保持效应的关系模型难以建立。(3)遥感技术与水土保持研究的结合是侵蚀区植被恢复度监测和研究的可靠途径。精确的区域植被结构多角度遥感反演与植被的水土保持效应研究相结合,为侵蚀区植被相对于水土保持“最优”结构的恢复程度之监测与评估提供了基础。本研究表明,福建省长汀县河田镇植被的总体恢复度较低,在季度和全年尺度都仅为27%,绿量恢复度<20%的面积占该镇总面积的71%以上,而绿量恢复度>80%的面积仅占18%。更多还原

【Abstract】 Water and soil loss is a focus of various ecological problems in China, and a root stock of ecological deterioration and poverty accumulation. The red soil area,covering2030000km2,21%of total land area in China, had once become a "red desert" in some regions for its widest and severest water and soil loss in southern China. With forest planting and conservation by forestry and soil and water conservation departments in recent decades, forest-coverage rate had increased obviously in many provinces, most of them exceeded50%, some were more than70%. Nevertheless, soil erosion occurred moderately or even intensely on forested areas, for the bareness of under-forest soil without close-cover of vegetation or residue. This under-forest soil erosion requires scientifically monitoring and calculating of water and soil conservation effects for vegetations on forested areas, through feasible vegetation indices indicating occurrence mechanism of under-forest soil erosion, and objectively evaluating of vegetation restoration degree on such water-eroded areas, so as to make a foundation for planning and management of water and soil conservation. On12experimental plots built in early2010in Hetian town, Changting county, Fujian province, this paper focuses on the analysis of the relationships between vegetation structure and its effects on water and soil conservation, with56measurements of vegetation, runoff, and soil erosion parameters, at3time scales, single rain fall, seasonal, and all-year. Three parameters of vegetation structures, leaf area index (LAI), vegetation fractional coverage (VFC), and live vegetation volumn (LVV,the production of LAI and VFC) were used to indicate spatial distribution density of vegetation, and the water and soil conservation effects of sparse Pinus massoniana and forest-grass combination were analyzed through comparing the runoff and soil erosion on the plots planted with sparse Pinus massoniana and forest-grass combination with those on the control plot; Through field measurement and remote-sensedly inversion of vegetation structures in different vegetation succession communities, grass, shrub, coniferous forest, coniferous-broad-leaf forest,and broad-leaf forest, local vegetation restoration degree (VRD) was evaluated based on "best" vegetation structures evaluated for water and soil conservation. The main conclusions are presented below:(1) Water and soil conservation effects of forest, grass, and the combination of forest-grass are different. Compared to the control plot, the seasonal and all-year surface runoff increased by5%-16%on the plots planted with pure Pinus massoniana, while decreased by1%～33%on the plots planted with pure grass(Paspalurn wettsteinii), and decreased by15%～54%on the plots planted with the vegetation combination of tree (Pinus massoniana) and grass (Paspalurn wettsteinii). The seasonal and all-year soil loss decreased by56%～70%on the plots planted with pure Pinus massoniana,and decreased by91%～99%on the plots planted with pure Pinus massoniana or with the vegetation combination of tree (Pinus massoniana) and grass (Paspalurn wettsteinii). Water and soil conservation effects of forest-grass are not those additions of pure forest and grass, and there are temporal changes in the water and soil conservation effects. Consequently, it is required to finely and quantitatively indicate different vegetation structures and their effects on. water and soil conservation,so as to systematically study the relationships between them.(2) The accuracies of results estimating water and soil conservation effects of forests and grasses are different for the models based on different vegetation indices and time scales. Vegetation fractional coverage (VFC) and live vegetation volumn (LVV) could be used to indicate the water and soil conservation effects of vegetation on forested area. Among various models calculating the water and soil conservation effects of forests and grasses, some seasonal models are better than all-year ones while others the reverse, or close to each other. Under single rainfall, however, it is difficult to establish models between vegetation structures and water and soil conservation effects.(3) The union of remote sensing techniques and researches on water and soil conservation effects of vegetation is a credible means to monitor and study vegetation restoration degree on eroded areas. The monitoring and evaluation of local to regional vegetation restoration degree, compared with the "best" vegetation structures for water and soil conservation,are supported by accurate inversion of vegetation structures by remote sensing, and studies on water and soil conservation effects of vegetation on eroded areas. Our study showed that, vegetation restoration degree of the study area are at the low level, and are only30%at both seasonal and all-year time scales. The area where live vegetation volumn (LVV) restoration degree<20%and>80%respectively covered more than60%and less than20%of the total area in Hetian town, Changting county, Fujian province.更多还原