【作者】 滕明君；

【导师】 周志翔；

【作者基本信息】 华中农业大学 ， 园林植物与观赏园艺， 2011， 博士

【摘要】 城市化在促进区域经济和社会发展的同时也导致一系列城市生态环境问题,严重威胁区域可持续发展。随着人类城市化活动的加剧,减缓和应对城市化带来的负面效应迫在眉睫。优化区域景观结构,构建生态安全格局,保育和恢复必要的区域生态过程已成为应对城市化环境问题的重要共识。但目前生态安全格局这一策略仍停留在概念阶段,缺乏切实可行的空间规划措施支持。武汉市是我国中部的中心城市,近年来正经历着前所未有的城市化进程,城市景观格局与环境急剧变化,在这一背景下针对区域生态环境问题探讨城市景观调控机制具有重要的理论与现实意义。本文以武汉市为研究对象,在景观生态学、保育生态学、城市生态学、环境科学及相关理论指导下,按照“格局动态—生态环境响应一生态安全格局构建”的研究思路,借助多时序遥感卫星影像,综合运用RS、GIS技术和相关分析方法研究了1987～2007a武汉市景观格局变化特征及驱动机制,分析了城市化背景下森林和湿地景观连接度变化特征和热环境空间分异特征及景观响应机制,在此基础上针对武汉市景观破碎化和城市热岛问题,从不同尺度上提出了武汉市生态安全格局的网络空间构型,为构建具有实践意义的城市生态安全格局提供了理论与方法依据。研究结果表明：(1)1987～2007a武汉市景观格局发生了显著变化,农田和水体面积大幅减少,城乡建设用地面积快速增加,森林面积明显提高,森林和水体景观破碎化程度加剧,城市建设用地集中程度增加,景观整体城市化特征显著。人类活动促使武汉市各类型景观斑块面积差异逐渐变小,复杂程度逐渐降低,景观粒度从粗粒景观向细粒景观转变。城市扩张、基础设施建设及其他人类活动已经明显影响区域景观格局,导致景观异质性持续增加。从1991a开始,武汉市景观整体动态变化度逐步增加,景观变化速度逐渐增加。整个研究期中,各景观类型面积变化具有一定波动性。疏林草地、建筑和裸地的年变化率较高,但不同阶段各景观类型年变化率存在一定差异。农业开发活动、城市化发展、森林建设和降水量增加是1987～1991a景观变化的主要直接驱动因素；城市人口数量增加、农业开发活动和积极的生态建设措施是1991～1996a景观格局变化的主要驱动力；1996～2001a景观变化的主要驱动力是城市化进程加速、产业比例结构的调整和林业生态建设措施等因素,在这一时期内,景观变化剧烈程度明显增加；社会经济发展、城市化进程加速和具有激励机制的林业生态政策是2001～2007a景观变化的主要驱动因素,在这一时期内,景观变化更为剧烈,城乡建设用地和森林两个类型的面积大幅增加。为改变武汉市景观破碎化加剧的状况,景观管理政策的制定必须综合考虑到变化条件下的多方利益,由保护单个的生态资源向保护生态资源系统和促进区域可持续发展转变。(2)基于土地覆盖数据,运用费用距离模型对1987-2007a武汉市景观连接度长期变化特征进行了定量评估。研究结果表明,随着城市化进程的不断加快,武汉市森林和湿地景观连接度迅速降低,高景观连接度景观的面积大幅减小,而低连接度景观的面积迅速增加,这揭示了武汉市生态资源景观连接度已经受到城市景观变化的严重影响。森林景观格局与景观连接度变化趋势不对称揭示出武汉市城市化进程对森林物种的影响已远大于森林建设带来的生态过程改善能力。因此,建议今后森林建设工程必须转变思路,从重视空间结构指标向注重功能连续性方面转变。水体景观连接度的变化与水体景观格局的总体变化趋势较为一致,随着水体面积减小和结构破碎化程度加剧,水体景观连接度持续下降。除水体面积和结构破坏之外,人类干扰活动对水体景观连接度的影响也非常明显。因此,对水体的保护和管理不能局限于保护水体面积和斑块结构不受损,而应充分考虑到水体边界景观对比度变化对水体景观连接度的影响。(3)基于地表温度反演的结果,将武汉市热环境按照地表温度值大小分为6类,其空间分异特征分析表明,武汉市热环境空间分布不均衡,呈明显的梯度分布特征,城市中心地区的温度明显高于城市外围地区。其中,高温景观和低温景观呈现组团式分布特征,高温景观主要分布在城市化较早,人口和建筑密集的中环线以内的城市建设区和工业区；城市低温区域主要分布在域内的大型湿地和森林地区,城市外围地区的低温景观面积明显高于城市建筑密集区。热岛斑块面积较小,且斑块间面积差异不大；最低温区和较低温区的平均斑块面积非常大。人类活动强度在增加地表温度的同时,也导致高温热力场分布向细粒景观转化。低温类型景观和高温类型景观分布较离散,而中等温度的景观类型斑块间距离较小。地表覆盖特征显著影响地表温度,城市建设用地均温最高,水体均温最低；高温区建设用地所占面积最大,低温区中水体所占面积最大；水体、裸地和城市建设用地中地表温度变化较大；从相对降温率来看,水体降温效率最高,滩涂湿地与森林也具有较高的降温效率,而城市建设用地会导致地表温度升高。植被覆盖度与地表温度之间存在极其显著的关系,其中水面覆盖区域植被覆盖度的增加与地表温度增加呈极其显著的正相关关系,植被覆盖度增加的温度降低作用规律依次为：裸地>建设用地>森林>草地>滩涂湿地>农田。此外,水体、滩涂湿地和森林对周边地表温度具有重要影响,多数情况下存在明显的距离效应和面积效应。降温景观对城市建设用地降温作用距离要高于其他景观类型上的作用距离,对植被覆盖区域(如农田、森林和草地等)降温距离略小。小型滩涂湿地斑块和小型森林斑块降温距离具有一定优势,因此,在城市中建立多个小型森林斑块比建立单一的大型森林斑块降温效果要好。研究同时表明,斑块面积对周围景观地表温度有一定影响,但影响能力普遍较弱。因此,为减缓武汉市城市热岛效应,需改善城市绿地质量,降低建设用地和裸地类型的斑块面积,建立城市湿地—森林一体化降温网络和改善城市风道。(4)针对城市区域景观破碎化加剧的趋势,本文提出构建城市绿道系统应对持续增加的城市绿地景观破碎化。通过集成最小费用路径模型、核密度估计、优先性原则和代理指标方法,本文构建了城市化地区多功能绿道构建方法框架。这一方法模型主要由识别规划目标与需求,决定潜在生态保育绿道、识别潜在游憩绿道、识别潜在水体保护绿道、设计综合绿道网络5部分内容构成。而对于单目标廊道识别则包括识别源汇斑块、设置费用权重、识别潜在连接和优化网络结构等4步骤。本研究通过将建设成本纳入最小费用路径分析改进了最小费用路径分析的现实性,使其模拟结果能够综合反映使用者和建设维护者需求,有助于促进城市绿道可持续性。运用这一方法,本研究提出了一个多目标优先性绿道网络体系。这一绿道网络体系由理想鸟类保育网络、理想小型哺乳动物保育网络、理想人类游憩网络、水体保护网络、骨干鸟类绿道网络、骨干小型哺乳动物保育网络、骨干人类游憩网络和综合性绿道网络等8个绿道网络构成,分别对应着低生态安全水平、中度生态安全水平和高生态安全水平等3个生态安全等级。在武汉市绿道建设中应该按照其生态安全等级进行分批建设分级管理。并针对不同生态安全水平上的绿道网络识别了其中重要绿道分布区和关键绿道。今后的研究应提高绿道规划的空间连续性、功能完整性和数据可靠性。(5)借鉴生态网络规划中景观连接度的理论,提出了构建城市“森林—湿地”降温网络改善城市热岛效应的思路和规划框架,并对武汉市城市森林一湿地降温网络构建进行了研究。这一规划框架运用源汇模型、最小费用路径模型和核密度分析模型识别了潜在的降温网络。城市“森林—湿地”降温网络由热梯度廊道、隔离廊道和低温廊道、绿色节点和降温斑块相互连接构成。武汉市城市“森林—湿地”降温网络研究共识别了6个不同层次和目的的降温网络,根据其网络结构的完整性可对应三个层次的环境安全层次：高环境安全等级水平、中度环境安全水平和低环境安全水平。而且,考虑到建设的可行性,综合性网络将冷热空气流动的需求整合,反映了降低城市热岛效应的最低要求,也是迫切需要付诸实施的廊道网络结构,它对应着最低的环境安全水平；骨干网络从不同降温作用方式更加完善的定义了区域降温网络的需求,对应着中度环境安全水平和中度优先性；而理想生态网络反映了最完善的降温需求,因此对应着最高环境安全水平和最低优先性。城市降温网络建设应以改善城市热岛区植被覆盖状况,缓解本身热量累积和构建城市植被降温廊道为主。通过建设一定结构特征的植被廊道将森林、湿地等降温斑块连接发挥其整体降温作用。(6)构建结构功能完整的生态网络对于快速发展地区森林保育具有重要意义。为改善武汉市森林景观连接度不断降低的趋势,本文以武汉市为研究对象,提出并应用了一个面向快速城市化环境的综合性的森林网络规划框架。研究结果显示通过整合物种包方法、多标准评价、最小费用路径模型、核密度估计和优先性原则,这一规划模型能够较好的适用于识别和规划生态网络。它使规划者能够将多物种需求整合到综合性生态网络中。本文识别了7个生态网络,对应着理想网络、骨干网络和综合性网络等三个保育优先性水平。这一综合性网络系统为规划者和利益相关者提供了一个深入和直观的视角来理解生态过程和提升区域生态系统可持续性。管理者应该尽快将综合性森林网络付诸实施,以保护当地物种多样性,应对城市化带来的森林生境破碎化问题。我们希望这一方法能够改善区域生态网络规划实践。(7)运用最小费用距离模型和焦点物种代理法识别了武汉市潜在的湖泊湿地生态网络,在此基础上运用中心度方法重点评价了武汉市湖泊湿地网络要素的重要性。研究结果显示,虽然武汉市湿地资源众多,但是能够作为湿地鸟类生境的湿地仅占武汉市湿地资源总量的21.53%,一些重要的生境斑块和廊道分布在中环线周围地区,这些地区正是今后城市扩张的主要位置。根据湿地核心生境斑块的节点度和中介度评估,本文将武汉市湖泊湿地生境重要性进行了排序,并分为最优先、中度优先和普通三个优先性类型。武汉市潜在湖泊湿地生态网络涉及到的56条廊道可分为3个优先层次。高优先性的廊道主要分布在网络内部,连接了多数高优先性生境斑块,构成了武汉市湿地生态网络的基本框架。更多还原

【Abstract】 Rapid urbanization has caused many eco-environmental problems when it improves our living condition, and put a serious threat to regional sustainability. Hence, how to mitigate these negative effects linking to rapid urbanization has become an urgent task and considerable challenge to maintain social, economic and environmental sustainability. The optimization of regional landscape structure, construction of the ecological security pattern as well as conservation and restoration the necessary regional ecological processes have become an increasing consensus on coping urbanization environmental problems. Yet, the ecological security pattern model is still a conceptual strategy in landscape planning due to lack of practical and feasible spatial planning approaches. Wuhan, as the largest city in the central of China, is undergoing the unprecedented urbanization process with dramatically changing in landscape pattern, ecological processes and environmental condition. Therefore, the researches on the approach to improve our urban landscape configuration will be of important theoretical and realistic significance to combat these eco-environmental problems. In this paper, the historical remote sensing imageries, climate data, basic topographical data, and field investigation data were collected to study the optimum ecological patterns of Wuhan. Hence, the objectives of our study are to focus on the landscape configuration dynamics and its driver-forces, landscape connectivity trends of forests and wetlands, and spatial characteristics of urban heat island (UHI), identify major eco-environmental problems in Wuhan, and develop a set of optimum ecological patterns proposal to eliminate these eco-environmental threatens, and initial to present an feasible landscape approach for urban eco-security pattern planning. The main results are shown as follows:(1) The landscape configuration has significantly changed during1987-2007a. Farmlands and water areas decrease dramatically in their area while built-up areas and forest quickly increase in their size. Such characteristics present a clear trend that the whole landscape configuration was changing from agricultural pattern to urban pattern. Increasing fragmentations were found in forest and water. The Aggregated Index of urban built-up areas increased during the whole research period. Because increasing human activities, the landscape in Wuhan is changing from coarse grain to fine grain, difference of patch area gradually reduced. Urban sprawl, infrastructure development and other human activities have significantly affected the pattern of regional landscape, resulting in continued increase in landscape heterogeneity. Since1991a, the landscape gradually increases the overall degree of dynamic change. The area of various landscape types fluctuated throughout the study period. Grasslands, built-up lands and bare lands have high annual rates of landscape changing but there are some differences in various stages.Agricultural development, urbanization, forest development, and the large precipitation are the main direct drivers of landscape change in1987-1991a. Increasing urban population, agricultural development, and positive eco-construction measures is the main driving force of landscape pattern changes during1991-1996a; The main factors affecting landscape change in1996-2001a, is the accelerated urbanization, the increasing proportion of industrial in GDP, and forestry ecological construction measures. In this period, the landscape changes significantly increased in intensity. Socio-economic development, accelerated urbanization process, and forestry policy with incentives is the main driver of landscape change in the period of2001-2007a. In this period, the landscape changes more intense, urban and rural construction land and forest increased significantly.To alleviate the exacerbated situation of landscape fragmentation in Wuhan, the deciders must consider the benefits of changing stockholder and integrate them into the landscape management policies, and change protection strategy of ecological resources from single patches to whole system to promote regional sustainable development.(2) Based on the generated land-cover data, a cost-distance model was used to estimate and assessed the landscape connectivity trends of Wuhan during the twenty years of1987-2007a. The results shown that, the landscape connectivity among forest patches and that among water patches decreased dramatically in the process of rapid urbanization of Wuhan. Additionally, the landscape areas with high connectivity decreased clearly while the landscape areas with low connectivity increased quickly. This trend indicates that the rapid landscape changes have influenced connectivity of eco-resources. The trend difference in pattern and connectivity of forest indicates that the negative effects on forest connectivity resulted from rapid urbanization have offset the positive effects of afforesting. Therefore, future forest construction projects must change their thinking from a focus on indicators of spatial structure to focus on changes in functional continuity. Trends of water connectivity is similar to the trends of landscape pattern, with the water area decreasing and fragmentation degree increasing, landscape connectivity degree of water continued to decline. In addition, the increasing human activities also influence water connectivity. Therefore, the protection and management of water bodies should not only focus on patch area and patches structure, but also fully take changes in water contrast landscape border into account.(3) The urban land surface temperature (LST) of Wuhan was retrieved using Landsat TM5remote sening imagery and classified into six types according to their temperature value. Based on the LST data, the thermal pattern and thermal environment effect of urban landscape pattern were analyzed. The results shown that that the spatial distribution of thermal environment in Wuhan was uneven, and showed a clear urban-rural gradient distribution. The temperature of downtown significantly higher than outlying areas of the city, UHI areas concentrated on urban central area while the low-temperature landscape mainly appear at outlying areas of the city with significant group distribution characteristics. The patch size of UHI is very small while the patch size of the lowest-and lower-temperature patches is much larger. The Euclidean Nearest-Neighbor Distance (ENN) of the lowest-and highest-temperature landscape is larger than the ENN of the medium-temperature patches.The results also demonstrated that land-cover characteristics significantly affected LST of research area. The average LST of built-up areas is the highest, while the average LST of water is the lowest. The built-up areas share the largest area of highest-temperature zones while water occupied the largest area of the lowest-LST zone Water, beach wetlands and forests influence significantly LST of the neighbor landscape. In most case, there is a clear distance effect and size effect to decrease LST of their neighbor landscape. The effective cooling-distance on built-up area by low-temperature landscape is largest among various land-cover types while the effective distance on vegetation covered areas (i.e. farmlands, forest and grassland).is smaller than others.Additionally, small patches of beach wetlands and forest has relative higher efficiency to decrease surrounding LST. Hence, the cooling effect by several small patches is better than one single large patch. Patch size of cooling landscape could influence the cooling-effects to surrounding landscape in a certain extent, but their impacts are very weak. In order to mitigate urban heat island, several strategies, to improve vegetation coverage of the green space and other landcover types, reduce patch size of built-up areas and bare lands, and establish urban wetlands-forest network and improve urban air duct are proposed(4) Urban greenways are proven strategies to counter ecological fragmentation and improve urban sustainability. However, significant challenges remain when seeking to integrate multiple functions into priority-oriented planning. In this paper, we focus on the city of Wuhan (China) to present an innovative approach for developing multi-functional greenway networks. Based on remote sensing and geographical information systems, this approach applies a least-cost path model, a kernel density analysis and a proxy index to identify and integrate multiple needs into priority greenway networks. Our priorities are bird and small mammal conservation, human recreation and water protection. In the cost settings of the least-cost path model, we include simulated construction costs to realistically identify corridors. This method insures that our results reflect the expenses necessary for the construction of greenways. Our study demonstrates that the kernel density method is an effective tool to illustrate priorities by revealing the potential utilization intensity of corridors. Based on the proposed method, we identify eight potential greenway networks, each highlighting different sets of priorities. We divide these priority greenway networks into three categories of ecological security:ideal greenways, backbone greenways and the comprehensive greenway network. The latter represents the integration of animal conservation, human recreation and water protection objectives. Because priorities can be added, interchanged or weighted according to local specificities, our study offers a methodological framework applicable to developing and developed cities.(5) In order to mitigate UHI, this paper presented a framework for Forest-Wetlands Cooling network (FWC network) planning. The proposed framework was demonstrated by a case of Wuhan. Using the landscape connectivity theory, the presented approach combined source/sink theory, least-cost path model and kernel density estimate to identify the potential cooling network. Urban FWC network is composited of thermal gradient corridor, barrier corridor, low-temperature corridor, green hub and cooling patch.In this paper, the six networks, responding various functions were identified and classified into three environmental-security level of high, medium and low level. The environmental security levels reflect the priorities for their constructions. The comprehensive FWC network represents the lowest needs for eliminate UHI effects and responses the highest priority for construction. The ideal FWC network represents the optimum network to mitigate UHI, and therefore response to the highest environmental security level and lowest priority. Considering the construction cost, main issues of establishing forest-wetlands cooling network are to improve cooling rate of built-up by increase vegetation coverage and to develop cooling corridors.(6) Construction, preservation and connections between forested areas are proven strategies to counter human-induced fragmentation of natural ecosystems in urban areas and improve regional sustainability as a whole. Yet designing forest networks that target multiple species in urban areas remains a challenging task for planners. In this section, we present a modeling framework for multi-species forest network planning in rapidly urbanizing contexts using a case study of Wuhan, a city in Central China. Based on remote sensing and geographical information systems, our approach integrates the species assemblage method, multi-criterion assessment, least-cost path model, kernel density estimates, and priority principle to plan forest networks for regional biodiversity conservation. Focal species assemblages were developed to represent local species with high sensitivity to fragmentation. We include stepping-stone functions in the cost settings of the least-cost path model to realistically identify optimum network for different species assemblages. Our study demonstrates that the kernel density method is an effective tool to illustrate priorities as it reveals potential utilization intensity of corridors. Using the proposed method, we identified seven forest networks accounting for three priority levels of biodiversity conservation within Wuhan:the ideal network, backbone network and comprehensive network. The proposed approach enables planners to readily integrate multi-species demands and develop comprehensive forest networks. This approach may contribute to forest networks planning in rapidly changing areas, especially urban.(7) Assessing functional significance of landscape elements is considered as the important context of conservation planning and the basis of sustainable ecological network. Based on conservation needs of focal species, a potential wetlands network was identified. The centrality method was further employed to assess the functional significance of wetland ecological network in Wuhan. Our results indicated that:the potential wetland network was consisted of29core-habitat patches and56corridors with least travel costs. According to the degree and betweenness, these core-habitat patches and corridors can be categorized into three priority levels, including the highest priority, the medium priority and low priority groups. When the proposed network is implemented, deferent scheduling priorities should be given respectably to various significant levels. Moreover, our simulated results also illustrated the effectiveness of the centrality method. Incorporating focal-assemblage surrogate model, multi-criterion suitability model and least-cost path model, the centrality method is available to rank and identify the key elements in regional ecological network. The future researches are expected to focus on the scale-effects of ecological network.更多还原