【作者】 廖晓丽；

【导师】 周新年； 刘健；

【作者基本信息】 福建农林大学 ， 森林经理学， 2012， 博士

【摘要】 沿海防护林是滨海的林业生态工程，由于所处地理位置的特殊性，在滨海地区的防风、固沙和防治水土流失的保护作用中起着关键的作用。而它同时也是一个脆弱的生态系统和全球气候变化的敏感区域，再加上沿海防护林早期建设时造林集中连片、造林模式单一、树种组成单一的模式，导致沿海防护林体系结构缺失，防护功能不健全，自然灾害发生频繁，自然因素和人为干扰的双重影响正在挑战沿海防护林的内部稳定性和防护功能，因此，如何对沿海防护林进行结构模式、树种组成的优化配置，实现沿海防护林的合理布局，以发挥生态效益、经济效益和社会效益的最大化是当前沿海防护林研究中亟待解决的首要问题。本文以沿海防护林的优化配置为目标，以2003年研究区域的遥感影像图、森林资源的一类、二类调查资料为基础，通过数据分析选择因子分析的方法，以固沙效益为着眼点，构建沿海防护林固沙效益模型，分析以固沙效益为目的的最佳植被盖度；以近自然林业为理论基础，确定最优的植被类型；通过分析沿海防护林主要优势树种的适生性，指导树种的选择和对位配置；并以以上研究的结果为沿海防护林优化配置的标准，对固沙效益效果不理想的区域进行小班调整，使其接近理想模式。主要研究结果如下：以植被指数（NDVI、VARI和MSAVI）、坡度、坡向、海拔为自变量，采用逐步回归的方法，利用多个自变量的最优组合建立回归方程，构建植被盖度的线性模型为Y=-0.101+1.064NDVI+0.129h，经验证，植被盖度估测值的平均精度达到83.77%；以裸沙占地面积比、气候侵蚀力因子、裸土指数、MSAVI、国内生产总值（GDP）5个因子作为固沙效益的评价因子，2004年沙化监测结果图的沙化等级为因变量，采用因子分析的方法，确定裸沙占地面积比、气候侵蚀力因子、裸土指数、MSAVI、国内生产总值的权重分别为：1.0184、0.3973、1.1823、0.7414、0.3592，构建固沙效益评价模型；采用2次多项式对植被盖度和固沙效益指数进行趋势线拟合，拟合的方程为：D=0.1388C~2-0.2441C+0.34，确定以固沙效益为目标的最佳植被盖度为0.8793；以近自然林业理论为基础，采用人工神经网络的方法对最优植被类型进行预测，配置后竹林、阔叶林、针叶林、混交林、灌木林的比例为4.15%、22.52%、17.92%、53.94%、1.46%。并用景观格局分析和生态质量评价配置的结果；以坡度、坡向、海拔、土层厚度和土壤类型为因子，以改算树高为因变量，采用数量化理论对沿海防护林主要优势树种木麻黄、湿地松、相思树、马尾松、杉木进行适生性分析，并划分适生等级，竹林通过耐寒性分析，划分竹林的适生性等级，为树种选择提供参考。通过分析固沙效益的效果，以最佳植被盖度、最优植被类型和优势树种的适生性为理想模型，对固沙效果不理想的区域进行以小班为单位的调整，使调整的结果尽量接近理想模式。更多还原

【Abstract】 The coastal protection forest is belong to coastal forestry ecologicalengineering, it has being played the key role on windbreak and sandfixation, erosion prevention in coastal areas due to its particularity ofgeographical position. On the other hand, it is also a fragile ecosystem andsensitive areas of global climate change, coupled with the earlyconstruction mode of centralized planting, single afforestation, and fewforest tree species, which lead to the lack of structure, distemperedness inprotection function, and frequency in natural disasters, The double impactfrom nature and human is threatening the internal stability and protectionfunction of coastal protection forest. Thus it becomes the primary problemto maximize ecological, economic and social benefits, needed to be solvedurgently in current coastal protection forest study, which is about to findinga way that is help to realize the Optimization of coastal protection foreststructure model and optimal allocation in tree species. This paper is aimedat optimizing the coastal protection forest distribution. In this paper, thecoastal protection forest sand fixation model,which focus on sand fixationeffectiveness and can be used to evaluate the proportion of VegetationCoverage, which is built by analyzing dates with factor analyse methods,the dates are based on the remote-sensor image in2003in study areas andthe result of forest management inventory; the optimal type of vegetationare determined in the guidance of Close to Nature Forestry Theory; theforest tree species and their locations are determined by the suitability ofdominant species;as a result of the optimization, subcompartments,whichdo not play an effective role in sand fixation, are adjusted to make themclose to the ideal model. The main results are as follows: 1. Having vegetation index (NDVI of VARI and MSAVI), slope, aspect,altitude as the independent variables, using stepwise regression method,the optimal combination of multiple variable regression equation wasestablished which is used to build the linear model of the vegetationcoverage. the model is Y=-0.101+1.064NDVI+0.129h, which is proved thatthe average accuracy of estimating the value of vegetation coverage canreach to83.77%;2. Having the rate of bare sand, climate erosivity factor, the bare soilindex,MSAVI, gross domestic product (GDP) as the evaluation factorsand desertification grade in desertification monitoring results graph in2004as the dependent variable, with factor analysis methods todetermine the weight of each factor, the area ratio of bare sand, ClimaticErosivity Factor, Bare soil index, MSAVI, the gross domestic product,whose weights are:1.0184,0.3973,1.1823,0.7414,0.3592, sandfixation efficiency evaluation model is built;3. AS a result of the trend line fitting of vegetation coverage and sandfixation efficiency index, using the two order polynomial, building thefitting equation: D=0.1388C2-0.2441C+0.34, and the best vegetationcoverage is determined to be0.8793;4. Based on the theory of Close to Nature Forestry, with the artificialneural network approach to predict the optimal configuration ofvegetation types, the optimal proportion of bamboo, broadleaf forest,coniferous forest, mixed forest, shrub forest is determined respectivelyto be4.15%,22.52%,17.92%,53.94%,1.46%. At the same time, the result ismeasured by landscape pattern and ecological quality;5. Having slope, aspect, elevation, soil depth and soil type as the factorsand calculated tree height as the dependent variable, with mathematicalmethods, suitability of the main dominant coastal protection speciesephedra, wetlands, pine, acacia, masson pine,Chinese fir is analyzed and their suitable grade is divided, and through the analysis of cold toleranceof bamboo, bamboo adaptability grade is dvided, which provide areference for the selection of tree species;6. According to the ideal suitability model of the optimal vegetationcoverage, the optimal vegetation types and the dominant species, thesubcompartments, which do not play an effective role in sand fixation,are adjusted to make them close to the ideal model, by analyzing thesand fixation effect.更多还原