【作者】 甘敬；

【导师】 马履一； 余新晓；

【作者基本信息】 北京林业大学 ， 森林培育， 2008， 博士

【摘要】 本文以系统科学、景观生态学等理论为依据,以样地调查、北京市“十五”森林资源二类调查结果和其它途径收集的图件为基础数据,从探讨森林健康的内涵与实质入手,在对森林景观进行区划的基础上,确立评价指标,采用正态分析及等距分组法确定评价指标标准,采用层次分析法确定评价指标权重,采用多级模糊综合评价法确定健康等级隶属度,并将北京山区森林健康评价等级划分为优质、健康、亚健康、不健康和疾病五级,从森林经营小班、景观和区域三个尺度对北京山区森林健康状况进行了评价,旨在为北京市森林健康可持续经营提供理论依据。主要研究结果与结论如下:(1)运用森林资源二类调查与遥感影像相结合的方法,以地类、森林类型、起源和优势树种作为北京山区森林景观区划的主要因子,区划得到北京山区共18个森林景观类型。森林景观分类精度达到83%以上。(2)在森林经营小班尺度上,以现代系统科学理论为基础,从完整性、稳定性和可持续性三个方面初步构建健康评价指标14个,经筛选,得到小班健康评价指标8个;在景观尺度上,以景观生态学理论为基础,从结构与格局、功能与过程两个方面初步构建健康评价指标6个,经筛选,得到景观健康评价指标4个。(3)在森林经营小班尺度上,建立了森林景观类型林分蓄积量—生物量回归模型、灌木层生物量—灌木高度回归模型、草本层生物量—草本高度回归模型各18类,以推求单位面积总生物量;建立了树高、胸径—叶面积指数回归模型18类,以推求小班叶面积指数。在景观尺度上,采用景观格局计算公式得到了分形维数、景观多样性、景观优势度、景观均匀度各18个指标值;通过统计各类景观样地每木调查数据的病虫害等级数量比例,计算其模糊隶属度的相对位置而得到了病虫害程度18个指数。(4)在小班尺度上,将小班评价指标分类为定性描述指标、无限制可比定量指标和半限制可比定量指标三种。确立了无限制可比定量指标、半限制可比定量指标标准各5级,在景观尺度上,确立分形维数、景观多样性指数指标标准各5级。(5)采用层次分析法确定指标权重,得到:小班健康评价指标权重向量为A_p=(0.0211 0.0646 0.2476 0.0333 0.3000 0.0861 0.2123 0.0349);景观健康评价指标权重向量为A_L=(0.1250 0.3750 0.4375 0.0625);北京山区森林整体健康评价指标权重向量为A_T=(0.0719 0.0602 0.0632 0.0618 0.0495 0.0670 0.0414 0.0476 0.0640 0.05230.0493 0.0496 0.0536 0.0477 0.0451 0.0668 0.0677 0.0411)。(6)评价结果表明:在小班尺度上,北京山区30164个森林经营小班中,处于亚健康的最多,达11265个,面积比例为42.21%,属于优质和疾病的小班较少,面积比分别占10%左右。在景观尺度上,北京山区18类森林景观中属于健康的4类,亚健康的12类,不健康的2类,无优质和疾病的景观类型。在区域尺度上,北京山区整体的健康状况为亚健康。以上结论说明北京市急需进行森林健康可持续经营,提高森林健康等级。更多还原

【Abstract】 Based on theories of systemic science and landscape ecology and in terms of data consisting of standard plots investigation, forest resources inventory and collected charts else, evaluation index system of forest health is built by exploring the conception and essence of forest health. A forest health evaluation study for Beijing mountain area is conducted, based on forest landscape zoning in this research. The evaluation indices of forest health is assigned by adopting the method of normal regression analysis and the way of equidistant dividing, the index weight is determined by applying the method of AHP, and membership degree is determined by the method of multilevel fuzzy comprehensive evaluation. Five forest health degrees, namely, Robust (Perfect), Healthy, Sub-healthy, Unhealthy and Disease are assigned, forest health status of Mountainous Beijing is evaluated at compartment, landscape and Region level respectively, in order to provide theoretical base for Beijing’s sustainable forest management. Main outputs of the paper conclude:(1) Eighteen forest landscape divisions in total are found in Beijing from this study, by using land tape, forest type, forest origin and dominant tree species as the main indices to analyze forest in Beijing mountainous area. Based on the forest management inventory data and remote sensing image classification, the classification accuracy is above 83%.(2) On the scale of sub-compartment, based on the theory of modern system science, this paper conducts 14 preliminary forest health indices by considering integrality, stability and sustainability of forest ecosystem, 8 forest health indices are selected after screening. On the scale of forest landscape division, based on the theory of landscape ecology, 6 preliminary forest health indices are conducted by considering the relationship of ecosystem structure and distribution pattern and relationship between ecosystem function and process, 4 forest health indices are selected after screening.(3) Regression analysis models are established to measure the indices that cannot obtain directly. On the scale of sub-compartment, to calculate the total biomass of each unit area, 18 volume-biomass regression models of stand layer, 18 height-biomass regression models of shrub layer and 18 height-biomass regression models of grass layer are established; Eighteen tree height, DBH and leaf area index models are established, to calculate leaf area index,.On the scale of forest landscape division, 18 indices are calculated, such as fractal dimension, landscape diversity, landscape dominance, evenness; based on the investigation data of forest pests and disaster grades collected from each landscape sample plot, 18 indices of forest pest and disease are calculated by calculating the relative position of fuzzy membership degree.(4). On the scale of sub-compartment, the healthy indices are categorized into 3 classes, namely, qualitative index, unrestricted comparable quantitative index, and semi-restricted comparable quantitative index. The standards of each unrestricted comparable quantitative index and semi-restricted comparable quantitative index are divided into 5 classes respectively. On the scale of forest landscape division, the standard of fractal dimension and landscape diversity is divided into 5 classes respectively.(5) Applying the method of AHP to confirm the index weight, index weight vectors are calculated as follow:The forest health index weight vector of sub-compartment scale is:A_P =(0-0211 0.0646 0.2476 0.0333 0.3000 0.0861 0.2123 0.0349);The index weight vector of landscape scale is:A_L =(0.1250 0.3750 0.4375 0.0625);The index weight vector of total Beijing mountain area forest health is:A_T =(0.0719 0.0602 0.0632 0.0618 0.0495 0.0670 0.0414 0.0476 0.0640 0.0523 0.0493 0.0496 0.0536 0.0477 0.0451 0.0668 0.0677 0.0411),(6) The evaluation results show that, in the scale of sub-compartment, 11265 sub-compartments are classified as in sub-healthy category, which accounts for 42.21% of the total Beijing Mountainous area, while sub-compartments are classified as in either robust or disease occupy 10% of the total area respectively. On the scale of forest landscape, there are 4 kinds that are classified as robust, 12 kinds as sub-healthy and 2 kinds as unhealthy among the total 18 forest landscape types. Neither high-quality nor diseased forest landscape types is found in Beijing mountainous area in this study. On the scale of whole Beijing mountain region, the healthy condition of Beijing mountainous forest is classified as sub health.The above conclusions shows that it is both urgent and necessary carry on forest health management, so as to improve forest health grade in Mountainous Beijing.更多还原