长白山林区林场级尺度景观格局演化与模拟

发布时间：2018-07-02 20:15

本文选题：景观格局 + 景观尺度　；参考：《东北林业大学》2016年博士论文

【摘要】：长白山林区是我国过伐林分布较多的地区之一,过去由于粗放经营,森林结构被破坏、功能逐渐退化。如何保护、经营天然林资源和恢复良好的林分结构己成为我国林业发展所迫切需要解决的问题。基于此,本文基于景观生态学基础理论和近自然林经营理念,选取典型区域吉林汪清林业局塔子沟林场(吉林省首次实施抚育间伐作业试点)为研究对象。借助于RS和GIS技术,以研究区1997年和2009年两期Landsat TM影像为主要数据源,辅助大量非遥感信息,基于景观格局尺度特征、景观格局尺度效应、景观格局空间结构特征、均匀度理论和景观格局的地形分布特征等,定量分析并揭示该区域12年间在景观尺度上的格局动态变化特征、时空演变规律、多尺度动态演变及其空间自相关特征。在此基础上,定量分析地形因素对森林植被空间格局影响状况,基于生物量地形因子趋势面定量分析,以其揭示海拔、坡度和坡向等各地形因子及其组合状况对其空间分布影响。基于CA-Markov模型和终极状态方程,模拟预测未来景观格局的发展趋势及其终极状态格局,基于森林景观自然度评价基础上,完成不同景观自然度森林经营模式研究和生态设计,从类型层次和景观尺度提出天然林景观合理配置模式,基于(?)原则和均匀度理论,完成区域森林经营抚育间伐设计和空间格局控制,在形成示范区的基础上,从景观水平上研究森林景观格局分布状态演变和解决森林生态经营规划技术和方法,以其为该区域森林景观生态过程模拟模型与森林经营决策模型的研究提供综合的、系统的方法和思路。本文主要取得以下的结论：(1)基于区域一级景观类型分析可知,总体景观格局以有林地景观为主,荒山荒地、沼泽地和居民用地等景观斑块分散于有林地中,各景观类型面积分布不均衡,斑块数目差异较为明显。森林景观在研究区域中占据绝对的优势地位,两期的面积比例分别为93.47%和95.39%,控制着研究区域的景观整体结构、功能和动态过程；人为景观和半自然景观明显处于支配地位。研究区森林类型海拔梯度分布为从河谷地低海拔榆树林和杨桦林到高海拔红松云冷杉林,地带性森林植被--阔叶红松林所形成的森林生态系统类型,主导着本地区森林景观格局的动态演变。(2)基于两期二级景观森林类型分析可知：研究区域是以阔叶混交林为主体,两期的面积比例分别为74.94%和82.45%,均以阔叶混交林景观、蒙古栎林景观和针阔混交林景观构成研究区域景观基质。12年间,该区域优势种演替顺序为先锋种(榆树、杨桦、落叶松)ì伴生种(水胡黄椴)→顶级种(云冷红-色柞),处于演替序列低级阶段、高级阶段和演替后期的杨桦林、阔叶红松林和针阔混交林面积减少,处于演替中期的阔叶混交林景观的面积增加,次生林区特征较为明显,景观异质性高,近自然程度较低,景观空间格局现状易受某些环境因子和干扰状况的综合控制。(3)1997-2009年,研究区域的聚集度和蔓延度指数增加,分别由88.4128和97.7300增加到91.7402和98.3815,分离度指数由1.144减少到1.1001,表明区域景观破碎化程度减缓,景观类型的分布由分散趋向集中交错,连通性增强,利于景观功能的整体发挥,但景观的优势度增加,景观的空间异质性和多样性减少,且景观优势度及多样性两个指数呈现出一定负相关,表明该区域各景观类型所占比例差异增大。(4)1997年的Moran's I系数为-0.11,2009年为0.42,基于General G-statistic的计算可知,1997年的Z Score的数值为0.92,而2009年为2.06；1997年基于均匀度理论计算的均匀度指数为0.3207,2009年为0.2883。基于Moran's I和均匀地理论结果相同,两结果都表明12年间研究区域的景观格局呈现出由随机格局到集聚格局的发展趋势,天然次生林的群落结构呈现出相对不稳定且2009年呈现出全局空间自相关在高值空间聚集(H-H聚集),但整体景观的优势度增加,单一优势景观类型对区域整体景观的控制作用增强。(5)基于ArcGIS 9.3的Geostatistical Analyst的分析可知,1997年SHDI最大和最小值分别为2.0463和0.2544,2009年分别为1.9722和0.2418,且SHDI的优势区域集中分布于该研究区域的中部地带,呈现西南向东北走向的条带状,并基于地统计学软件GS+计算实验区半方差函数和理论拟合模型进行分析,结果表明该尺度下森林景观多样性与空间地理位置具有中等的空间相关性和明显地域结构性,基于IDW和Kring插值获取不同尺度下森林景观多样性指数的空间分布图。(6)在ArcGIS的支持下,基于DEM与优势树种的空间叠加,精确地描述各景观要素的优势树种在空间格局上的分布、大小、地理位置或相对位置,并完成基于生物量地形因子趋势面的定量分析。高海拔和低海拔区域,两期的生物量都随着坡向的改变而呈现一定规律的变化,在高海拔、坡度大的区域生物量较大；坡度主要通过联合坡向对改善立地水分条件的效应发挥出对景观要素的生物量空间分异进行作用。局部坡度因子对生物量的影响没有成为整个森林生物量空间分布趋势的重要因子,仅影响到局部区域生物量波动起伏。(7)基于CA-Markov的终极状态概率计算可知,按照当前经营措施不发生改变的情况下,研究区域最终的状态是：阔叶混交林、针阔混交林和榆树林的面积增加,其他景观类型的面积都呈现减小,阔叶混交林的面积比例达到了59%,在研究区域中占据绝对的优势地位,顶级森林景观的云冷杉针叶混交林和阔叶红松林的面积比例仅为3.9%,预测最终的景观格局现状将不利于整个研究区域景观的整体稳定性发展,也不符合其森林景观经营目标。(8)基于群落连续带指数(CI)来定量确定群落的演替状态,以近自然林业理论和潜在天然植被生态修理念为指导,通过对森林景观生境、森林景观的演替情况、优势树种、立地条件森林景观的经营目的定量研究,首次提出景观自然度的概念,并划分为顶级景观自然度、高级景观自然度、中间过渡性景观自然度和初级景观自然度,基于森林景观自然度评价的基础上,以林场尺度为示范区,首次以景观自然度区域内的森林景观类型为经营管理单元,从类型层次和景观尺度提出经营目标、目标结构以及结构调整等主要内容的经营模式,为长白山林区森林经营提供理论依据和技术支撑。(9)在ArcGIS 9.3的支持下,利用VBA程序提取2009年三个大小为30m×30m演替初期、演替中期、演替亚顶级各阶段样地树种点相对坐标,并建立GIS专题数据库,基于C#和均匀度理论的支持下,完成样地树点最近点距离、独占体总面积及均匀度指数的定量计算,结果表明三个样地均匀度分别为0.331、0.288、0.275；按照(?)原则,选择性的去掉一些点后,重新计算三个样地的均匀度分别为0.463、0.346、0.290,得出按照(?)原则去点后的分布来生成树木,可以使林分格局达到优化。
[Abstract]:Changbai Mountain forest area is one of the areas with large distribution of over logged forests in China. In the past, the forest structure was destroyed and the function gradually deteriorated because of extensive management. How to protect, manage natural forest resources and restore good forest structure has become an urgent problem to be solved in the development of forestry in China. Based on this, this paper is based on the basic theory of landscape ecology. And the concept of near natural forest management, selected typical area Jilin Wangqing Forestry Bureau tower Zi Gou Forest Farm (Jilin province for the first time the practice of thinning operation pilot) as the research object. With the help of RS and GIS technology, the study area in 1997 and 2009 Landsat TM images as the main data source, auxiliary a large number of non remote sensing information, based on the landscape pattern scale characteristics, The scale effect of landscape pattern, the spatial structure feature of landscape pattern, the theory of evenness and the topography of landscape pattern, and so on. The quantitative analysis and Revelation of the dynamic change characteristics of the pattern, the temporal and spatial evolution law, the dynamic evolution of the multi scale and the spatial autocorrelation characteristics of the region in the 12 years, based on this, quantitative analysis of the terrain The influence of factors on the spatial pattern of forest vegetation is based on the quantitative analysis of the trend surface of biomass topographic factors to reveal the influence of the topographic factors and their combination conditions on the spatial distribution of altitude, slope and slope direction. Based on the CA-Markov model and the ultimate state equation, the development trend of the future landscape pattern and its ultimate state lattice are simulated and predicted. On the basis of forest landscape naturalness evaluation, the research and ecological design of forest management models with different landscape nature are completed. The rational allocation model of natural forest landscape is put forward from type and landscape scale. Based on (?) principle and uniformity theory, the design of regional forest tending and space pattern control is completed and the demonstration is formed. On the basis of the area, the paper studies the evolution of forest landscape pattern distribution and the technology and methods to solve the forest ecological management planning from the landscape level, and provides comprehensive, systematic methods and ideas for the study of the forest landscape ecological process simulation model and the forest management decision model. The main conclusions are as follows: (1) The analysis of landscape types at the regional level shows that the overall landscape pattern is dominated by woodland landscape, barren hills and wasteland, swamps and residential land and other landscape patches are scattered in the woodland, the area distribution of each landscape type is uneven and the number of patches is more distinct. The forest landscape occupies an absolute dominant position in the study area and the area of the two phases The ratios of 93.47% and 95.39% respectively control the overall landscape structure, function and dynamic process of the study area; the artificial landscape and the semi natural landscape are obviously dominant. The distribution of forest types in the study area is from the low altitudes of the valley and Yang Hualin to the high altitude red pine Cloud Fir Forest and the zonal forest vegetation -- broadleaf. The forest ecological system type formed by the red pine forest dominates the dynamic evolution of the forest landscape pattern in the region. (2) based on the analysis of the landscape forest types of two stages and two stages, the study area is based on the broad leaved mixed forest as the main body, the area ratio of the two periods is 74.94% and 82.45% respectively, both in the broad leaf mixed forest landscape, the Mongolia oak forest landscape and the coniferous and broad-leaved mixed forest landscape. During the period of.12, the succession sequence of the dominant species was the pioneer species (ELM, Yang Hua, Larix gmelinii) and the top species (cloud cold red tussah), at the low stage of succession sequence, the advanced stage and the late succession of Yang Hualin, the area of broad-leaved red pine forest and the coniferous and broad-leaved mixed forest decreased. The area of broad-leaved mixed forest in the middle of succession is increased, the secondary forest area is more obvious, the landscape heterogeneity is high, the near nature is low, the landscape spatial pattern is easily controlled by some environmental factors and interference conditions. (3) 1997-2009 years, the aggregation degree and spread index of the study area are increased from 88.4128 and 97.7300, respectively. Adding to 91.7402 and 98.3815, the index of separation degree decreased from 1.144 to 1.1001, indicating that the degree of landscape fragmentation slowed down, the distribution of landscape types was concentrated and interlaced, and the connectivity was enhanced, which was beneficial to the overall display of landscape function, but the dominance of landscape was increased, the spatial heterogeneity and diversity of landscape were reduced, and the landscape dominance was more than that of the landscape. The two indices showed a certain negative correlation, indicating that the proportion of the landscape types in the region increased. (4) the Moran's I coefficient in 1997 was 0.42 for -0.112009 years. Based on the calculation of General G-statistic, the value of Z Score in 1997 was 0.92 and 2009 was 2.06; and the evenness index based on the uniformity theory in 1997 was 2.06. For 0.32072009 years, the results of 0.2883. based on Moran's I and uniform theory were the same. The two results showed that the landscape pattern of the study area showed a trend from random pattern to cluster pattern in 12 years. The community structure of natural secondary forest was relatively unstable and the global spatial autocorrelation was gathered in high value space in 2009 (H- H aggregation), but the dominance of the overall landscape increased, and the single dominant landscape type enhanced the overall landscape control. (5) based on the analysis of the Geostatistical Analyst of ArcGIS 9.3, the maximum and minimum values of SHDI in 1997 were 1.9722 and 0.2418 respectively, respectively, and the dominant regions of SHDI were concentrated in the research area. In the middle area of the region, we present the strip in the north-east direction of the southwest, and analyze the semi variance function and the theoretical fitting model of the experimentation area based on the geostatistics software GS+. The results show that the forest landscape diversity and the spatial geographic location are of medium spatial correlation and obvious regional structure, based on IDW and Kring. Spatial distribution map of forest landscape diversity index under different scales was obtained. (6) under the support of ArcGIS, based on the spatial superposition of DEM and dominant species, the spatial distribution, size, location or relative position of dominant tree species in various landscape elements were accurately described, and the trend surface based on biomass topographic factors was completed. In high altitudes and low altitudes, the biomass of two phases is changed with the change of slope direction. In high altitude, the biomass of the area with large slope is larger. The slope degree mainly plays a role in the spatial differentiation of the landscape element by the effect of the combined slope to improve the soil water condition. The effect of factors on biomass did not become an important factor in the spatial distribution trend of the whole forest biomass, only affecting the fluctuation of local biomass. (7) according to the ultimate state probability calculation based on CA-Markov, the final state of the study area is the broadleaf mixed forest and the broad-leaved forest. The area of the mixed forest and the other landscape is decreased, the area ratio of the mixed broadleaf mixed forest has reached 59%, and it occupies an absolute dominant position in the study area. The proportion of the mixed forest and the broadleaf red pine forest in the top forest landscape is only 3.9%, and the final situation of the landscape pattern will not be predicted. The whole stability development of the whole research area landscape is not consistent with its forest landscape management goal. (8) based on the community continuous belt index (CI) to determine the succession state of the community, the succession of the forest landscape and the forest landscape, and the advantages of the succession of the forest landscape and the forest landscape, are based on the theory of the natural forest and the ecological repair of the potential natural vegetation. The concept of landscape naturalness is first proposed for the first time, and the concept of landscape naturalness is proposed for the first time. It is divided into top landscape nature, high landscape naturalness, intermediate transition landscape naturalness and primary landscape nature. Based on forest landscape naturalness evaluation, the forest farm scale is the demonstration area for the first time. The forest landscape type in the natural area is the management unit, the management target, the target structure and the structural adjustment from the type level and landscape scale, provide the theoretical basis and technical support for the forest management in Changbai Mountain forest area. (9) with the support of ArcGIS 9.3, the use of VBA program to extract three large 2009 In the early stage of 30m * 30m succession, the relative coordinates of the tree species at the top stages of the substage were replaced, and the GIS thematic database was established. Based on the support of the theory of C# and uniformity, the quantitative calculation of the nearest point distance, the total area and the evenness index of the sample plots was completed, and the results showed that the three plots of land uniformity were 0.331,0.288, respectively. 0.275, according to the (?) principle, after the selective removal of some points, the uniformity of the three plots is calculated to be 0.463,0.346,0.290, and the distribution of the trees following the (?) principle is obtained to optimize the stand pattern.
【学位授予单位】：东北林业大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：S718.5

【参考文献】