不同群落蒙古栎种群空间格局的地统计学分析

以蒙古栎天然次生林中的蒙古栎种群为研究对象,在吉林省汪清林业局塔子沟林场设置2块1 hm²的处于不同演替阶段的样地(A、B).采用相邻网格调查法将每块样地划分为 100 个10 m×10 m的调查单元,对单元内每株林木的空间坐标进行精确定位,调查所有胸径≥1 cm的林木基本信息.采用地统计学分析的半方差函数法和分维数对蒙古栎种群各林木属性的空间异质性程度、组成、尺度、方向进行分析;运用克里格插值法对具有空间自相关的树木属性进行无偏估计并绘制分布图,分析其空间分布格局.结果表明: 两块样地各林木属性的最优半方差函数以指数模型和球状模型为主,呈聚集分布,但样地A较样地B的空间自相关程度更高,空间连续性更大;两块样地内部,胸径和东西冠幅均表现出较强的空间异质性和空间自相关性.两块样地各林木属性均在南北方向上表现出较强的空间异质性.此外,样地A在西北-东南方向上也存在较强的空间异质性,而样地B则在东北-西南方向存在较强的空间异质性.两者相比,样地A的空间异质性强度更高、尺度更大.样地A中胸径和东西冠幅变异明显,而样地B中东西冠幅和南北冠幅变异明显.分维数值反映的结果与标准半方差函数值的结果基本一致.样地A各林木属性变量以斑块状和条带状分布为主,空间分布格局和变化趋势明显,而样地B各林木属性变量分布破碎,格局复杂.上述结果说明,种群属性特征、群落发育程度、空间尺度大小和空间水平方向可能影响种群的空间格局.基于地统计学的分析方法有助于定量、直观地描述种群的生长现状和发展趋势,可为东北林区大面积的蒙古栎天然次生林的可持续经营提供理论基础.

Geostatistical analysis on the spatial pattern of Quercus mongolica population in different communities.

Taking Quercus mongolica population in the secondary forest of Q. mongolica as the research object, two plots in different stages of succession (A and B) were set up in Tazigou Forest Farm of Wangqing Forestry Bureau, Jilin Province. By applying the method of adjacent grid survey, each plot was divided into 100 units of 10 m×10 m and the spatial coordinates of each tree in the unit were accurately located to survey all the basic information of trees with diameter at breast height (DBH)≥1 cm. The degree, composition, scale and pattern of spatial heterogeneity of individual tree of Q. mongolica were analyzed by means of semi-variance function and fractal dimension of geostatistics. By using Kriging interpolation method, unbiased estimation of tree attribute with spatial autocorrelation was carried out, distribution map was drawn and spatial distribution pattern was analyzed. The results showed that the best semi-variance function of tree attributes in two plots was mainly distributed in an exponential model and a spherical model with an aggregated distribution. The degree of spatial autocorrelation and continuity of plot A were higher than that of plot B. The DBH and the east-west crown (C_EW) had strong spatial heterogeneity and autocorrelation in the two plots. The tree attributes of both plots showed strong spatial heterogeneity in the north-south direction. In addition, there was strong spatial heterogeneity in the northwest-southeast direction of plot A and in the northeast-southwest of plot B. The strength of the spatial heterogeneity was higher and the scale being larger in plot A. The variations of DBH and C_EW were obvious in plot A, while the variations of C_EW and south-north crown (C_SN) were obvious in plot B. The fractal dimension and semi-variogram function showed the same result. The tree attributes of plot A were mainly patchy and stripe, and the variation trend of spatial distribution pattern was obvious. The tree attributes of plot B was broken, with complex pattern. Those results indicated that the characteristics of population, community development, spatial scale and spatial horizontal direction might affect the spatial pattern of populations. The geostatistical analysis method is helpful to quantitatively and directly describe population growth and development trend, which can provide a theoretical basis for the sustainable management of Q. mongolica secondary forests in Northeast China.