| 摘要: |
| 【目的】分析森林结构参数与取样面积之间的关系，确定评价天然林状态特征的最小取样面积，为指导天然林生态监测与经营提供理论依据。【方法】以小陇山林区天然林作为研究对象，选取3块面积为1 hm2的天然林固定监测样地，采用巢式取样法，按照10 m×10 m，20 m×20 m，…，90 m×90 m的嵌套方式设置取样面积，分析树种数、树种多样性（Shannon-Wiener指数）、林木株数、胸高断面积、树种隔离程度（混交度）、林木分布格局（角尺度）与取样面积之间的关系，采用对数模型、幂函数模型和逻辑斯蒂模型，分别对样地树种-面积、Shannon-Wiener指数 面积进行拟合。【结果】树种数、树种多样性与取样面积的曲线特征均为初期急剧上升，而后逐渐趋于平缓，曲线拐点出现在取样面积1 600和2 500 m2处。树种-面积模型和Shannon-Wiener指数 面积模型均以逻辑斯蒂模型为最优。不同取样面积下每公顷林木株数和胸高断面积的估算值与实测值比较结果表明，当3块样地取样面积增大到1 600 m2时，每公顷林木株数估算值与实测值的误差均降至10%以下；1、2号样地取样面积增大到1 600 m2、3号样地取样面积增大到2 500 m2时，每公顷胸高断面积估算值与实测值的误差均降至10%以下。混交度分析表明，当取样面积较小时，混交度各频率分布表现为无规律，3块样地平均混交度趋于稳定的取样面积依次为1 600，2 500和900 m2。角尺度分析表明，1号样地中各取样面积的平均角尺度均表明林地中林木为聚集分布，但当取样面积＜2 500 m2时平均角尺度上下波动较大；2号样地中当取样面积≥900 m2时平均角尺度趋于稳定，林地中林木表现为聚集分布；3号样地中当取样面积≥2 500 m2时平均角尺度趋于稳定，林地中林木表现为随机分布。【结论】当取样面积达到2 500 m2时，各参数所表现的规律均趋于稳定，因此，在小陇山林区研究天然林结构特征的最小取样面积应不小于2 500 m2。 |
| 关键词: 小陇山林区 巢式取样 树种 物种多样性 混交度 胸高断面积 角尺度 取样面积 |
| DOI： |
| 分类号: |
| 基金项目:国家自然科学基金项目“基于相邻木关系的林分结构多样性研究”（31670640） |
|
| Determination of minimum sampling area for natural forest investigation in Xiaolongshan forest region |
|
YUAN Yichao,LI Chunlan,ZHAO Zhonghua,et al
|
| Abstract: |
| 【Objective】This study determined the minimum plot area needed for evaluating natural forest characteristics by analyzing the relationship between forest structure parameter and plot area to provide basis for ecological monitoring and forest management.【Method】This study focused on natural forest in Xiaolongshan forest region and three 1 hm2 permanent plots were selected to set plots with areas of 10 m×10 m,20 m×20 m,…,90 m×90 m using the nested sampling method.The relationships of plot area with trees species,Shannon-Wiener index,mingling,stem number,basal area and uniform angle index were investigated,respectively.The logarithmic model,power function model and logistic model were used to fit tree species with plot area and Shannon Wiener index with plot area.【Result】The tree species,diversity and plot area increased dramatically at the early stage and tended to stable later with peaks at plot areas of 1 600 m2 and 2 500 m2.The logistic model had the best fitting for species with plot area and Shannon Wiener index with plot area.When the plot area was 1 600 m2,the error between estimated and measured values of trees per hectare was less than 10%.When the areas of plots 1 and 2 were 1 600 m2 and the area of plot 3 was 2 500 m2,the error between estimated and measured values of basal area hectare was less than 10%.The analysis of mingling indicated that the frequency distribution of mingling had no regularity with very small plot area. The average mingling became stable when the plot areas of three plots were 1 600 m2,2 500 m2,and 900 m2.The analysis of uniform angle index indicated that average uniform angle index of plot 1 was aggregated with different plot areas,but it varied greatly when plot area was less than 2 500 m2.The average uniform angle index value of plot 2 was aggregated and tended to be stable when sampling area was larger than 900 m2,while the average uniform angle index of plot 3 was random and tended to be stable when plot area was larger than 2 500 m2.【Conclusion】When the plot area was larger than 2 500 m2,all stand parameters tended to be stable.Therefore,the minimum sampling area for analyzing characteristics of natural forest in Xiaolongshan forest region was 2 500 m2. |
| Key words: Xiaolongshan forest region the nested sampling method tree species species diversity mingling basal area uniform angle index sampling area |
