小兴安岭阔叶红松林木本植物种-面积关系

Woody plants species-area relationships in a broad-leaved Korean pine forest in the Xiaoxing'an Mountains

种-面积关系研究是了解植物群落结构的重要途径,是群落生态学的基本问题。不同的研究方法对种-面积关系影响很大。利用黑龙江省小兴安岭两个10.4 hm2样地和5个1.0 hm2样地的调查数据,采用移动窗口法确定各样地的最小取样面积,避免了巢式取样法及随机样方法的不足。并采用4种种-面积关系模型进行拟合,评价各关系模型的适合度。在此基础上,基于最小面积进行模拟随机取样,探讨取样大小对物种数估计精度的影响。研究结果表明:由于拟合曲线模型的适用性及曲线外推可靠性问题的存在,采用拟合曲线的方法所估计的最小面积与实际值偏差较大。实际调查得到的各样地最小面积40 m×40 m—45 m×45 m,说明小兴安岭地区阔叶红松林群落所需的最小面积基本一致,但各样地群落结构的差异却在对取样数量的要求上体现出来。其中丰林与大亮子河样地物种数分布相对均匀,所需最小样方数量较少;而方正与胜山样地物种数分布异质性较大,差异的机理还有待于进一步研究。

The species-area relationship is an important approach for understanding the structure of the plant community, which represents a fundamental issue in ecology, sampling area of a community can be determined and is an important component of community by a species-area curve. However, different surveys. The minimum sampling methods have dramatic effects on the species-area relationship. Previous studies have shown that the minimum sampling area can be determined by species-area relationship models （e.g., logarithmic function, power function, and logistic function）, which are constructed from nested plots and quadrate combination methods. However, the forms and parameters of the species-area relationship models vary with sampling method, regional location, and spatial scale. The nested plot method may increase the probability of appearance of rare species, resulting in exaggeration of the number of species, while the quadrate combination method may neglect the aggregation distribution. Therefore, further research is needed to develop methods with decreased uncertainty. Here, based on the inventory data from two 10.4 hm^2 plots （Shengshan and Liangshui） and five 1.0 hm^2 plots （Fangzheng, Tangwanghe, Daliangzihe, Dongzhelenghe, and Fenglin） in the broad-leaved Korean pine （Pinus koraiensis） mixed forests in Xiao Xing＇an mountains, Heilongjiang Province, the minimum sampling areas were determined by species-area curves with the moving window method. This method overcame the limitations of the nested sampling and quadrate combination methods and prevented the interference of sampling error on the species-area relationship caused by the community spatial heterogeneity. Additionally, the species-area curves of the seven plant communities were fitted with four saturation curve models to simulate the minimum sampling areas, and the suitability of these fits was evaluated. Furthermore, random sampling simulation was carried out with the minimum sampling area to examine the effects of sampling number on the accuracy of the species number estimation. The results showed that the measured minimum sampling areas of the seven forest communities with the moving window method varied from 40 m × 40 m to 45 m × 45 m, demonstrating that the minimum sampling areas of the broad-leaved Korean pine forest communities were similar. The R2 values of the four species-area relationship models were all greater than 0.8 for the seven plots, but the minimum sampling areas determined by the models differed greatly from that measured because of the suitability of the models and the reliability of the curve extrapolation. Moreover, the proportions of the species of minimum sampling areas to that of the entire plot were overestimated by the models. By simulating the process of random sampling, we examined the effects of sampling number on the accuracy of species number estimation and found that the differences in community structures could be reflected in the demands of sampling numbers. Plots at Fenglin and Daliangzihe required only a small number of samples, indicating an even spatial distribution of species. However, the plots at Fangzheng and Shengshan required a large number of samples, indicating high spatial heterogeneity in the species distribution. Further studies are required to determine the mechanisms responsible for this difference.