莫索湾南缘沙漠植物群落多样性抽样方法的研究

Sampling methods of plant community diversity in the desert of the southern edge of Muosuowan region,Xinjiang

2005年9月中下旬,在莫索湾南缘沙漠设置的200m×40m样带内共观测到植物物种22个,其中草本物种占77．3%。通过种一面积方程拟合和临界抽样面积的求解发现要达到抽取群落60%以上物种,临界样方面积至少在518．33m^2以上,而草本物种在735．49m^2以上。样方间相似性指数分析发现样方布样经历的生境异质性越高,物种替代性增加,相似性降低,抽样效率提高。研究发现沿垂直于沙丘走向的方向布样,以4m×4m为基本样方,抽样272m^2即达到60%群落物种,远低于临界抽样面积的求解结果,表明基本样方大小和设置方向对抽样效率产生重要影响。由于沙漠物种偶见种多,频度和密度较低,并且分布的环境特异性强,认为对沙漠植物群落观测,样线法、巢式样方法以及对草本观测的梅花样方法均具有一定局限性,建议对草本和灌木物种多样性观测,在确定合理的布样方向基础上,以4m×4m为基本样方。提出对沙漠植物群落多样性结构和功能有全面了解,必须将局部抽样与区域抽样结合,增加抽样频率和强度。

The research was conducted in Muosuowan region in the southern edge of the Gurbantunggut Desert during the period of September 25 - 30, 2005. In a plot of 8 000 m^2（200 m × 40 m）, twenty two species of plants were observed, of which 22.7% belonged to shrub and others were herbaceous species. Species such as Ceratoides ewersmanniana , Ceratocarpus arenarius , Perosimonia sibirica and Perosirnonia soongorica played a dominant role in this area. After fitness test of the species-area equation and a calculation of the critical sampling area （CSR）, it was discovered that CSR should be above 518.3 m^2 in order to ensure that more than 60% of all the species in the community being included in the experiment. Because of the low distributional frequency and habitat-specific characteristic of herbaceous species, the optimal CSR of herbaceous species should be above 735.49 m^2. Through the analysis of the similarity index among quadrats, it was discovered that the replacement of species increased, the similarity among species decreased, and the sampling efficiency improved with an increase in habitat heterogeneity of quadrats. If 4 m × 4 m is taken as a basic sampling quadrat, a sampling area of 272 m^2 would be large enough to include 60 96 of all the species in the community. As the results of this sampiing area were far below those derived from calculation of the critical sampling area, the paper concluded that the setting size and direction of a basic sampling quadrat could have a great influence on the efficiency of each sampling. Because there were many occasional species in the desert and the distribution of many species were dependent largely on the habitat, the paper concluded that neither the methods of sampling line and nest method for all species nor the sampling method of the pattern of plum blossom quadrat for the herbaceous species could be suitable for observation in the desert. And the paper suggested that sampling in desert should take 4 m × 4 m as basic quadrat towards vertical sand hill direction. In this paper, it is pointed out that in order to get a general understanding of the structure and function of the diversity of plant community in the Gurbantunggut desert, the observation should combine local sampling with regional sampling and increase the frequency and intensity of each sampling.