PLANT COMMUNITY DIVERSITY IN DONGLINGSHAN MOUNTAIN,BEIJING,CHINA
北京东灵山地区植物群落多样性研究 Ⅵ.几种类型植物群落物种数目的估计

Empirical Bayes method(EBM),nonparametric methods (NPMs) (with include jack knife and bootstrap methods) and extrapolation of species area curves were used to estimate the number of species ( NS )in each of the five types of plant communities in Dongling Mountain,Beijing,China,and their performances were also examined.The following are the results.(1)as sampling intensity ( Sl ) was not very low,good point estimates be obtained with EBM,but the estimated standard deviations were greater than those with NPMs;(2) NPMs would also give good point estimates under adequate Sl .as Sl was too low,the estimated NS be low.On the contrary,as Sl was too high.the estimated NS wold be high.Under all of the Sls,the estimated standard deviations with NPMs were less than those with EBM;(3)Species area curves 2,3,4,5,6 gave good point estimates of NS .Curve 1 overestimated the NSs while curves 7,8,9,10 underestimated the NSs for many cases;(4)Synthesizing the results obtained with the three kinds of methods.we can give the estimated NSs for the five communities and their three layers for each community:16,19,21,12,16 for trees,13,15,21,14,21 for shrubs,49,46,54,82,55 for herbs,and 79,80,96,107,92 for communities.