国外植被数量分类与排序研究进展OA北大核心CHSSCDCSCD

Numerical classification and ordination of vegetation can reveal the ecological relationships among plant species and their interactions with environmental factors more objectively and accurately.Over the past 70 years,methods for numerical classification and ordination have been continuously developed,gradually expanding their application fields,and analytical software and statistical tools have consistently developed.This paper elaborates on the international research progress in vegetation numerical classification and ordination in terms of classification methods,evaluation of classification methods,ordination methods,integrated application of ordination and classification methods,and analytical software,and a particular focus is put on the progress since 1990,including classification methods such as multivariate regression trees(MRT)and adaptive affinity propagation clustering(AAPC),classification evaluation methods like geometric and non-geometric evaluation,ordination methods such as non-metric multi-dimensional scaling(NMDS),combined application of ordination and classification methods such as two-way indicator species analysis(TWINSPAN)and detrended correspondence analysis(DCA),as well as statistical analysis tools such as CANOCO and JUICE.Compared to using a single quantitative classification or ordination method,the synergistic application of multiple methods can yield more scientific classification and ordination results.