Abstract: The herbaceous vegetation in wetlands of the Yellow River Delta, China was classified into 7 community types by TWINSPAN. Double principal coordinate analysis (DPCoA) was then applied for the ordination of the communities. The results from DPCoA ordination indicated that in species composition, Ass. Phragmites australis+ Suaeda salsa was a transitional type of Ass. Phragmites australis and Ass.Suaeda salsa+Phragmites australis, Ass. Suaeda salsa, Ass. Suaeda salsa+Limonium sinense+Suaeda glauca, and Ass. Limonium sinense, while Ass. Phragmites australis+Myriophyllum spicatum showed a great difference from others. The distribution of different community types was significantly correlated with soil salinity and soil pH, but not significantly correlated with soil total phosphorus, total nitrogen, and organic matter. Compared with principal component analysis (PCA) and detrended correspondence analysis (DCA), DPCoA may keep more accumulative information from the original data. It congregated the communities in the ordination chart which had fewer differences in their species composition and species taxonomy, but dispersed the communities which had greater differences in their species composition and species taxonomy. Therefore, DPCoA may be more efficient in investigating the structural relationships between different commnnty types and their relationships to environmental variables.

Key words: Plant polyphenols, Property, Agriculture, Environment, Application