Abstract:

【Objective】There are few quantitative researches concerned about the relationships between the ecological characteristics of wetland plant and water level which could probably be key to solving the environmental issues in Sanjiang Plain, Northeast, China. Therefore, the purpose of this paper is to analyze the influences of the hydrological regime on the wetland plant community by exploring the relationships between the hydrological factors and the types of wetland plant community, as well as the coupling relationships between hydraulic gradient and wetland plant’s ecological characteristics in the wetland of Sanjiang Plain.【Method】Our team chose the Honghe National Nature Reserve (HNNR) in the Sanjiang Plain of Northeast China as our research area. We had conducted several field surveys to select 9 line transects with the obvious hydraulic gradient, totally 90 sampling spots as the experimental data. The data were further measured to establish a matrix with the size of 21×90 that the sampling spots as ordinate and the species as abscissa. The major analyzing approaches consist of three main steps. The first thing was to use the PC-ORD 5 to analyze the matrix data by Two-Way Indicator Species Analysis (TWINSPAN) in order to execute the quantitative classification of wetland plant communities and analyze the relationship between the plant community and hydraulic gradient in the HNNR. Secondly, we used the CANOCO 5 to execute the Detrended Correspondence Analysis (DCA) with the matrix data to explore the relationship between the distributions of the wetland plant community and hydraulic gradient. Finally, we used the Gaussian logistic regression model to determine a relationship between the plant height, biomass of typical wetland vegetation and water level. The aim of the last step is to explore the influences of the hydraulic gradient on the ecological characteristics of wetland vegetation.【Result】The results of TWINSPAN indicated that the wetland plant in HNNR can be divided into 12 different types of community. The trend of plant types is transited from the hygrophilous vegetation to xeromorphic vegetation with the hydraulic gradient. This experimental result proved the change of water level is critical to succession and distribution of wetland plants. Moreover, the sequencing diagram of DCA reflected that the distribution of the sample showed certain regularity. The types and distributions of wetland plants were affected by the hydraulic gradient as well. Furthermore, the sequencing consequences of DCA species also showed that the hygrophilous vegetation, the wet mesophytic vegetation and the xeromorphic vegetation were all distributed obviously with exact scopes and boundaries. These results could probably be the strong evidences to prove that the main environmental factor affecting the distributions of the wetland plant in HNNR is the hydraulic gradient. Moreover, the results of the Gaussian logistic regression model demonstrated that the height and biomass of the three typical wetland plants all increased first and then decreased with the changing trends of water level. The most suitable ecological amplitude of water level of Calamagrostis angustifolia, Carex lasiocaropa and C. pseudocuraica were: [4.46 cm, 20.04 cm], [8.30 cm, 28.40 cm] and [40.87 cm, 48.71 cm], respectively. The height and biomass of the three wetland plants corresponding to the maximum water level were all sorted as: Calamagrostis angustifolia < Carex lasiocaropa < C. pseudocuraica which was consistent with the classification and sequencing results. So, the height and biomass of Calamagrostis angustifolia were most sensitive to the hydraulic gradient.【Conclusion】We concluded the hydraulic gradient is consequently the major environmental factor influencing the types and distributions of wetland plant communities in HNNR. The distributions of dominant species represented the distributing characteristics of the plant community and have functions as indicator habitats. Finally, our results also proved that the hydraulic gradient also has extremely significant impacts on the ecological characteristics of wetland plants in the reserve area and differences of the adaptability to the flooded water level among different plants could resulted in the obvious differences of their most suitable ecological amplitude of water level. Importantly, this study is to provide a scientific basis for the ecological restoration and ecosystem management of the wetlands in HNNR.

Key words: wetland vegetation, two-way indicator species analysis, detrended correspondence analysis, water level, ecological amplitude, Sanjiang Plain Wetland,Northeast China