Vegetation types and flood water level are dominant factors controlling the carbon sequestration potential in Dongting Lake floodplain, China.

Wetlands are important carbon sinks. However, the carbon sequestration potential of flooded wetlands may be weakened owing to water regime changes induced by anthropogenic disturbances. Using the eddy covariance technique, this study quantified the effects of the water level and vegetation types on the net ecosystem CO exchange (NEE), gross primary production (GPP), and ecosystem respiration (R) from a reed marsh (Miscanthus sacchariflorus) and a sedge meadow (Carex spp.) in the Dongting Lake floodplain from 2014 to 2016. Our results indicated that the sedge meadow (-89.49 to -186.47 g C m y) and reed marsh (-246.12 to -513.94 g C m y) were carbon sinks on the interannual timescale. However, the sedge meadow changed from a carbon sink to a carbon source during the flooding season. The effect of flooding on the carbon sink function in the reed marsh was dependent on the water level. The carbon sink function of the reed marsh was enhanced by moderate flooding (water level under 30.5 m in Chenglingji) owing to the inhibition of R, but was weakened by extremely high-water levels (over 33 m in Chenglingji) during the flooding season. Seasonal variations in NEE, GPP, and R were closely related to photosynthetic photon flux density, soil water content, water level, soil temperature, and air temperature. We can conclude that the increase in reed area combined with the decrease in flooding days in the sedge meadow can potentially enhance the carbon sink function of the Dongting Lake floodplain.