Ecosystem comparison and pursuit of vegetation indexes as predictive tools
The present study was conducted to (i) investigate parameters influencing the fluxes of the greenhouse gas methane (CH4) in Danish riparian wetlands with contrasting vegetation characteristics and (ii) develop models relating CH4 emissions to soil and/or vegetation parameters integrating the spatial and temporal variability in the fluxes. Fluxes of CH4 were monitored in 12 wetland plots over a year using static chambers, yielding a dataset with more than 800 measured fluxes of CH4. Yearly emissions of CH4 ranged from −0.2 to 38.3 g CH4-C m−2 year−1, and significant effects of groundwater level, soil temperature (10 cm depth), peat depth, sulfate, nitrate, and soil carbon content were found. Two methods based on easily available environmental parameters to estimate yearly CH4 emissions from riparian wetlands are presented. The first uses a generalized linear model (GLM) to predict yearly CH4 emissions based on the humidity preference of vegetation (Ellenberg-F), peat depth and degree of humification of the peat (von Post index). The second method relies solely on plant species composition and uses weighted-average regression and calibration to link the vegetation assemblage to yearly CH4 emission. Both models gave reliable predictions of the yearly CH4 fluxes in riparian wetlands (modeling efficiency > 0.35). Our findings support the use of vegetation, possibly in combination with some soil parameters such as peat depth, as indicator of CH4 emission in wetlands.