Wetland Flux Controls: How does interacting water table levels and temperature influence carbon dioxide and methane fluxes in Northern Wisconsin?

Daily (a) and annual (b) average water table level vs average discharge at Bear River. Existing water table data and discharge at Bear River were correlated both at annual (R2 = 0.954, p value = 0.0002, 95% CI shown) and daily (R2 = 0.628, p value \0.0001, 95% CI shown) scales, when using daily data above low discharge threshold. Time series of water discharge at Bear River at Lost Creek Wetland from 2001 to 2015, with circles showing the average discharge of the month with the highest average in each year and solid line showing period when water table depth data was available (c). NDVI at Lost Creek wetland from 2000 to 2016 based on MODIS observations (d)

Citation

Pugh, C.A., Reed, D.E., Desai, A.R., Sulman, B.N., 2018. Wetland Flux Controls: How does interacting water table levels and temperature influence carbon dioxide and methane fluxes in Northern Wisconsin?, Biogeochemistry, 137(1–2), 15–25, doi:10.1007/s10533-017-0414-x.

*** Undergrad first author ***

Abstract

Wetlands play a disproportionately large role in global terrestrial carbon stocks, and from 1 year to the next individual wetlands can fluctuate between carbon sinks and sources depending on factors such as hydrology, temperature, and land use. Although much research has been done on short-term seasonal to annual wetland biogeochemical cycles, there is a lack of experimental evidence concerning how the reversibility of wetland hydrological changes will influence these cycles over longer time periods. Five years of drought-induced declining water table at Lost Creek, a shrub fen wetland in northern Wisconsin, coincided with increased ecosystem respiration (Reco) and gross primary production (GPP) as derived from long-term eddy covariance observations. Since then, however, the average water table level at this site has increased, providing a unique opportunity to explore how wetland carbon fluxes are affected by interannual air temperature differences as well as changing water table levels. Water table level, as measured by water discharge, was correlated with Reco and GPP at interannual time scales. However, air temperature had a strong correlation with Reco, GPP, and net ecosystem productivity (NEP) at monthly time scales and correlated with NEP at inter-annual time scales. Methane flux was strongly temperature-controlled at seasonal time scales, increasing an order of magnitude from April to July. Annual methane emissions were 51 g C m-2 . Our results demonstrate that over multiyear timescales, water table fluctuations can have limited effects on wetland net carbon fluxes and instead at Lost Creek annual temperature is the best predictor of interannual variation.