Factors Controlling Diffusive Co2 Production and Transport in the Cedarburg Bog, Saukville, Wisconsin

Abstract

Wetlands are vital components of the carbon cycle, containing an estimated 20-30% of the global soil carbon reservoir. The Cedarburg Bog of southeastern Wisconsin boasts a myriad of wetland habitats including the southernmost string bog found in North America. Soil carbon dioxide (CO2) behavior in these systems is the response of multiple interdependent variables that are, collectively, not well understood. Many studies have measured and modeled soil CO2 flux (soil respiration) based on isolated, intermittent measurements that do not account for the full range of soil CO2 flux intensity. In the Cedarburg Bog, high-resolution measurements of soil CO2 flux were recorded over two field seasons using a LI-COR 8100A soil gas flux analyzer at 30 minute (May-Nov., 2014) and 60 minute (June-Oct., 2015) intervals. Additionally, soil moisture and temperature data were collected, and weather station variables (atmospheric temperature, radiation, wind, pressure, precipitation) were acquired for correlations. Stable isotope signatures were interpreted from a peat core (δ13C, δ15N) and from gaseous CO2 at the surface (δ13C) to determine sources of soil respired CO2. The intensity of soil CO2 flux was broadly distributed across the entire data set, ranging from less than 1 to over 650 mg/min-m2. Average for all soil CO2 flux measured was 6.49 mg/min-m2, with a median of 3.39 mg/min-m2. Soil respiration was attributed to two main sources: 1) microbial respiration, and 2) root respiration (including rhizo-microbial respiration). Microbial respiration was, in part, influenced by soil temperature, and produced a constant, low flux ( 50 mg/min-m2) were observed, but were not correlated to the included parameters. Although high flux occurred much less frequently, it produced a significant amount of the CO2 mass respired from the soil. Correlations between soil CO2 flux and controlling parameters were addressed using JMP; multiple linear regression models presented weak and significant correlations due to the absence of lag/response time variables for assimilation and transport mechanisms of CO2. Wetland soils are structurally complex, and can be highly variable through time; improving correlations for soil respiration models requires high-resolution data sets, and determination of lag/response times of CO2 transport processes above ground and in the soil.