Net Carbon Fluxes Over Burned and Unburned Native Tallgrass Prairie

Abstract

Prescribed burning of aboveground biomass in tallgrass prairie is common and may influence dynamics and magnitudes of carbon (C) movement between the surface and atmosphere. Carbon dioxide (CO2) fluxes were measured for 2 yr using conditional sampling systems on two adjacent watersheds in an ungrazed tallgrass prairie near Manhattan, Kansas. One watershed was burned annually (BA) and the other biennially (BB). Leaf and soil CO2 fluxes were measured in the source area. Net ecosystem exchange (NEE) of CO2 reached a maximum daily gain of 26.4 g CO2 m-2 d-1 (flux toward surface is positive) in July 1998 (year when both sites were burned and precipitation was above normal); gains were similar between sites in 1998. The maximum daily NEE loss of CO2 was 221.8 g CO2 m-2 d-1 from BA in September 1997 (year when only BA was burned and precipitation was below normal). When data were integrated over the two years, both sites were net sources of atmospheric CO2; NEE was –389 g C m-2 2 yr-1 on BA and –195 g C m-2 2 yr-1 on BB. Burning increased canopy size and photosynthesis, but the greater photosynthesis was offset by corresponding increases in respiration (from canopy and soil). Carbon losses from fire represented 6-10% of annual CO2 emissions (bulk came from soil and canopy respiration). Data suggest that annual burning promotes C loss compared to less-frequently burned tallgrass prairie where prairie is not grazed by ungulates. Greater precipitation in 1998 caused large increases in biomass and a more positive growing season NEE, indicating that C sequestration appears more likely when precipitation is high. Because C inputs (photosynthesis) and losses (canopy and soil respiration) were large, small measurement or modeling errors could confound attempts to determine if the ecosystems are long-term CO2 sources or sinks.