Wednesday, November 15, 2006

Temporal Stability of Soil Respiration in Tallgrass Prairie: Towards Watershed-scale Estimates of Carbon Fluxes.

Kira Arnold1, Jay Ham1, Clenton Owensby1, and Patrick Coyne2. (1) Kansas State Univ, Dept of Agronomy, Manhattan, KS 66506, (2) Kansas State Univ Ag. Research Cntr. Hays, 1232 240th Ave., Hays, KS 67601-9228

The spatial variation in soil CO2 flux was determined on tallgrass prairie south of Manhattan, KS between May and October, 2006. Fluxes were measured weekly along two 400-m transects using a hand-held Licor LI-8100 CO2 flux analyzer The sampling transects surveyed a wide range of landforms (e.g., uplands and lowlands), as well as variations in species composition and soil types. Soil collars were installed every 15 m for a total of 28 sampling points per transect. Digital elevation maps were used to develop elevations and topographic indices for each sample site. Peak aboveground biomass was determined by clipping small plots near each soil collar in late July. Because fluxes were not stationary during the two hours required to read each transect, a detrending technique was developed to normalize a set of readings to a single moment in time. Data were analyzed for temporal stability. Results showed that soil CO2 fluxes were highly variable across the landscape. The difference between the minimum and maximum flux along a transect would often exceed 10 μmol m-2 s-1 . The rank of certain sampling locations was very stable among sampling dates. Spearman rank and mean relative differences were used to test for the presence of temporal stability. Results will help determine the number of sampling locations required to make areal estimates of watershed scale soil CO2 flux in tall grass prairie.