Patterns of carbon dioxide and water vapor flux following harvest of tallgrass prairie at different times throughout the growing season

Abstract

Most rangelands are harvested at some point during the year and removal of plant leaf area and biomass alters a host of ecosystem processes including gas exchange. An experiment was conducted in 2005 and 2006 to study the effects of clipping tallgrass prairie at different dates on water vapor and CO2 fluxes. A portable, non-steady-state chamber was designed to measure CO2 and water vapor fluxes from small plots in less than 40 s. A combination of sunlit and shaded readings allowed measurements of net carbon exchange (NCE) and ecosystem respiration (RE); by summing NCE and RE, gross canopy photosynthesis (GCP) was calculated. Throughout the two-year study, the chamber had a minimal effect on microclimate, i.e., average chamber temperature increased 2.9° C, while chamber pressure increased only 0.3 Pa during measurements, and photosynthetically active radiation attenuation was 10%. The immediate effect of all clipping treatments was a loss of leaf area that led to reductions in GCP, NCE, and RE and in most cases decreased water vapor flux. Further patterns of carbon flux were governed by the amount of water stress during canopy development, while water vapor flux rates varied with water availability. Canopies that developed during periods of low water stress quickly increased carbon flux rates following precipitation after a mid-season drought. However, flux rates of canopies, which developed during the mid-season drought, responded considerably slower to subsequent water availability. A separate experiment was conducted from June-October of 2006 to estimate GCP, leaf area index (LAI), and total aboveground biomass with a hyperspectral radiometer. Indices such as the Normalized Difference Vegetation Index and the Simple Ratio were used to estimate LAI and biomass had poor correlations with measured values. However, GCP was significantly correlated to all six indices derived in this study. While GCP measured from June-October was significantly correlated with all indices, removal of the senesced canopy scans recorded during October greatly increased the relationship.