Response of water vapor and carbon dioxide fluxes in semi-arid plant communities to variations in precipitation

Eddy covariance measurements of Net Ecosystem Exchange (NEE) of water vapor and CO2 were made over three Great Basin ecosystems: crested wheatgrass (Agropyron desertorum), Utah juniper ( Juniperus osteosperma), and sagebrush (Artemisia tridentata ) in 2002 and 2003. To our knowledge, this is the first study of NEE of CO2 and water vapor in these ecosytems. Juniper and sagebrush had significantly higher water fluxes compared to crested wheatgrass, presumably due to better access to deeper soil water.; Evapotranspiration (ET) decreased during the summer at all sites. However, the juniper and sagebrush communities remained physiologically active, and showed significant ET rates throughout the summer, while crested wheatgrass senesced. Each plant community had significant net assimilation of CO 2 during the daytime in the spring of both years. Juniper had the largest negative CO2 fluxes, with sagebrush intermediate and crested wheatgrass the lowest. Sagebrush and juniper were able to fix carbon during the warm and dry summer conditions, while crested wheatgrass became senescent. The very low turbulence at night resulted in an inability to quantify CO 2 fluxes under most nighttime conditions. The effects of intermittent precipitation events on water vapor and CO2 fluxes were determined for the three plant communities. Significant responses to rain events were observed, although the magnitude of the response varied with season and among the three plant communities. The greatest response of fluxes to rain events was observed in juniper, followed by sagebrush and crested wheatgrass. Net CO2 exchange changed significantly at the juniper and sagebrush sites but very little at the crested wheatgrass site, due to senescence. Variations in precipitation appear to affect the water and carbon balance of the three communities differently. Comparisons between an eddy covariance and Bowen ratio (BR) system were made over crested wheatgrass from April to July, 2002. ET fluxes compared well in the spring, but the BR system estimated larger ET values during summer. There were large discrepancies in CO2 fluxes throughout the measurement period. Given that the eddy covariance systems were able to close the energy balance reasonably well, this suggests that the BR approach is unable to estimate fluxes in this type of ecosystem. Results of this research will increase our understanding of carbon and water fluxes in semiarid ecosystems and help to better predict their sensitivity to global change.