Effects of increased soil water availability on grassland ecosystem carbon dioxide fluxes

There is considerable interest in how ecosystems will respond to changes in precipitation. Alterations in rain and snowfall are expected to influence the spatio-temporal patterns of plant and soil processes that are controlled by soil moisture, and potentially, the amount of carbon (C) exchanged between the atmosphere and ecosystems. Because grasslands cover over one third of the terrestrial landscape, understanding controls on grassland C processes will be important to forecast how changes in precipitation regimes will influence the global C cycle. In this study we examined how irrigation affects carbon dioxide (CO₂) fluxes in five widely variable grasslands of Yellowstone National Park during a year of approximately average growing season precipitation. We irrigated plots every 2 weeks with 25% of the monthly 30-year average of precipitation resulting in plots receiving approximately 150% of the usual growing season water in the form of rain and supplemented irrigation. Ecosystem CO₂ fluxes were measured with a closed chamber-system once a month from May-September on irrigated and unirrigated plots in each grassland. Soil moisture was closely associated with CO₂ fluxes and shoot biomass, and was between 1.6% and 11.5% higher at the irrigated plots (values from wettest to driest grassland) during times of measurements. When examining the effect of irrigation throughout the growing season (May–September) across sites, we found that water additions increased ecosystem CO₂ fluxes at the two driest and the wettest sites, suggesting that these sites were water-limited during the climatically average precipitation conditions of the 2005 growing season. In contrast, no consistent responses to irrigation were detected at the two sites with intermediate soil moisture. Thus, the ecosystem CO₂ fluxes at those sites were not water-limited, when considering their responses to supplemental water throughout the whole season. In contrast, when we explored how the effect of irrigation varied temporally, we found that irrigation increased ecosystem CO₂ fluxes at all the sites late in the growing season (September). The spatial differences in the response of ecosystem CO₂ fluxes to irrigation likely can be explained by site specific differences in soil and vegetation properties. The temporal effects likely were due to delayed plant senescence that promoted plant and soil activity later into the year. Our results suggest that in Yellowstone National Park, above-normal amounts of soil moisture will only stimulate CO₂ fluxes across a portion of the ecosystem. Thus, depending on the topographic location, grassland CO₂ fluxes can be water-limited or not. Such information is important to accurately predict how changes in precipitation/soil moisture will affect CO₂ dynamics and how they may feed back to the global C cycle.