Isotopic air sampling in a tallgrass prairie to partition net ecosystem CO2exchange

Abstract

Stable isotope ratios of various ecosystem components and net ecosystem exchange (NEE) CO2fluxes were measured in a C3‐C4mixture tallgrass prairie near Manhattan, Kansas. The July 2002 study period was chosen because of contrasting soil moisture contents, which allowed us to address the effects of drought on photosynthetic CO2uptake and isotopic discrimination. Significantly higher NEE fluxes were observed for both daytime uptake and nighttime respiration during well‐watered conditions when compared to a drought period. Given these differences, we investigated two carbon‐flux partitioning questions: (1) What proportions of NEE were contributed by C3versus C4species? (2) What proportions of NEE fluxes resulted from canopy assimilation versus ecosystem respiration? To evaluate these questions, air samples were collected every 2 hours during daytime for 3 consecutive days at the same height as the eddy covariance system. These air samples were analyzed for both carbon isotope ratios and CO2concentrations to establish an empirical relationship for isoflux calculations. An automated air sampling system was used to collect nighttime air samples to estimate the carbon isotope ratios of ecosystem respiration (δR) at weekly intervals for the entire growing season. Models of C3and C4photosynthesis were employed to estimate bulk canopy intercellular CO2concentration in order to calculate photosynthetic discrimination against13C. Our isotope/NEE results showed that for this grassland, C4vegetation contributed ∼80% of the NEE fluxes during the drought period and later ∼100% of the NEE fluxes in response to an impulse of intense precipitation. For the entire growing season, the C4contribution ranged from ∼68% early in the spring to nearly 100% in the late summer. Using an isotopic approach, the calculated partitioned respiratory fluxes were slightly greater than chamber‐measured estimates during midday under well‐watered conditions. In addition, time series analyses of our δRmeasurements revealed that occasionally during periods of high wind speed (increasing the sampling footprint) the C3cropland and forests surrounding the C4prairie could be detected and had an impact on the carbon isotopic signal. The implication is that isotopic air sampling of CO2can be useful as a tracer for evaluating the fetch of upwind airflow in a heterogeneous ecosystem.