Increasing CO2 from subambient to elevated concentrations increases grassland respiration per unit of net carbon fixation

AbstractRespiration (carbon efflux) by terrestrial ecosystems is a major component of the global carbon (C) cycle, but the response of C efflux to atmospheric CO2 enrichment remains uncertain. Respiration may respond directly to an increase in the availability of C substrates at high CO2, but also may be affected indirectly by a CO2‐mediated alteration in the amount by which respiration changes per unit of change in temperature or C uptake (sensitivity of respiration to temperature or C uptake). We measured CO2 fluxes continuously during the final 2 years of a 4‐year experiment on C3/C4 grassland that was exposed to a 200–560 μmol mol−1 CO2 gradient. Flux measurements were used to determine whether CO2 treatment affected nighttime respiration rates and the response of ecosystem respiration to seasonal changes in net C uptake and air temperature. Increasing CO2 from subambient to elevated concentrations stimulated grassland respiration at night by increasing the net amount of C fixed during daylight and by increasing either the sensitivity of C efflux to daily changes in C fixation or the respiration rate in the absence of C uptake (basal ecosystem respiration rate). These latter two changes contributed to a 30–47% increase in the ratio of nighttime respiration to daytime net C influx as CO2 increased from subamient to elevated concentrations. Daily changes in net C uptake were highly correlated with variation in temperature, meaning that the shared contribution of C uptake and temperature in explaining variance in respiration rates was large. Statistically controlling for collinearity between temperature and C uptake reduced the effect of a given change in C influx on respiration. Conversely, CO2 treatment did not affect the response of grassland respiration to seasonal variation in temperature. Elevating CO2 concentration increased grassland respiration rates by increasing both net C input and respiration per unit of C input. A better understanding of how C efflux varies with substrate supply thus may be required to accurately assess the C balance of terrestrial ecosystems.