The terrestrial biosphere regulates atmospheric CO₂ concentrations via photosynthesis (GPP) and ecosystem respiration (Re) and is strongly sensitive to climate variability and changes in resource supply. There is still considerable uncertainty over how GPP, RE and the net carbon balance of ecosystems (net ecosystem exchange; NEE) will be affected by interacting global change factors.

The Prairie Heating and CO₂ Enrichment (PHACE) experiment evaluates the interactive effects of rising CO₂ (present ambient [388 ppm], and elevated [600 ppm] CO₂) and elevated temperature ([1.5/3.0 ºC warmer day/night]) (n=5) on the ecology of the semi-arid northern mixed grass prairie near Cheyenne, WY, USA. An irrigation treatment is also included in the experimental design to help characterize potential CO₂ effects on ecosystem processes as direct effects (due to photosynthetic stimulation) versus indirect effects (due to enhanced soil water availability).  Canopy gas-exchange measurements were conducted using an open path infra-red gas analyzer over diurnal time scales throughout three growing seasons to quantify NEE, Re and GPP (by difference), and the C flux responses to soil moisture and temperature.

Results/Conclusions

Averaged over three growing seasons, all treatments lost C to the atmosphere, and losses from the elevated CO₂ plus warming treatment were double those from ambient (150 vs. 75 g C m^-2 y^-1).  GPP increased under elevated CO₂ during two of three experimental years, consistent with expectations from modeling studies and other experiments. However, when combined with warming, elevated CO₂ did not stimulate GPP, suggesting that the effects of CO₂ on production were mediated by water availability and counteracted by desiccation caused by warming. Comparisons with the irrigated treatment also suggest that greater water use efficiency and higher soil moisture largely drive enhanced GPP under elevated CO₂ at ambient temperature. Elevated CO₂ consistently stimulated Re at both ambient and elevated temperatures, while warming alone suppressed Re. We speculate that priming of older soil organic matter may be responsible for a portion of the C lost under warming plus elevated CO₂, and additional measurements are underway to address this possibility. The consistent net C losses under elevated CO₂ plus warming run counter to expectations from model predictions for the PHACE experiment and for grasslands generally, demonstrating critical gaps in modeling interactive effects of climate change.