Background/Question/Methods
Climate change is likely to play a major role in the carbon balance of temperate forests in the future, although the sign and magnitude of the resulting feedbacks to the climate system are uncertain. A long-term soil warming experiment at the Harvard Forest in central Massachusetts aims to examine the responses of northern temperate forests to these changing climate conditions. The field experiment consists of two 30 x 30 m plots (control and heated) in a deciduous forest, where soil temperatures on the heated plot are maintained at 5°C above ambient. The experiment has been operating for eight years, including one year of pretreatment data collection. Soil respiration and dendrology sampling occur once a month from April to November. Carbon dioxide fluxes are measured using infrared gas analysis and static field chambers. Carbon storage in trees is allometrically determined from dendrometer band measurements. Fine root (<1mm) respiration was measured once a month during the growing season in 2008 and 2009. Changes in root biomass in response to heating were estimated from samples taken in 2007 and 2008. 

Results/Conclusions

In the first seven years of warming, an average net source of 920 kg C ha^-1 yr^-1 was lost from the ecosystem. This was driven mainly by increases in soil respiration, which released an average of 1850 kg C ha^-1 yr^-1. Warming also increased nitrogen availability, leading to increases in vegetation C storage of 930 kg C ha^-1 yr^-1, which is equivalent to 50% of soil organic matter loss. These results suggest that vegetation carbon storage can significantly offset carbon losses from soil in a warmer world.

While overall, the average carbon balance over seven years reflects a net source of CO₂ to the atmosphere, the effect of soil warming on C storage has changed over time. Warming caused a net source of C in the first six years of warming, ranging between 575 and 1770 kg C ha^-1 yr^-1 relative to the controls.  In the seventh year, however, warming resulted in a net ecosystem carbon sink of 670 kg C ha^-1. Microbial respiration spiked initially, but has decreased over time in this experiment, while vegetation C storage has steadily increased. This trend is evidence that the carbon and nitrogen cycles in soil have been de-coupled; microbes have exhausted the labile carbon pool while trees continue to thrive on increased nitrogen availability and a slightly elongated growing season.