Soil tillage and other methods of soil management may influence CO2 emissions because they accelerate the mineralization of organic carbon in the soil. This study aimed to quantify the CO2 emissions under conventional tillage (CT), minimum tillage (MT) and reduced tillage (RT) during the renovation of sugarcane fields in southern Brazil. The experiment was performed on an Oxisol in the sugarcane-planting area with mechanical harvesting. An undisturbed or no-till (NT) plot was left as a control treatment. The CO2 emissions results indicated a significant interaction (p < 0.001) between tillage method and time after tillage. By quantifying the accumulated emissions over the 44 days after soil tillage, we observed that tillage-induced emissions were higher after the CT system than the RT and MT systems, reaching 350.09 g m(-2) of CO2 in CT, and 51.7 and 5.5 g m(-2) of CO2 in RT and MT respectively. The amount of C lost in the form of CO2 due to soil tillage practices was significant and comparable to the estimated value of potential annual C accumulation resulting from changes in the harvesting system in Brazil from burning of plant residues to the adoption of green cane harvesting. The CO2 emissions in the CT system could respond to a loss of 80% of the potential soil C accumulated over one year as result of the adoption of mechanized sugarcane harvesting. Meanwhile, soil tillage during the renewal of the sugar plantation using RT and MT methods would result in low impact, with losses of 12% and 2% of the C that could potentially be accumulated during a one year period.
