Long-term effects of pre-harvest burning and nitrogen and vinasse applications on yield of sugar cane and soil carbon and nitrogen stocks on a plantation in Pernambuco, NE Brazil

Since the 1970s the area under sugarcane in Brazil has increased from 2 million to over 5 million ha (M ha), and it is expected to pass the 7 M ha mark in 2007. More than half of the cane is harvested to produce bioethanol as a fuel for light vehicles. The distilleries produce approximately 13 L of distillery waste (vinasse) for each litre of ethanol produced. In the 1980s there was considerable concern over the long-term effects of the disposal of this material (containing about 1% carbon and high in K) on cane yields if it was applied to the field. At the same time there was a growing movement to abandon the practice of pre-harvest burning and some research was showing that some Brazilian varieties of sugar cane were able to obtain significant contributions of N from plant-associated biological nitrogen fixation (BNF). For these reasons an experiment was installed on a cane plantation in the state of Pernambuco, NE Brazil to investigate the long-term effects of vinasse and N fertiliser additions and the practice of pre-harvest burning on crop and sugar yield, soil fertility parameters, N balance and soil C stocks. The results showed that over a 16-year period, trash conservation (abandonment of burning) increased cane yields by 25% from a mean of 46 to 58 Mg ha(-1). Vinasse applications (80 m(3) ha(-1) crop(-1)) increased mean cane and sugar yield by 12 to 13% and the application of 80 kg N ha(-1) as urea increased cane yields by 9%, but total sugar yield by less than 6% (from 7.0 to 7.4 Mg ha(-1) crop(-1)). The total N balance for the soil/plant system when only the surface 20 cm of the soil was considered was positive in plots where no N fertiliser was added. However, the data indicated that during the 16 years of the study considerable quantities of soil organic matter were accumulated below 20 cm depth such that the N balance considering the soil to 60 cm depth was strongly positive, except where N fertiliser was added. The data indicated that there were considerable BNF inputs to the system, which was consistent with its low response to N fertiliser and low N fertiliser-use-efficiency. There were no significant effects of vinasse or urea addition, or trash conservation on soil C stocks, although the higher yields proportioned by trash conservation had potentially significant benefits for increased mitigation of CO2 emissions where the main use of the cane was for bioethanol production.