Nitrous oxide release from arable soil:: importance of perennial forage crops

N2O emission rates from a sandy loam soil were measured in a field experiment with 2 years of perennial forage crops (ryegrass, ryegrass-red clover, red clover) and 1 year of spring barley cultivation, Spring barley was sown after the incorporation of the forage crop residues. All spring barley plots received 40 kg N ha(-1) N fertiliser. Ryegrass, ryegrass-red clover and red clover plots were fertilised with 350 kg N ha(-1), 175 kg N ha(-1) and 0 kg N ha(-1), respectively, From June 1994 to February 1997, N2O fluxes were continuously estimated using very large, closed soil cover boxes (5.76 m(2)). In order to compare the growing crops, the 33 months of investigation were separated into three vegetation periods (March-September) and three winter periods (October-February). All agronomic treatments (fertilisation, harvest and tillage) were carried out during the vegetation period. Large temporal changes were found in the N2O emission rates. The data were approximately log-normally distributed. Forty-seven percent of the annual N2O losses were observed to occur during winter, and mainly resulted from N2O production during daily thawing and freezing cycles. No relationship was found between the N2O emissions during the winter and the vegetation period. During the vegetation period, N2O losses and yields were significantly different between the three forage crops. The unfertilised clover plot produced the highest yields and the lowest N(2)Q losses on this soil compared to the highly fertilised ryegrass plot. Total N2O losses from soil under spring barley were higher than those from soil under the forage crops, this was mainly a consequence of N2O emissions after the incorporation of the forage crop residues.