EVIDENCE AGAINST ASSOCIATIVE N-2 FIXATION AS A SIGNIFICANT N SOURCE IN LONG-TERM WHEAT PLOTS

In monocropped cereal systems, annual N inputs from non-fertilizer sou rces may be more than 30 kg ha(-1) We examined the possibility that th ese inputs are due to biological N-2 fixation (BNF) associated with ro ots or decomposing residues. Wheat was grown under greenhouse conditio ns in pots (34 cm long by 10 cm diameter) containing soil from a plot cropped to spring wheat since 1911 without fertilization. The roots an d soil were sealed from the atmosphere and exposed to a N-15(2)-enrich ed atmosphere for three to four weeks during vegetative, reproductive or post-reproductive stages. This technique permitted detection of as little as 1 mu g fixed N plant(-1) in plant material and 40 mu g fixed N plant(-1) in soil. No fixation of N-15(2) occurred during either of the first two labelling periods. In the final labelling period, straw returned to the soil was significantly enriched in N-15, especially i n a pot with a higher soil moisture content. Total BNF in this pot was 13 mu g N plant(-1), or about 30 g N ha(-1). In a separate experiment with soil from the same plot, we detected BNF only when soil was amen ded with glucose at a high soil moisture content. Measured associative BNF was insufficient to account for observed N gains under field cond itions.