Exploring the boundaries of N2-fixation in cereals and grasses:: An hypothetical and experimental framework

Despite more than 40 years of research on free-living and endophytic bacteria associated with cereals and grasses, conclusive examples of impacts of non-symbiotic N-2-fixation in agriculture are lacking. All available methods for measurement of N-2-fixation associated with cereals and grasses have been employed, and N-2-fixation has been demonstrated to occur under controlled conditions, but this is insufficient evidence to prove a significant role for N-2-fixation by heterotrophic bacteria in the field. Recently attention has focused on endophytic N-2-fixation with claims of major inputs of N from the atmosphere, particularly in sugar cane for which the largest body of information exists, although some methodological concerns remain unanswered. Here we analyse the evidence for N-2-fixation in the rhizosphere and within graminaceous plants. We provide guidelines suggesting the type of information required to allow a critical analysis of the amounts of non-symbiotic N-2-fixation contributed in agriculture. Given the substantial availability of C as a substrate for N-2-fixation in sugar cane it remains the most likely candidate plant to benefit from non-symbiotic N-2-fixation. Measurements of elevated respiration in zones of high bacterial colonization may give an indication of whether these really are focal points of the rapid metabolic activity required to support significant N-2-fixation activity. The only conclusive evidence for a major role of non-symbiotic N-2-fixation in the field will come from long-term N balance studies in which ALL processes of potential gain and loss are measured. We conclude that evidence for a large contribution from heterotrophic, non-symbiotic N-2-fixation in both natural and agricultural systems is weak. Although amounts of non-symbiotic N-2-fixation in natural systems may be small (<5 kg N ha(-1)) they could be significant in the long term over many years.