Field application of the N-15 isotope dilution technique for the reliable quantification of plant-associated biological nitrogen fixation

To apply the isotope dilution (ID) technique, it is necessary to grow the ''N-2-fixing'' crop in a soil where the mineral N is labelled with N-15. Normally the ''N-2-fixing'' crop and a suitable non-N-2-fixing control crop are grown in the same labelled soil and the N-15 enrichment of the control crop is assumed to be equal to the N-15 enrichment of the nitrogen (N) derived from the soil in the ''N-2-fixing'' crop. In this case the proportion of unlabelled N being derived from the air via biological N-2 fixation (BNF) in the ''N-2-fixing'' crop will be proportional to the dilution of the enrichment of the N derived from the labelled soil. To label the soil, the technique most often used is to add a single addition of N-15-labelled N fertilizer shortly before, at, or shortly after, the planting of the crops. Data in the literature clearly show that this technique results in a rapid fall in the N-15 enrichment of soil mineral N with time. Under these conditions, if the control and the ''N-2-fixing'' crops have different patterns of N uptake from the soil they will inevitably obtain different N-15 enrichments in the soil-derived N. In this case the isotope dilution technique cannot be applied, or if it is, there will be an error introduced into, the estimate of the contribution of N derived from BNF. Several experiments are described which explore different strategies of application of the ID technique to attempt to attenuate the errors involved. The results suggest that it is wise to use slow-release forms of labelled N, or in some cases, multiple additions, to diminish temporal changes in the N-15 enrichment of soil mineral N. The use of several control crops produces a range of different estimates of the BNF contributions to the ''N-2-fixing'' crops, and the extent of this range gives a measure of the accuracy of the estimates. Likewise the use of more than one N-15 enrichment technique in the same experiment will also give a range of estimates which can be treated similarly. The potential of other techniques, such as sequential harvesting of both control and test crops, are also discussed.