Potential precision of the δ15N natural abundance method in field estimates of nitrogen fixation by crop and pasture legumes in south-west Australia
Precision of estimation of the proportion of legume N derived from N2 fixation (%Ndfa) was assessed in relation to subterranean clover (Trifolium subterraneum L.) pastures and crops of pea (Pisum sativum L.) and lupin (Lupinus angustifolius L.) under south-west Australian conditions. By using a standardized 10-point sampling procedure of paired sampling of legume and reference plant and reference plant 15N natural abundance (S15N) values in the range from +2.9 to +4. 0%o, %Ndfa of sample crops of lupin and field pea and a clover pasture were assessed with respective precisions of 93� O.6%, 76� 2.4% and 91�1.3% (� s.e., n = 10). Effects on S15N due to isotope discrimination during fixation and subsequent distribution of N by the three study legumes were studied using sand-cultured, fully symbiotic plant material. The resulting S15N data (B values) showed consistently more negative values for shoots than roots (all species), no significant effects of cultivar on B values (all species), a marked effect of rhizobial strain on B value (subclover) and a tendency for B values to fall with plant age (pea and lupin). The likely magnitude of errors in %Ndfa estimates due to incorrect choice of B value was indicated. By using data for reference plant S15N values from field surveys and previously assessed error factors in mass spectrometric measurement of S15N, precision of estimation of %Ndfa by using bulked material from the 10-point field sampling procedure was predicted for situations ranging from where a legume was obtaining only minimal amounts (10%) through to the bulk (90%) of its N by atmospheric fixation.