Nitrogen fixation in chickpea. II. Comparison of 15N enrichment and 15N natural abundance methods for estimating nitrogen fixation

The 15N enrichment and 15N natural abundance methods for estimating N2 fixation in chickpea were compared over a range of soil NO3-N levels at crop establishment varying from 10 to 326 kg N/ha (0-120 cm depth). Barley was used as a non-N2 fixing control crop. Both methods estimated reduced N2 fixation as soil NO3-N levels at crop establishment increased. Similar estimates of % N2 fixation were obtained at high values, but at low values the enrichment method gave lower estimates, some of which were negative. The 15N natural abundance method provided realistic estimates of % N2 fixation across all soil N03-N levels at crop establishment. An asymptotic curve described a close ( R2 = 0.95) relationship between these factors. Standard errors of estimates of means for the 15N natural abundance method remained acceptable and relatively stable over the full range of measurements; however, with the 15N enrichment method they became unacceptably large at low values of % N2 fixation. These large errors may have been partly due to legume and control plants assimilating mineral N of differing 15N enrichment. High mineral N levels associated with low values of % N2 fixation were also shown to reduce reliability of N2 fixation values estimated by the 15N enrichment method. These errors caused potentially greater inaccuracy at low values of % N2 fixation than at high values. To compare N2 fixation means statistically, transformations were necessary to stabilize variance and to impart lower weightings to plots with low values of % N2 fixation.