SOIL SPATIAL VARIABILITY AND SYMBIOTIC NITROGEN-FIXATION BY LEGUMES

The isotope dilution method for the estimation of N2 fixation by legumes is analyzed, comparing the application of 15N-enriched fertilizer with 15N-labeled soil. Soil variability of other dynamic processes in the soil are discussed in light of the distribution of the 15N label in the system. Field data were collected along six transects, 45 m long, with 30 plots (replicates) each. The legume (Vicia faba L.) was used as a fixing crop, barley (Hordeum vulgare L.) and oil radish (Raphinus sativus L.) as nonfixing standard crops. Isotope methods were also compared with the yield difference method. Results show that isotope methods are very sensitive to the distribution of the label in the soil and that dynamic processes involving N can significantly affect this distribution over a whole field. My best results were obtained with 15N-labeled soil. The particular site used, having been farmed for more than 20 years with 15N trials, showed a homogeneous residual 15N label that made it possible to estimate N2 fixation without the application of extra label. Estimates of N2 fixation with the isotope method were well correlated with the yield difference method when fertilizer use efficiency of the fixing and nonfixing crops were similar. My results also indicate that a good reference crop for one method might not be the best for the other method, and one reason for this is the variability of soil parameters and of dynamic processes occurring in the soil. © 1990 Williams and Wilkins.