A model to predict the accuracy of measurements of legume N rhizodeposition using a split-root technique

Legume N rhizodeposition is an important process for understanding the N turnover in legume-based cropping systems. Different 15N labeling techniques have been developed to estimate the rhizodeposition of legume-derived N into the soil. However, it is not known how the 15N-based experiments have to be designed to achieve a defined degree of accuracy in measuring the amount of N derived from rhizodeposition (Ndfr). As a consequence therefore, a model for the split-root technique was developed on the basis of experimental data to (i) test the effects of various experimental conditions on soil 15N enrichment, (ii) evaluate the accuracy of the measurements, and to (iii) deduce instructions for designing efficient experiments using 15N techniques for quantifying legume N rhizodeposition. It turned out that the coefficient of disproportional 15N distribution between non-15N-fed roots grown in soil (RootSoil) and total plant biomass (D coefficient) is an indispensable component in the development of such a model for 15N-based split-root experiments. The model showed the coefficients of variation in measuring the Ndfr with regard to the analytical accuracy in determining not only the isotope composition of both the RootSoil and the soil, but also the N content of the soil itself. Suggestions for designing specific experimental conditions to achieve a high accuracy in quantifying Ndfr were deduced from the model for the split-root technique, particularly in the choice of the amount of soil N at the start of the experiment and the 15N enrichment of the fertilizer being used. The coefficients of variation in measuring Ndfr are presented regarding the 15N abundance of RootSoil and the quotient of the amount of Ndfr and soil N at the start of the experiment.