Spatial variability of symbiotic N2 fixation in grass-white clover pastures estimated by the 15N isotope dilution method and the natural 15N abundance method

Aiming at estimating the spatial variability in N-2 fixation, and to evaluate the appropriateness of the N-15 isotope dilution (ID) method and the natural N-15 abundance (NA) method in reflecting spatial variability under the influence of cattle grazing, the symbiotic N-2 fixation in grass-white clover mixture was studied. At the Foulum site, where the ID method was used, differences in the climatic conditions between the two years of investigations caused a considerable difference in plant growth rates and proportion of clover. Consequently, the total N-2 fixation in ungrazed reference plots was significantly less in 1998 than in 1997, being 5.9 and 12.5 g N m(-2), respectively. In both years there was a wide range in concentration of inorganic N in the soil with coefficients of variance of approximately 60-190% for ammonium and 70-340% for nitrate. Significant negative correlations between pNdfa, determined by the ID method, and the log-transformed values of inorganic N and total N in grass were found. The NA method was applied on three nearby commercial dairy farms. They also showed high coefficients of variation. The coefficient of variance for NO3--N ranged from 37 to 282% and for NH4+-N from 29 to 237%. Average estimates of pNdfa values, which in the NA method were calculated using apparent B values ranging from -2.10 to -2.59 parts per thousand, were generally lower (0.7-0.87) for these farms than for the Foulum site (0.89-0.95) using the ID method. For the NA method the delta N-15 values, i.e. deviation in N-15 concentration from atmospheric N-2, ranged from -7.0 to 5.7 parts per thousand for the grass N, which in several cases was lower than for clover N. Due to this high variability of the delta N-15 values, probably caused by deposition and plant assimilation of N-15 depleted urinary N in the pastures, the NA method was marginal for accurate determination of pNdfa. Consequently no significant correlation between the pNdfa determined by this method, and the log-transformed values of inorganic N in soil or total N in grass were found.