Turnover of residual 15N-labelled fertilizer N in soil following harvest of oilseed rape (t Brassica napus L.)

We studied the fate of ¹⁵N-labelled fertilizer nitrogen in a sandy loam soil after harvest of winter oilseed rape (Brassica napus L. cv. Ceres) given 100 or 200 kg N ha^-1 in spring, with or without irrigation. Our main objective was to quantify the temporal variations of the soil mineral N, the extractable soil organic N and soil microbial biomass N, and fertilizer derived N in these pools during autumn and winter. Nitrogen use efficiency of the oilseed rape crop varied from 47% of applied N in the 100N, irrigated treatment to 34% in the 200N, non-irrigated treatment. However, only in the latter treatment did we find significantly higher fertilizer derived soil mineral N than in the three other treatments which all had low soil mineral N contents at the first sampling after harvest (8 days after stubble tillage). Between 31% and 42% of the applied N could not be accounted for in the harvested plants or 0-15 cm soil layer at this first sampling. Over the following autumn and winter none of the remaining fertilizer derived soil N was lost from the 0–5 cm depth, but from the 5–15 cm depth a marked proportion of N derived from fertilizer was lost, probably by leaching. Negligible amounts of fertilizer derived extractable soil organic and mineral N (<1 kg N ha^-1, 0-15 cm) were found in all treatments after the first sampling.Soil microbial biomass N was not significantly affected by treatments and showed only small temporal variability (±11% of the mean 76 kg N ha^-1, 0- 15 cm depth). Surprisingly, the average amount of soil microbial biomass N derived from fertilizer was significantly affected by the treatments, with the extremes being 5.5 and 3.1 kg N ha^-1 in the 200N, non-irrigated and 100N, irrigated treatments, respectively. Also, the estimated exponential decay rate of microbial biomass N derived from fertilizer, differed greatly (2 fold) between these two treatments, indicating highly different microbial turnover rates in spite of the similar total microbial biomass N values. In studies utilising ¹⁵N labelling to estimate turnover rates of different soil organic matter pools this finding is of great importance, because it may question the assumption that turnover rates are not affected by the insertion of the label.