中文摘要：

采用盆栽实验，第1季水稻生长期内施入^15N标记硫酸铵，在以后的各季作物生长期内使用非标记硫酸铵，连续5季实施稻麦轮作，在每一季结束后采样测定^15N标记硫酸铵的去向，并利用土壤中^15N残留量数据计算了不同施肥强度下土壤氮被替换的速率，以此反映人们对土壤氮素干预的程度。计算结果是：假设肥料氮一旦进入土壤就看成是土壤氮，那么肥料氮替换50％的土壤氮因施肥量不同需要7～60a（14～121季）；如果将作物成熟以后残留于土壤的肥料氮看成是土壤氮，那么替换20％的土壤氮因施肥量不同需要3～7a（5～15季），替换30％的土壤氮需要无限长时间。

英文摘要：

A pot experiment was conducted to study fate of ^15N-labelled ammonium sulfate incorporated into soil under the rice-wheat rotation system for 5 seasons of cropping. Residue fertilizer N from previous crops would accumulate as the cropping and fertilization practices continued. Assuming that the cropping system and soil management did not change and the total soil nitrogen content remained constant, the replacement rates of soll nitrogen by fertilizer nitrogen could be calculated based on change in residualnitrogen in the soil. By fitting the data of residualnitrogen in the soil into different mathematic equations, a best fitting equation could be picked out in light of their correlation efficiency （ R^2） .The equation could be used conveniently to figure out the replacement rate, which is really high, accounting for 20% within only 3 ～ 5 years. The calculated replacement rate, however, could not be extrapolated into field conditions anywhere and the calculation methods presented in this paper is only applicable to similar cases where the assumptions hold water to a certain extent. The significance of replacement rate was discussed.