A field experiment using 15N was carried out at the Indian Agricultural Research Institute, New Delhi during the crop year 1985–1986 to investigate the possible economy in fertilizer nitrogen applied to rice and succeeding wheat by growing a summer leguminous green manure (Sesbania aculeata) and incorporating it in soil before transplanting rice. Grain and straw yields of rice with 60 kg N/ha after green manuring were slightly greater than with 120 kg N/ha after summer fallow. Thus summer green manuring can lead to an economy of 60 kg N/ha in rice. 15N analysis showed that green manuring contributed 4·7–5·6% N of the total taken up by rice and 1·9–4·8% of that taken up by the succeeding wheat, depending upon the rate of fertilizer N applied to rice.

15N analysis showed that at 60 kg N/ha applied to rice, 35·4% was recovered by rice, 41% by the succeeding wheat and 16·7% was left in the 0–45 cm soil layer after summer fallow; after green manuring 32·8% N was recovered by rice, 3·0% by succeeding wheat and 22·6% was left in the soil. Thus 56·2 and 58·4% of applied N could be accounted for after summer fallow and green manuring, respectively. At 120 kg N/ha about 60–61% of applied N could be accounted for. Thus about 40–44% of N applied to rice was lost by the various loss mechanisms operating in rice fields.