In the experiment through sod-podzolic soil, nitrogen flows were estimated in agrocoenosis when applied under spring wheat using 15N ammonium nitrate labeled with isotope and white mustard biomass as a siderate. It was established that the increase of grain yield from the use of N45 is 42%, the siderate – 58%, their joint introduction – 99%, from seed inoculation with the Rizoagrin biopreparation – 16%. 60 - 80% of spring wheat harvest is formed due to soil nitrogen, the share of "extra" nitrogen reaches 14%. Plants use 41% of ammonia nitrate nitrogen for crop formation, and 23% of the siderate. Seed inoculation with a biological product increases plant nitrogen intake of mineral fertilizer by 5 - 9%. 26% of ammonium nitrate nitrogen and 33% of siderate nitrogen are fixed in the soil, the biological product does not affect this process. Unaccounted losses of 15N mineral fertilizer make up 32% of the applied amount and reduced by 8% when adding ammonium nitrate and inoculating seeds. The loss of labeled siderate nitrogen is 25% of the amount deposited, the use of the biological product increases the nitrogen consumption of N-fertilizers and provides a positive tendency for green manure nitrogen to fix in the soil and reduces its loss by 4%. On sod-podzolic light loamy soil, agroecosystem functions in the homeostasis mode (norm) when using green manure, in the stress mode (permissible) – when sharing green manure and mineral nitrogen fertilizer. When using only ammonium nitrate, agroecosystem operates in resistance mode. Differences in the performance of agroecosystems among variants with the inoculation of spring wheat with Rizoagrin and without it have not been established.