Direct pathway of nitrate produced from surplus nitrogen inputs to the hydrosphere.

In PNAS, Sebilo et al. (1) present the fate of labeled nitrate in the soil–plant system over three decades and attempt to determine “to which extent and over which time periods fertilizer N stored in soil organic matter is rereleased for either uptake in crops or export into the hydrosphere.” I suspect that this mechanism of immobilization-mineralization turnover described by Sebilo et al. (1) is valid only in cases when the external N supply has been low or in balance within the soil–plant system. In another case when the external N supply is in surplus the retention capacity, the soil matrix would have excess nitrate either directly derived from NO3 -based fertilizer or by nitrification of NH4 -based fertilizer (2, 3), possibly with even larger amounts exported rapidly to the hydrosphere. The extremely high external N inputs have led to severe groundwater nitrate pollution in China within only one or two decades, resulting from the direct diffusion of fertilizer-derived nitrate (2, 4). To clarify, we can define N retention capacity as the soil N saturation point at which net mineral N production is nearly zero during a given period; that is, the amount of N immobilization by plant root, soil microbial assimilation, and soil organic matter or minerals is more than or almost equal to the amount of external N input, in which the soil matrix would not have additional nitrate directly derived from external N inputs. We can further divide the process into roughly three stages. First, the former is much larger than the latter under very low external N inputs, and the small residual nitrate in the soil matrix and the plant will deplete the soil N pool and the soil cannot support goodplant growth. Second, the former is nearly equal to the latter under moderate external N inputs and maintenance of soil N balance, with no additional residual nitrate in the soil matrix but sufficient plant available N to support target crop yields. Third, the former is far smaller than the latter under high external N inputs, with much more residual nitrate in the soil matrix and with anN supply that is surplus to plant requirements and consequently much greater N loss to the environment (5). We can regard the three N status levels above as tight, balanced, and surplus N cycling in the soil–plant system. I am left with unanswered questions from the report by Sebilo et al. (1): (i) How do we sample soil organic matter annually? If soil is only sampled from the top 28 cm, how do we calculate the annual N trace retention? If soil samples were taken annually from 200 cm, this might destroy the water flow distribution. (ii) The y axis in figure 2 of ref. 1 should be “% of the initial N input.” (iii) If the nitrate in seepage water is derived from remineralized N-labeled soil organic N, then its δN(‰) should be close to the δN(‰) of soil organic N, but the former is much higher than the latter.