15N abundance of soils and plants along an experimentally induced forest nitrogen supply gradient.

15N abundances of soils and a grass species (Deschampsia flexuosa (L.) Trin.) were analysed in a forest fertilization experiment 10 years after the last fertilization. Nitrogen had been given as urea, at seven doses, ranging from 0 to 2400 kg N ha-1. Previously, we have shown that plants in systems experiencing large losses of N become enriched with 15N. This was explained by the fact that processes leading to loss of N, e.g. ammonia volatilization, nitrification followed by leaching or denitrification and denitrification itself, tend to fractionate against 15N. In this experiment, 15N abundance increased with dose of N applied in both grass and soil total-N, but more so in the grass. This was interpreted to be due to the grass sampling small but active pools of N subject to losses. In contrast, soil total-N largely consists of inactive N that does not immediately exchange with pools of N from which fractionating losses occur. Hence, soil total-N shows a large pretreatment 15N memory effect, and is, therefore, and integrator of the long-term N balance. When short-term changes (years, decades) in N balances are monitored using variations in 15N abundance, plants are more suitable indicators of such change than is soil total-N.