15N abundances of current needles of Norway spruce collected during 23 yrs of a forest fertilization experiment were studied in order to follow ecosystem gains and losses of N. Unlabelled ammonium nitrate at four rates (N0–N3), phosphorus at three rates (P0–P2), and potassium plus other elements including micronutrients at two rates (K0–K1), had been applied to plots in a complete factorial design. Nitrogen had been applied annually at average rates of 0, 34, 68 and 102 kg N ha-1 yr-1. Tree growth had responded positively to additions of N, but the response was remarkably more positive to the N2P2K1 treatment.
In N1 treatments, δ15N (‰) declined over time. This was consistent with an earlier study, and should reflect a change in 15N abundance towards that of fertilizer N (minus discrimination during uptake), which in turn means accretion of most of the N added. As in the earlier study, in which N3 plots lost most of the N added, the present N3 plots showed an increasing δ15N (‰). This pattern was not significantly affected by additions of P and K plus other elements, although a weak negative effect of P on N accretion was indicated, i.e. there was a tendency δ15N (‰) to be higher when P was added. This, and another recent result based on an N budget, shows that so-called revitalization fertilization may well increase growth of trees, but also promotes losses of N from the ecosystem. As in the previous study, a decline in δ15N (‰) on control plots provided evidence of contamination. Given a removal of 100 kg N ha-1 at stem harvest and a leaching of 2 kg N ha-1 yr-1, our data on 15N suggest that a load of 9 kg N ha-1 yr-1 would saturate the ecosystem after 100 years. This load is only about twice the annual deposition at the site.