Terrestrial nitrogen (N) availability continues to increase globally, due to chronic direct fertilizer application and atmospheric N deposition. Concurrently, land management changes affect vegetation growth and plant community composition, and consequently terrestrial carbon (C) cycling processes. In the context of these changes, it is imperative to learn (1) whether biological responses to N fertilization are similar under different land management scenarios and (2) whether chronic N fertilization changes the biological capacity for N removal upon cessation of fertilization. Our project addressed these two questions, with the general hypothesis that land management changes which increase soil C availability would also increase soil denitrification potential, and thus promote N loss during and following cessation of chronic N fertilization. We collected soils monthly over the 2017 growing season from a replicated 30-y plot-scale field experiment that crosses annual burning (or no burning) with 10 g N m-2 (ammonium nitrate) addition. During the sampling season, whole-experiment fertilization was ceased, and only sub-plots received fertilizer. We compared soil N mineralization, nitrification and denitrification potential rates, soil N concentrations, and soil microbial community composition and functional gene abundances among all treatments.
Results/Conclusions
Thirty years with no prescribed burning, and thirty years of fertilization, each significantly increased soil N mineralization, nitrification and denitrification rates (P < 0.05), with no significant interaction effects. Nitrification and denitrification potentials were lowest in April, highest during early summer (June and July) and intermediately high in late summer (August and September). During this first season of cessation of fertilizer addition, N mineralization and nitrification rates were lower in the recovering soils than the fertilized soils, but still higher than unfertilized soils (P < 0.05). In contrast, denitrification rates in recovering soils and fertilized soils were statistically indistinguishable from one another, and significantly higher than unfertilized soils. Estimated nosZ gene abundances were not related to soil denitrification potential activity, but amoA gene abundances were correlated with soil nitrification potential activity; and total soil microbial community composition was affected by N fertilization to a far greater extent than fire management. We hypothesise that lack of recovery of denitrification from chronic N fertilization is related to either shifts in the denitrifiying microbial community, or to differences in soil C availability (electron donor availability), either of which might support sustained denitrification potential.