Globally, nitrogen (N) availability has increased, primarily through N fertilization and atmospheric deposition. Changes to fire disturbance regimes have also resulted in altered N dynamics, with fire suppression allowing for soil N accumulation. Soil microbial communities often respond to increased N availability with changes in composition, and with higher nitrification and denitrification potentials; thus microbial responses can promote loss of excess N from ecosystems. While increases in microbial N cycling following N addition are expected, it is less clear whether these processes remain elevated following the cessation of long-term fertilization. We investigated changes in soil nitrification and denitrification populations and potentials after cessation of a chronic fertilization experiment, expecting that, of these, N-reliant functional groups would recover more quickly than the whole soil microbial community. Soils were collected monthly from April-September 2017 from a 30-y plot-scale experiment that had manipulated available N through annual burning (or no burning) and 10 g N m-2 y-1 (NH4NO3) until 2016, when fertilization was ended. Subplots (1-m2) continued to receive N fertilizer to assess microbial recovery. Archaeal and bacterial amoA, clades I and II nosZ, and bacterial 16S rRNA gene populations, nitrification and denitrification potentials, and microbial biomass carbon were compared among all treatments.
Results/Conclusions
During the first growing season without N fertilization in 30 years, soil bacterial amoA gene population and nitrification potential (NP) declined to levels that were statistically indistinguishable from unfertilized soils, but burning did not facilitate the level of recovery. In contrast, soil archaeal amoA gene populations remained elevated in recovery plots. A fertilization history by month interaction on NP hinted that elevated net N mineralization might restrict complete recovery of soil NP. Denitrification potential (DNP) also decreased following cessation of fertilization, but remained elevated above unfertilized levels, and burning did not facilitate recovery. The nosZ clade I gene populations were not significantly different between fertilized, recovering, and unfertilized treatments, but the clade II gene population was lower in historically fertilized treatments than unfertilized soils. The bacterial 16S rRNA gene population, and microbial biomass carbon, was also highest in unfertilized soils, and 16S rRNA gene community composition mirrored this response. Thus, a portion of the heterotrophic microbial community appears to have declined with N fertilization, and not yet recovered. Soil nitrifiers appear more sensitive and resilient to fertilizer addition and cessation, but the lagged recovery of heterotrophic microbes could support sustained N loss potential even after cessation of chronic fertilization.