Resistant Soil Microbial Communities Show Signs of Increasing Phosphorus Limitation in Two Temperate Forests After Long-Term Nitrogen Addition

Abstract

Forest soils harbor diverse microbial communities that are responsible for the cycling of elements including carbon (C), nitrogen (N) and phosphorus (P). Conversely, anthropogenic N deposition can negatively feedback on soil microbes and reduce soil organic matter (SOM) decomposition. Mechanistically, this includes reductions of decomposer biomass, especially fungi, and decreases in activities of lignin-modifying enzyme (LMEs). Moreover, N inputs can decrease the C:N imbalance between microbial decomposers and their resources by lowering resource C:N, resulting in slowed microbially-mediated decomposition and larger SOM pools. Here, we studied the long-term impact of N addition on soil microbes and associated decomposition processes along the topsoil profile in two temperate coniferous forests in Switzerland and Denmark. We measured microbial biomass C and N, phospholipid fatty acid (PLFA) biomarkers and potential enzyme activities. In particular, we investigated shifts in community level homeostasis and relative elemental limitation after two decades of N addition. Contrary to prevailing theory, microbial biomass and community composition were remarkably resistant against twenty years of 780 and 1280 kg ha-1 of cumulative N inputs at the Swiss and Danish site, respectively. While N reduced fungal-specific PLFAs and lowered fungi:bacteria ratios in some horizons, it increased the fungi:bacteria ratio in other horizons. We did not find a consistent reduction of lignin-modifying enzymes (LMEs). This questions prevalent theories of responses of lignin decomposition and SOC storage to elevated N inputs. We further showed that microbial communities responded in part non-homeostatically to decreasing resource C:N, likely through adaptations in microbial elemental use efficiencies. In contrast, the expected increased allocation to C- and decreased allocation to N-acquisition enzymes was not found. Microbial investment into P acquisition (acid phosphatase activity) increased in nutrient-poor Podzols (but not in nutrient-rich Gleysols), while enzyme vector analysis showed decreasing C but increasing P limitation of soil microbial communities at both sites. We conclude that simulated N deposition in two independent, long-term experiments led to physiological adaptations of soil microbial communities with implications for tree nutrition and SOC sequestration. However, we expect that microbial adaptations are not endless and may reach a tipping point when ecosystems experience nitrogen saturation.