Soil moisture drives the response of soil microbial nutrient limitation to N and P additions in an Inner Mongolian meadow steppe

The metabolic activity of soil microorganisms is often limited by soil nutrient availability. Fertilization can increase available nutrient content, but nutrient limitations may persist because of imbalances in nutrient inputs. However, the mechanisms driving the response of soil microbial nutrient limitation to N and P application in grasslands remain unclear. To address this issue, we applied fifteen fertilization treatments composed of five N levels (0, 1.55, 4.65, 13.95, 27.9 g N m(-2) y(r-1)) and three P levels (0, 5.24, 10.48 g P m(-2) yr(-1)) to a meadow steppe in Inner Mongolia across three years using a split-plot experiment design. Soil microbial biomass and extracellular enzyme activities were analyzed in samples collected from each plot in May, July, and August. The addition of N significantly reduced microbial biomass carbon (MBC) in samples collected in May and increased microbial biomass nitrogen (MBN) in July samples, thus decreasing the ratio of MBC:MBN in both months. P addition significantly increased microbial biomass phosphorus (MBP), whereas it reduced the ratio of MBC:MBP and MBN:MBP. Using vector analysis, we found that vector angle was less than 45 degrees across all sampling dates, indicating that soil microbial metabolism was predominately limited by N rather than P. The severity of microbial N limitation was attenuated by N addition, but was worsened by P addition in May and July. The severity of microbial C limitation was significantly intensified by N addition in May and July, and forced by P addition in July and August. Visual partitioning analysis showed that soil physicochemical and microbial properties explained 37% and 70% of variation in microbial C and N limitation, respectively. Besides soil available nutrient concentrations, soil water content (SWC) and pH were identified as the key factors driving microbial C and N limitations. The relative influence of SWC on microbial N limitation was highest across all sampling dates. According to PLS-SM modeling, SWC had a total effect of -0.349 on microbial N limitation, which was significantly higher than the effects than N addition (-0.192) and P addition (0.131). Overall, this study indicates that soil moisture was the primary control over the response of microbial nutrient limitation to N and P additions in a meadow steppe in Inner Mongolia.