Abstract: Soil microbial biomass is critical for biogeochemical cycling and serves as precursor for carbon (C) sequestration. The anthropogenic nitrogen (N) input has profoundly changed the pool of soil microbial biomass. However, traditional N deposition simulation experiments have been exclusively conducted through infrequent N addition, which may have caused biased effects on soil microbial biomass compared with those under the natural and continuous N deposition. Convincing data are still scarce about how the different N addition frequencies affect soil microbial biomass. By independently manipulating the frequencies (2 times vs. 12 times N addition yr^–1) and the rates (0–50 g N m⁻² yr⁻¹) of N addition, our study aimed to examine the response of soil microbial biomass C (MBC) to different N addition frequencies with increasing N addition rates. Soil MBC gradually decreased with increasing N addition rates under both N addition frequencies, while the soil MBC decreased more at low frequency of N addition, suggesting that traditional studies have possibly overestimated the effects of N deposition on soil microbial biomass. The greater soil microbial biomass loss with low N frequency resulted from the intensified soil acidification, higher soil inorganic N, stronger soil C and N imbalance, less net primary production allocated to belowground and lower fungi to bacteria ratio. To reliably predict the effects of atmospheric N deposition on soil microbial functioning and C cycling of grassland ecosystems in future studies, it is necessary to employ both the dosage and the frequency of N addition.

Key words: nitrogen addition frequency, temperate grasslands, nutrient imbalance, fungi to bacteria ratio, plant–microbe interaction, belowground net primary production, Shannon–Weiner diversity index