Precipitation affects soil nitrogen fixation by regulating active diazotrophs and nitrate nitrogen in an alpine grassland of Qinghai-Tibetan Plateau.

Soil asymbiotic nitrogen (N) fixation provides a critical N source to support plant growth in alpine grasslands, and precipitation change is expected to lead to shifts in soil asymbiotic N fixation. However, large gaps remain in understanding the response of soil asymbiotic N fixation to precipitation gradients. Here we simulated five precipitation gradients (10 % (0.1P), 50 % (0.5P), 70 % (0.7P), 100 % (1.0P) and 150 % (1.5P) of the natural precipitation) in an alpine grassland of Qinghai-Tibetan Plateau and examined the soil nitrogenase activity and N fixation rate for each gradient. Quantitative PCR and high-throughput sequencing were used to measure the abundance and community composition of the soil nifH DNA (total diazotrophs) and nifH RNA reverse transcription (active diazotrophs) gene. Our results showed that the soil diazotrophic abundance, diversity and nifH gene expression rate peaked under the 0.5P. Soil nitrogenase activity and N fixation rate varied in the range 0.032-0.073 nmol·CH·g·h and 0.008-0.022 nmol·N·g·h respectively, being highest under the 0.5P. The 50 % precipitation reduction enhanced the gene expression rates of Azospirillum and Halorhodospira which were likely responsible for the high N fixation potential. The 0.5P treatment also possessed a larger and more complex active diazotrophic network than the other treatments, which facilitated the resistance of diazotrophic community to environmental stress and thus maintained a high N fixation potential. The active diazotrophic abundance had the largest positive effect on soil N fixation, while nitrate nitrogen had the largest negative effect. Together, our study suggested that appropriate precipitation reduction can enhance soil N fixation through promoting the abundance of the soil active diazotrophs and decreasing soil nitrate nitrogen, and soil active diazotrophs and nitrate nitrogen should be considered in predicting soil N inputs in the alpine grassland of Qinghai-Tibetan Plateau under precipitation change.