Objectives Nitrous oxide (N₂O) in soil is a product of nitrification and denitrification, while manganese can react with the products of nitrification or denitrification to produce N₂O or N₂. Studies have shown that high levels of manganese can affect soil nitrification. Therefore, in this experiment the biochemical coupling reactions of manganese with soil nitrification and denitrification are considered as an entry point to study the effect of manganese addition on soil N₂O release rate and microorganisms, so as to provide a new perspective to understand the interaction between N₂O release and soil environmental factors.

Methods A microcosmic incubation was conducted, using red paddy soil as tested material. Birnessite was added by 0, 0.1%, 0.3%, 0.7%, 1.5% into the soils, and pre-incubated for 7 days. Ammonium sulfate was then added in rate of N 100 mg/kg in each well and incubated for another14 days. On the 1st, 3rd, 7th, and 14th day of incubation, gas sample of 10 mL was collected using gas-tight syringe from each treatment and the N₂O content was determined with gas chromatography. At the same time, soil samples were collected, the ammonium and nitrate nitrogen contents in the soil were determined by colorimetric method, the 16S rDNA and ammonia oxidizing bacteria (AOB) amoA gene copy numbers by real-time fluorescent quantitative PCR, and the composition and diversities of microbial communities were explored by high throughput sequencing technology.

Results The addition of 0.1% birnessite increased the release rate and the accumulative emission of N₂O from soil and also significantly increased soil pH and apparent N₂O yield (N₂O-N/NO₃⁻-N). In all treatments, the soil ammonium nitrogen decreased rapidly with the cultivation time, but the nitrate nitrogen content was reversed. Birnessite significantly increased the copy number of soil bacteria 16S rDNA and ammonia-oxidizing bacteria (AOB) amoA gene, and significantly increased the ratio of soil 16S rDNA to AOB amoA gene copy number. However, these promotion effects were not found when the addition rate of birnessite was higher than 0.1%. Actinomycetes, proteobacteria and bacteroides were the dominant phylums in all treatments. According to the result of non-metric multi-dimensional analysis, the microbial communities varied significantly among the treatments, with the largest discrepancy between the 0% treatment and the 1.5% treatment.

Conclusions Low addition percentage of birnessite could increase the N₂O release from the red paddy soil by increasing the amount of AOB, and affecting microbial species abundance and community structure. High addition rate did not show the effect as respected. Therefore, soil manganese content should be considered as the influencing factor of soil N₂O release.