Effects of different fertilizer application methods on the community of nitrifiers and denitrifiers in a paddy soil

PURPOSE: Nitrifiers and denitrifiers are the key drivers of N cycling in paddy soil. Little is known about the effects of different fertilization methods, especially side bar fertilization, on the community of nitrifiers and denitrifiers in paddy soils. Here we assess the relationships between soil physicochemical properties, denitrification and nitrification activities, and the underlying microbial communities in a surface layer of paddy field soil treated with different fertilization methods. MATERIALS AND METHODS: Soil was unfertilized (control), treated with conventional chemical fertilizer (CF), CF plus pig manure (MC), or slow-release fertilizer (SR), or by slow-release side bar fertilization (SB). Soil was sampled after one season of early and late rice growth. We determined soil physicochemical properties, potential nitrification rates (PNR), and denitrification enzyme activities (DEA). Ammonia-oxidizing archaeal (AOA) and bacterial (AOB) communities were assessed via their ammonia monooxygenase (amoA) genes, and denitrifiers via nitrite reductases (nirK and nirS) and nitrous oxide reductase (nosZ). Quantitative PCR was used to assess gene abundance, terminal restriction fragment polymorphism (T-RFLP) to investigate fertilization effects on microbial communities, and clone library sequencing and phylogenetic analysis to explore the taxonomic diversity of the nitrifiers and denitrifiers. RESULTS: Fertilization significantly increased the amount of NH₄⁺-N in the soil of SB and MC treatments, whereas MC lowered the NO₃⁻-N level. DEA was higher for MC and CF than the other treatments. The PNR in MC-treated soil was significantly lower than that in CF-treated soil. There were no significant differences in AOA and nirS copy numbers; however, nirK and nosZ copy numbers were higher for MC compared with CF. The number of AOA terminal restriction fragments (TRFs) increased significantly with N addition and reached the highest level for SB, whereas the number of AOB TRFs did not change significantly between treatments. Similarly, the number of nirK TRFs increased under fertilization, with the highest number obtained for SR; however, no significant change was observed for nirS and nosZ TRFs across different treatments, except for their relative abundance. All AOA amoA genotypes were in archaeal group 1.1b, whereas 95% AOB were in Nitrosospira cluster 3a. More than 40% of nirS OTUs were affiliated with Herbaspirillum, a key N-cycle player in this paddy soil. CONCLUSIONS: The SB and MC treatments had significant effects on soil N, DEA, and PNR levels, and affected the community of N-functional microbes. SB in combination with pig manure would contribute to the improvement of paddy soil fertility.