Differential response of denitrifying community to the application of green manure and reduced chemical fertilizer in a paddy soil

Denitrification could lead to N loss from agricultural soil resulting in a low rate of N availability by crops. We investigated the response of nirK and nirS denitrifiers in a paddy soil to the incorporation of green manure and reduced chemical fertilizer. Soil samples were collected from plots of a short-term fertilization experiment initiated in 2009. The treatments were no chemical fertilizer, no Chinese milk vetch (Astragalus sinicus L., CK), chemical fertilizer (NPK), Chinese milk vetch (MV) plus 80% chemical fertilizer (MF80), MV plus 60% chemical fertilizer (MF60) and MV plus 40% chemical fertilizer (MF40). Abundance and community composition of nirK and nirS denitrifiers were analyzed using quantitative PCR and Miseq sequencing. Reduced chemical fertilizer did not reduce content of total N and available N in soils amended with green manure. Abundances of nirK and nirS genes in different treatments were 1.04 x10(8) to 4.89x10(8) and 1.22 x10(7) to 7.04 x10(7) copies g(-1) soil, respectively, which were significantly higher in NPK treatment than those in soils with green manure. NirK abundance was positively correlated with the potential denitrifying activity (PDA) 0.827, p < 0.01). Green manure combination with reduced chemical fertilizer significantly changed the community structure of nirK denitrifiers but not nirS denitrifiers relative to soils amended with chemical fertilizer and unfertilized soil. Shifts of community structure of nirK denitrifiers were closely associated with soil organic matter (r(2) = 0.623, p = 0.003), available N (r(2) = 0.507, p = 0.01), pH (r(2) = 0.661, p = 0.006), and PDA (r(2) = 0.633, p = 0.005). In conclusion, nirK-type denitrifying community was more sensitive to the incorporation of green manure and reduced chemical fertilizer and they played a more important role in the denitrification process in this study.