Abstract:
Microorganisms with nitrite reductase genes can reduce nitrite to nitric oxide (NO), which is an important influence in the biological nitrogen cycle. A field study was conducted to investigate soil denitrifying bacteria (
nirK- and
nirS-type) communities and diversity in a farmland (
Zea mays field) and
Medicago sativa land established based on different times (2, 9, and 18 years, respectively expressed as L
2019, L
2012 and L
2003). Illumina MiSeq high-throughput sequencing and real-time fluorescent quantitative PCR technology were used to investigate the structure and diversity of denitrifying bacterial communities under four treatments (Farmland, L
2003, L
2012 and L
2019). Redundancy analysis and molecular ecological network analysis were used to evaluate the relationship between soil physical and chemical properties and denitrifying bacterial community. The results indicated that the abundance of
nirK gene was significantly higher than that of
nirS gene. The abundance of
nirK gene varied from 4.91×10
7 to 6.33×10
7 copies∙g
−1, whereas the abundance of
nirS gene varied from 1.02×10
7 to 1.86×10
7 copies∙g
−1. The years of
M. sativa cultivation did not affect the diversity of
nirK- and
nirS-type denitrifying bacteria. Proteobacteria had the highest abundance in the denitrifying bacterial community. The dominant genera of the
nirK-type denitrifying bacteria were
Paracoccus (1.10%–39.94%),
Achromobacter (0.07%–12.50%), and
Sinorhizobium (0.50%–7.60%). The relative abundance of
Paracoccus in
M. sativa soil was significantly higher than that in maize soil (
P<0.05), and the relative abundance gradually increased with increasing age of
M. sativa stands. The relative abundance of
Achromobacter in
M. sativa soil was significantly lower than that in maize soil (
P<0.05), and the abundance decreased gradually with increasing age of the
M. sativa stand. The dominant genus of
nirS-type denitrifying bacteria was
Rhodobacter (1.42%–5.20%). There was no significant difference in
Rhodobacter abundance between the maize fields and
M. sativa fields. Correlation analysis showed that the abundance of
nirK-type denitrifying bacteria had no significant response to soil environmental factors, but the abundance of
nirS-type denitrifying bacteria had a significant positive correlation with soil organic carbon (
r=0.762), total nitrogen (
r=0.776), and microbial biomass carbon (
r=0.622) and a significant negative correlation with soil water (
r=–0.678) and available phosphorus (
r=–0.628). RDA analysis indicated that soil water (
P=0.002) and organic carbon (
P=0.020) were the main environmental factors affecting the community structure of
nirK-type denitrifying bacteria, and soil available phosphorus (
P=0.006) was the main environmental factor affecting the community structure of
nirS-type denitrifying bacteria. The proportion of positively correlated edges in the
nirK-type denitrifying bacterial ecological network was 98.37%, and the proportion of negatively correlated edges was 1.63%; however, all edges in the
nirS-type denitrifying bacterial ecological network were positively correlated. This indicated that the relationship between bacterial communities of both types of denitrifying bacterial was mainly synergistic. In summary, long-term planting of
M. sativa significantly affected the composition of soil denitrifying bacterial community. Our results provide a scientific basis for further studies on the microbial mechanism of denitrification in the Loess Plateau after years of
M. sativa planting.