Introduction

Nitrous oxide (N2O) is one of the most prevalent greenhouse gases, and its warming potential is approximately 265 times that of CO2 on the 100-year scale1. In addition, N2O is an important substance that destroys the ozone layer2. Annual average N2O emissions from global cultivated fields total 6.4 × 1012 g [calculated as nitrogen (N)], which accounts for 25% of total global N2O emissions3. Agricultural fertilization management is the main source of N2O emissions. Application of nitrogen fertilizer stimulates N2O emission from cultivated soils4,5. Therefore, it is urgent to find an effective strategy for decreasing N2O emissions.

N2O is produced mainly through biological and abiotic processes. Denitrification is one of the main ways N2O is produced under microbe-driven conditions in soil6. Denitrification is a process in which nitrate (NO3–N) is sequentially reduced to nitrite (NO2–N), nitric oxide (NO), nitrous oxide (N2O), and finally dinitrogen gas (N2)7. It is essentially the inverse process of nitrification. Nitrite-reducing bacteria are encoded by nitrite reductase genes (nirK and nirS). In addition, reduction of N2O to N2 is one of the key steps in the whole process, and is encoded by the nosZ gene8,9. Nos-catalyzed N2O reduction plays a significant role in N2O emissions. Therefore, the abundance of nirS, nirK and nosZ denitrifying genes was used to study soil nitrogen cycling and N2O emissions10,11. At present, nitrogen addition affecting N2O emissions has attracted wide concern12,13. Wang et al.14 reported that nitrogen addition at 250 kg ha−1 promoted denitrification, thus contributing to comparable stimulation of N2O emissions. Albanito et al.15 proposed a linear relationship between N2O emissions and amounts of nitrogen fertilizer applied in tropical and subtropical regions. To reduce N2O emissions, combined application of organic materials and inorganic fertilizer has been endorsed by some scholars. The addition of organic material altered the abundance of denitrifiers and thus influenced the denitrification process16,17. Wang et al.18 found that organic materials stimulated the activity of the nosZ gene and decreased N2O emissions. However, some studies have suggested that organic materials (i.e., wastewater sludge, manure and vermicompost) provide carbon sources for denitrifiers and accelerate denitrification, thus increasing N2O emissions11,19. Although many studies have been focused on combined application of organic and inorganic fertilizers and its effect on N2O emissions in agricultural soils, the research conclusions are not yet consistent. Therefore, it is necessary to carry out research to determine the drivers of N2O emissions and propose reasonable fertilization strategies.

In addition, the identification of major denitrification-driving genes plays an important role in predicting N2O emissions. Yin et al.20 indicated that denitrifiers were more sensitive to inorganic fertilization than to organic fertilization. Harter et al.21 revealed that the nosZ gene is the decisive factor for N2O emissions. Hai et al.22 found that organic fertilizers increased the abundance of nirS and nirK genes and enhanced N2O emissions. Huang et al.23 reported that co-addition of urea and cattle manure decreased the abundance of nirS and nirK but increased N2O emissions. In addition, Xu et al.24 suggested that nirS-type denitrifying genes showed strong correlations with significant increases in N2O emissions from soils undergoing organic fertilization. Henderson et al.25 questioned whether the relationship between N2O emissions and the abundance of nirS and nosZ genes was significant. Other studies have focused on the impacts of co-amendment with inorganic and organic fertilizers on the abundance of denitrification bacterial communities and changes in soil properties21,26,27,28. However, there was still a knowledge gap regarding the dominant drivers of N2O emissions, including denitrifiers and environmental factors.

Therefore, this study was focused on the impacts of co-additions of urea and four organic additions (OAs), including wheat straw, pig manure, compost products and improved compost products, on N2O emissions. Additionally, fluorescence quantitative PCR and 16S rDNA sequencing were used to study changes in the abundance of denitrifying genes in soil caused by OAs and clarify the mechanism for the response of denitrifying gene abundance to added OAs. Particularly, the purposes of this research were (i) to compare the impacts of different OAs with urea on changes in N2O emissions; (ii) to understand the abundance of denitrifiers affected by different OAs; and (iii) to identify the relationships among N2O emissions, soil properties and the abundance of denitrifiers.

Materials and methods

Experimental soil collection

The tested soil was collected from the experimental station around Northwest A&F University in Shaanxi Province, China. The tested plot lies in a typical semiarid region with average annual precipitation of 533–631 mm. The basic soil properties before the experiment were as follows: pH = 7.53, soil organic carbon (SOC) = 0.79%, total nitrogen (TN) = 0.075%.

Experimental materials collection

The four OAs included compost product (CP), wheat straw (WS), pig manure (PM) and improved-compost product (IC). PM and WS were collected from the farm and field of the experimental station. CP and IC products were produced in the greenhouse on the campus. CP was manufactured with PM and WS. IC was produced by the addition of bean dregs and biochar to the compost product. Previous studies confirmed that IC products increased SOC and TN contents more than compost products29,30. All OA materials were dried at 55 °C and then sieved through a 2 mm mesh. Basic physicochemical properties for all OAs are shown in the Supplementary Material.

Experimental design

Six incubation experiments, including no fertilization (CK), urea N alone (U), 70% urea N plus 30% N from PM (U + PM), 70% urea N plus 30% N from WS (U + WS), 70% urea N plus 30% N from CP (U + CP) and 70% urea N plus 30% N from IC (U + IC), were conducted with three replicates to understand the responses of N2O emissions and denitrification genes to different fertilization practices. Fresh soil samples were first weighed (700 g for dry weight basis) and placed into plastic bottles. A total of 0.5 mg N g−1 of dry soil was added to each treatment. The amounts of urea and OAs were calculated by their total nitrogen content. The content of soil moisture was controlled at 60% of the water-holding capacity. Incubation was performed in a 25 °C environment for 77 days. Each plastic bottle was weighed, and distilled water was added to maintain constant moisture in the soil. All soil samples were divided into three parts after incubation: one was used for determination of soil properties after drying, one was used for measurements of NO3–N and NH4+-N at 4 °C, and the other was used for DNA determination at − 80 °C.

Properties determination

Soil NO3–N and NH4+-N were determined with an AA3 flow analyzer by extraction with 2 mol L−1 KCl. Soil organic carbon (SOC) was measured through K2Cr2O7 digestion. Total nitrogen (TN) was measured by a Kjeldahl nitrogen analyzer31. In addition, soil moisture content was determined by weighing after drying. In addition, the soil microbial biomass carbon (MBC) and microbial biomass nitrogen (MBN) contents were determined using the chloroform fumigation-extraction method32,33. Briefly, triplicate soil samples were weighed and fumigated with ethanol-free chloroform for 24 h at 25 °C. The three fumigated and unfumigated samples from each treatment were extracted with 100 mL of 0.5 mol L−1 K2SO4 on a shaker at 220 rpm for 30 min and then filtered and determined with a TOC/TN analyzer. Dissolved organic carbon (DOC) was determined by a TOC-analyzer after extraction and filtering34. Dissolved organic nitrogen (DON) was measured using a flow analyzer after extraction with H2O35. N2O samples were collected on days 0, 3, 7, 14, 21, 28, 35, 49, 63 and 77 during incubation. All bottles were sealed for 10 min before sampling to ensure concentration of N2O emissions. N2O samples were determined by gas chromatography. The N2O fluxes and cumulative N2O emissions were calculated as described by Huang et al.36. The specific calculation formula is as follows:

$$ F = \frac{{\Delta {{C}}}}{\Delta t} \cdot \frac{V}{m} \cdot \frac{M}{\textit{22.4}} \cdot \frac{\textit{273}}{{\textit{273} + T}} $$

F is the N2O fluxes (μg/kg/day), ∆C/∆t is the variation of concentration per unit time (μg/kg/day), V is the volume of the device (L), m is the dry soil weight (g), M/22.4 is the mass density of standard gases (g/L), and T is the incubation temperature (°C).

Cumulative N2O emission is the accumulation of the N2O fluxes during the incubation period (μg/kg).

Microbial analyses

Soil DNA was extracted from 0.5 g of fresh soil using a DNA kit (Omega, GA, USA). The DNA solution extracted from each sample was studies with 1% agarose gel electrophoresis. The nirK, nirS and nosZ denitrifying genes were amplified by PCR using primers (Supplementary Materials). The abundances of denitrifiers were quantified by quantitative real-time PCR (qPCR) on an Illumina MiSeq platform in Beijing.

Data analyses

Figures were prepared with Origin 2018. Significant differences among the treatments were analyzed by Duncan's multiple range test using IBM SPSS Statistics (version 20.0). Automatic linear modeling was conducted using IBM SPSS Statistics (version 20.0). Redundancy analyses (RDA) and principal component analyses (PCA) were conducted with Canoco (version 4.5).

Results

Changes in N2O emission

N2O emissions influenced significantly by different OAs are shown in Fig. 1a. A similar trend for N2O emissions among all treatments was observed. N2O emissions increased within the first 7 days and declined thereafter. N2O emissions peaked on day 3 for the CK, U, U + WS and U + PM treatments and peaked on day 4 for the U + IC and U + CP treatments. The OAs increased N2O emissions during the whole incubation process compared with the CK treatment. The maximum peak value for N2O emissions among all OAs was that seen with the U + PM treatment (76.04 mg/kg/day), and the minimum value of N2O emissions was that seen with the U + IC treatment (21.44 mg/kg/day). Compared with other OA treatments, compost amendments (U + IC, U + CP) decreased the peak values of N2O emissions. Furthermore, the highest cumulative N2O emissions were seen with the U + PM treatment, which were 241.8% higher than the levels seen with the CK sample; this was followed by the U + WS (111.1%), U + CP (44.63%), U (41.61%) and U + IC (8.70%) treatments (Fig. 1b). In addition, cumulative N2O emissions for the U + WS, U + PM and U + CP treatments were 49.05%, 141.3% and 2.12% higher relative to the U treatment, but the cumulative N2O emission for the U + IC treatment was 23.24% lower compared with that of the U treatment. These results indicated that relative to application of urea alone, mixing urea with wheat straw, pig manure and compost products promoted N2O emissions during incubation. In addition, IC retarded N2O emissions to a greater degree compared with other OA additions.

Figure 1
figure 1

The changes of N2O emission (a) and cumulative N2O emission (b) during the incubation affected by different OAs. CK no chemical fertilizer, U urea, U + WS urea plus wheat straw, U + PM urea plus pig manure, U + CP urea plus compost, U + IC urea plus improved compost.

Full size image

Changes in soil properties

The basic properties of soils are shown in Table 1. The highest SOC content was observed for the U + WS treatment, followed by the U + IC and U + CP treatments, and these levels were significantly higher than that of the CK treatment (P < 0.05). However, there was no significant difference in SOC contents for the U + PM and CK treatments (P > 0.05). Relative to the CK treatment, OAs significantly increased the levels of DON, NH4+-N and NO3–N. The highest NO3–N, NH4+-N and DON contents were found for the U + PM treatment, and these were higher than those seen for other OA treatments (P < 0.05). However, the DOC contents with all OA treatments except for U + IC differed significantly from that of the CK treatment (P < 0.05). The maximum DOC concentration was observed for the U + WS treatment. In addition, only the U + PM treatment, among all OA treatments, failed to increase the MBN content significantly (P > 0.05). The U treatment did not significantly increase any soil properties during incubation relative to the CK treatment except for NO3–N (P > 0.05).

Table 1 Soil properties in different treatments with OAs.
Full size table

Abundances of nirK, nirS and nosZ denitrifying genes

The relative abundances of denitrification genes (nirK, nirS and nosZ) were determined during incubation for all treatments (Fig. 2). The copy numbers for nirK, nirS and nosZ varied within the ranges 5.26 × 103–4.47 × 104, 1.26 × 103–9.36 × 103 and 1.04 × 103–2.51 × 103, respectively. The nirK-type denitrification genes exhibited the highest abundance after incubation. Compared with the CK treatment, all OA treatments (U + CP, U + IC, U + PM and U + WS) significantly increased the abundance of the nirK gene. The highest copy number of the nirK gene was seen for the U + PM treatment, followed by the U + WS, U + CP and U + IC treatments. In addition, there were no significant differences between the U + IC and U + CP treatments. However, the nosZ gene had the lowest abundance after incubation. Compared with the CK treatment, all OA treatments except the U + WS treatment significantly increased the copy number of the nosZ gene. The maximum abundance of the nosZ gene was observed for the U + PM treatment. There was no significant difference between the CK and U treatments. For abundance of the nirS gene, the U and OA treatments all significantly increased the abundance relative to the CK treatment. The highest nirS gene abundance was seen for the U + PM treatment, followed by the U + WS, U + CP and U + IC treatments. However, there was no significant difference between the U and U + CP treatments. Among all denitrification genes, the U treatment significantly increased only the abundance of nirS genes. These results indicated that co-additions of urea and OAs increased the abundance of nirK genes more, while urea alone increased the abundance of nirS genes more.

Figure 2
figure 2

Relative abundance of denitrification genes during the incubation affected by different OAs. Error bars indicate standard deviations (n = 3). The different letters indicate significant differences (P < 0.05) (nirS (a); nirK (b); nosZ (c); CK no urea and OAs, U urea, U + WS urea plus wheat straw, U + PM urea plus pig manure, U + CP urea plus compost, U + IC urea plus improved compost).

Full size image

Correlations of denitrifying gene abundance and soil properties

PCA was used to analyze the abundances of denitrifiers among all treatments (Fig. 3a). The results showed that the explanations of Axis 1 and Axis 2 contributed 73.66% and 10.56%, respectively, to changes in denitrifying genes (nirS, nirK and nosZ). The OA treatments were largely separated from the CK and U treatments along PCA1. Additionally, the U + PM and U + WS treatments were segregated from the U + CP and U + IC treatments along PCA2. The results suggested that OA application significantly influenced the abundance of the three types of denitrifiers, and OAs had stronger effects on the abundance of denitrifiers than urea application alone. Correlations among soil properties, accumulative N2O emissions and denitrifying gene abundance showed that the abundance of the nirS gene was significantly associated with accumulative N2O emissions (r = 0.876, P < 0.05) and DOC (r = 0.859, P < 0.05). The abundance of the nirK gene was correlated with DON (r = 0.977, P < 0.1), NO3–N (r = 0.880, P < 0.05), DOC (r = 0.865, P < 0.05), MBN (r = 0.852, P < 0.05) and NH4+-N (r = − 0.880, P < 0.05). In addition, the abundance of the nosZ gene was mainly related to SOC (r = 0.965, P < 0.1), MBN (r = 0.836, P < 0.05) and NH4+-N (r = − 0.931, P < 0.1). In addition, the results of automatic linear modeling showed that nirK abundance, DOC, NO3–N, MBN and DON were the main factors influencing N2O emissions (Fig. 3b).

Figure 3
figure 3

Principal component analysis (PCA) of soil properties, cumulative N2O emission and abundance of denitrification genes (a); predictive importance of denitrifier gens abundance, and soil properties on cumulative N2O emission as determined by automatic linear modeling (b) (DOC dissolved organic carbon, MBN microbial biomass carbon, DON dissolved organic nitrogen, SOC soil organic carbon, CK no urea and OAs, U urea, U + WS urea plus wheat straw, U + PM urea plus pig manure, U + CP urea plus compost, U + IC urea plus improved compost).

Full size image

Discussion

Different fertilization practices influenced N2O emissions from soil. Soil N2O emissions peaked on day 3 for the CK, U, U + WS and U + PM treatments, while those for the U + CP and U + IC treatments peaked on day 7. This may be attributed to the fact that WS and PM provided more rapidly available N than CP and IC37. In addition, compared with urea application alone, the U + CP, U + WS and U + PM treatments increased cumulative N2O emissions by 2.13–49.06%, which was within the increase range of 27–74% seen with pig slurry and compost amendments38. However, the U + IC treatment decreased cumulative N2O emissions by 23.24% relative to urea alone. Therefore, the combined application of IC and urea reduced N2O emissions in soil because IC was more stable than CP, PM and WS. Additionally, OA treatment increased cumulative N2O emissions during incubation relative to the CK treatment. This might be because OAs provided more substrates for denitrification through mineralization. Denitrification could result in a rapid reduction in NO3–N content and promote the emission of N2O. Previous studies showed that the NO3–N concentration was an important factor affecting the denitrification rate and N2O release39,40,41,42,43. Increasing the concentration of NO3–N could significantly increase N2O release. Compared with urea application alone, the OA additions increased the content of NO3–N in soil; furthermore, PM application provided more NO3–N than WS, CP and IC. Therefore, PM application promoted changes in N2O emissions, which was consistent with the variations in NO3–N. In addition, SOC increased by addition of OAs to the soil. The importance of SOC as a factor affecting denitrification and N2O emissions has been reported by Chen et al.44. However, in this study, the effect of SOC on denitrifiers and N2O emissions was lower than that of DOC due to the narrow range of SOC changes occurring in a short-term incubation experiment. Carbon availability was the key controlling factor for denitrification in soil45. Compared with the U + PM treatment, the U + WS treatment produced a higher DOC content but lower N2O emissions, which may be due to the higher C/N ratio in WS. Microbial fixation of carbon in the straw treatment limited carbon effectiveness. In addition, of all treatments involving addition of organic material, the modified compost addition treatment produced the lowest DOC content due to its higher degree of humification, which reduced the abundance of nirK genes and therefore reduced N2O emissions from the soil37.

N2O emissions were highly related to soil physiochemical properties and the abundance of denitrifiers, which was consistent with the findings of Sun et al.46. Denitrifying bacteria are active in soil biological denitrification11,47. Many studies showed that fertilization increased the number of denitrifying microorganisms in the soil by providing substrates and energy for denitrifying bacteria and promoting their growth and reproduction37,46,48. In this study, the OA addition increased the abundance of denitrifying genes in soil. The contributions of nirK to N2O emissions were higher than those of nirS and nosZ. In addition, application of pig manure significantly increased nirK gene abundance and thus promoted N2O emissions more than other OAs. Yoshida et al.49 also found that application of organic manure increased the abundance of nirK genes in rice paddy soil more than that of nirS genes. However, there are some studies showing the opposite results. For example, Yin et al.50 found that organic manure changed the abundance of the nirS gene community in black soil during long-term treatment but not that of nirK. Barrett et al.51 confirmed that a higher abundance of nirS-type genes was observed in carbon- amended soil relative to other genes. This may be related to the different sources and nature of available carbon in amended soil. Environmental factors significantly influenced the abundance of denitrifiers in this study52,53. Positive correlations were observed between NO3–N, DON, DOC and the abundance of the nirK gene. These findings suggested that NO3–N, DON and DOC levels were important factors affecting the abundance of the soil denitrifying bacterial community, thus affecting denitrification with different OA applications. This may be due to differences in the availability of C and N in OA-amended soil.

Conclusion

Fertilization treatments increase denitrification and N2O emissions relative to the CK treatment. In addition, compared with urea alone, combined application of pig manure and urea provided more available N (DON and NO3–N) and increased the abundance of nirK genes, thus increasing cumulative N2O emissions from the soil. However, combined applications of improved compost products and urea reduced accumulated N2O emissions by decreasing the abundance of nirK genes. Overall, the different fertilization practices affected the abundance of denitrifiers, denitrification and soil properties. From the perspective of soil N2O emission reduction, we recommend the application of improved compost products and urea to the soil.