Response of N2O emission and denitrification genes to different inorganic and organic amendments

Denitrification is a key biochemical process in nitrogen cycling and nitrous oxide (N2O) production. In this study, the impacts of different inorganic and organic amendments (OAs) on the abundance of denitrifying genes (nirS, nirK and nosZ) and the level of N2O emission were examined with incubation experiments. Six treatments included the indicated applications: (i) no fertilization (CK); (ii) urea application alone (U); (iii) wheat straw plus urea (U + WS); (iv) pig manure plus urea (U + PM); (v) compost product plus urea (U + CP); and (vi) improved compost product plus urea (U + IC). The results indicated that all fertilization treatments increased accumulative N2O emissions compared with the CK treatment. The U + WS, U + PM and U + CP treatments increased N2O emissions by 2.12–141.3%, and the U + IC treatment decreased N2O emissions by 23.24% relative to the U treatment. nirK was the dominant denitrification gene rather than nirS and nosZ found in soil. Additionally, the highest abundance of nirK gene was that with the U + PM treatment, and the lowest was that with the U + IC treatment. Additionally, changes in the nirK gene were highly correlated with levels of dissolved organic carbon (DOC), dissolved organic nitrogen (DON) and nitrate nitrogen (NO3–N). Automatic linear modeling revealed that N2O emission was closely related to the nirK gene, DOC and NO3–N. Overall, the use of urea and improved compost as co-amendments retarded N2O emission to a considerable degree compared with other OA additions.

been focused on combined application of organic and inorganic fertilizers and its effect on N 2 O emissions in agricultural soils, the research conclusions are not yet consistent. Therefore, it is necessary to carry out research to determine the drivers of N 2 O emissions and propose reasonable fertilization strategies.
In addition, the identification of major denitrification-driving genes plays an important role in predicting N 2 O 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 N 2 O emissions. Hai et al. 22 found that organic fertilizers increased the abundance of nirS and nirK genes and enhanced N 2 O emissions. Huang et al. 23 reported that co-addition of urea and cattle manure decreased the abundance of nirS and nirK but increased N 2 O emissions. In addition, Xu et al. 24 suggested that nirS-type denitrifying genes showed strong correlations with significant increases in N 2 O emissions from soils undergoing organic fertilization. Henderson et al. 25 questioned whether the relationship between N 2 O 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 properties 21,[26][27][28] . However, there was still a knowledge gap regarding the dominant drivers of N 2 O 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 N 2 O 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 N 2 O emissions; (ii) to understand the abundance of denitrifiers affected by different OAs; and (iii) to identify the relationships among N 2 O 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 products 29,30 . 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 N 2 O 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 NO 3 -N and NH 4 + -N at 4 °C, and the other was used for DNA determination at − 80 °C.
Properties determination. Soil NO 3 -N and NH 4 + -N were determined with an AA3 flow analyzer by extraction with 2 mol L −1 KCl. Soil organic carbon (SOC) was measured through K 2 Cr 2 O 7 digestion. Total nitrogen (TN) was measured by a Kjeldahl nitrogen analyzer 31 . 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 method 32,33 . 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 K 2 SO 4 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 filtering 34 . Dissolved organic nitrogen (DON) was measured using a flow analyzer after extraction with H 2 O 35 . N 2 O 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 N 2 O emissions. N 2 O samples were determined by gas chromatography. The N 2 O fluxes and cumulative N 2 O emissions were calculated as described by Huang et al. 36 . The specific calculation formula is as follows: F is the N 2 O 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).  -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 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 × 10 3 -4.47 × 10 4 , 1.26 × 10 3 -9.36 × 10 3 and 1.04 × 10 3 -2.51 × 10 3 , 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 Table 1. Soil properties in different treatments with OAs. 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. Values are mean ± standard deviation (n = 3). The different letters indicate significant differences at the 0.05 probability level. www.nature.com/scientificreports/ 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.

Results
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). www.nature.com/scientificreports/ 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 N 2 O emissions and denitrifying gene abundance showed that the abundance of the nirS gene was significantly associated with accumulative N 2 O 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), NO 3 -N (r = 0.880, P < 0.05), DOC (r = 0.865, P < 0.05), MBN (r = 0.852, P < 0.05) and NH 4 + -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 NH 4 + -N (r = − 0.931, P < 0.1). In addition, the results of automatic linear modeling showed that nirK abundance, DOC, NO 3 -N, MBN and DON were the main factors influencing N 2 O emissions (Fig. 3b).

Discussion
Different fertilization practices influenced N 2 O emissions from soil. Soil N 2 O 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 IC 37 . In addition, compared with urea application alone, the U + CP, U + WS and U + PM treatments increased cumulative N 2 O emissions by 2.13-49.06%, which was within the increase range of 27-74% seen with pig slurry and compost amendments 38 . However, the U + IC treatment decreased cumulative N 2 O emissions by 23.24% relative to urea alone. Therefore, the combined application of IC and urea reduced N 2 O emissions in soil because IC was more stable than CP, PM and WS. Additionally, OA treatment increased cumulative N 2 O 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 NO 3 -N content and promote the emission of N 2 O. Previous studies showed that the NO 3 -N concentration was an important factor affecting the denitrification rate and N 2 O release [39][40][41][42][43] . Increasing the concentration of NO 3 -N could significantly increase N 2 O release. Compared with urea application alone, the OA additions increased the content of NO 3 -N in soil; furthermore, PM application provided more NO 3 -N than WS, CP and IC. Therefore, PM application promoted changes in N 2 O emissions, which was consistent with the variations in NO 3 -N. In addition, SOC increased by addition of OAs to the soil. The importance of SOC as a factor affecting denitrification and N 2 O emissions has been reported by Chen et al. 44 . However, in this study, the effect of SOC on denitrifiers and N 2 O 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 soil 45 . Compared with the U + PM treatment, the U + WS treatment produced a higher DOC content but lower N 2 O 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 N 2 O emissions from the soil 37 . N 2 O 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 denitrification 11,47 . 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 reproduction 37,46,48 . In this study, the OA addition increased the abundance of denitrifying genes in soil. The contributions of nirK to N 2 O emissions were higher than those of nirS and nosZ. In addition, application of pig manure significantly increased nirK gene abundance and thus promoted N 2 O 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 study 52,53 . Positive correlations were observed between NO 3 -N, DON, DOC and the abundance of the nirK gene. These findings suggested that NO 3 -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 N 2 O 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 NO 3 -N) and increased the abundance of nirK genes, thus increasing cumulative N 2 O emissions from the soil. However, combined applications of improved compost products and urea reduced accumulated N 2 O 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 N 2 O emission reduction, we recommend the application of improved compost products and urea to the soil.

Data availability
All data generated or analysed during this study are included in this published article (and its Supplementary  Information files).