Biochar amendment reduces paddy soil nitrogen leaching but increases net global warming potential in Ningxia irrigation, China

The efficacy of biochar as an environmentally friendly agent for non-point source and climate change mitigation remains uncertain. Our goal was to test the impact of biochar amendment on paddy rice nitrogen (N) uptake, soil N leaching, and soil CH4 and N2O fluxes in northwest China. Biochar was applied at four rates (0, 4.5, 9 and13.5 t ha−1 yr−1). Biochar amendment significantly increased rice N uptake, soil total N concentration and the abundance of soil ammonia-oxidizing archaea (AOA), but it significantly reduced the soil NO3 −-N concentration and soil bulk density. Biochar significantly reduced NO3 −-N and NH4 +-N leaching. The C2 and C3 treatments significantly increased the soil CH4 flux and reduced the soil N2O flux, leading to significantly increased net global warming potential (GWP). Soil NO3 −-N rather than NH4 +-N was the key integrator of the soil CH4 and N2O fluxes. Our results indicate that a shift in abundance of the AOA community and increased rice N uptake are closely linked to the reduced soil NO3 −-N concentration under biochar amendment. Furthermore, soil NO3 −-N availability plays an important role in regulating soil inorganic N leaching and net GWP in rice paddies in northwest China.

slow carbon and N release 15,16 via the high content of recalcitrant organic carbon in the biochar and concomitant changes in soil properties, which affect microbial activity 17 . Recent reviews have highlighted biochar as a possible method to decrease soil CH 4 and N 2 O emissions 18,19 and reduce N leaching 20 .
The effects of biochar on paddy soil CH 4 emissions remain controversial depending on the biochar type, climatic conditions and soil properties 21 . A laboratory incubation study showed that amendment with bamboo biochar and rice-straw biochar decreased paddy soil CH 4 emissions by up to 51% and 91%, respectively 22 , while wheat-straw biochar amendment increased soil CH 4 emission by 37% 23 . In addition, no significant difference in soil CH 4 flux was found between a biochar plot and a control plot in Germany 24 . Biochar application can affect N transformation and N fate in soil 16,25 . The soil N 2 O flux increased significantly in some studies 26,27 but substantially decreased or remained unchanged in others 23,28 . These contrasting results emphasize the need for more studies to assess the role of biochar in mitigating paddy soil CH 4 and N 2 O fluxes. Moreover, the mechanisms of action are not well understood, which has impeded the adoption of biochar in a wide range of rice ecosystems.
Nitrification, through which microorganisms oxidize ammonium (NH 4 + ) to create nitrate (NO 3 − ), releasing N 2 O as a by product, has long been a concern of scientists in paddy soils. Many studies found lower N leaching after biochar amendment in laboratory and field experiments 20,29 . However, the underlying mechanisms are still controversial. Recent studies have demonstrated that increased water-holding capacity, enhanced microbial biomass and altered bacterial community structure in soils may contribute to the reduction of N leaching 20 . Other studies suggest that reduced N leaching may result from improved plant N use 30,31 . NH 3 oxidation to NO 2 − , the first and rate-limiting step of nitrification, is catalyzed by ammonia-oxidizing archaea (AOA) and ammonia-oxidizing bacteria (AOB) 32 . Due to the homology of methane monooxygenase and ammonia monooxygenase enzymes, the same habitats, and the variety of analog substrates, CH 4 in the soil can be simultaneously oxidized by both methanotrophs and ammonia oxidizers 33,34 . It is essential to explore the links between ammonia oxidizers and the soil CH 4 and N 2 O fluxes in the field under biochar amendment.
In the upper reaches of the Yellow River basin of China, preliminary studies revealed that biochar amendment significantly improves N use efficiency 35 and reduces total inorganic N leaching 36 . Consequently, we hypothesized that biochar amendment decreases soil N 2 O emission and reduces NO 3 − -N and NH 4 + -N leaching. We also hypothesized that biochar amendment increases soil CH 4 emissions due to the labile carbon input and the positive priming effects of biochar and also increases crop productivity. The aim of the present study was to provide insight into the effects of biochar amendment on paddy soil NO 3 − -N and NH 4 + -N leaching and soil CH 4 and N 2 O emissions throughout the entire growth period. By considering the net global warming potential (GWP) of the soil CH 4 and N 2 O fluxes and the NO 3 − -N and NH 4 + -N leaching under different treatments, it should be possible to determine the optimal amount of biochar application for China's rice paddies.

Materials and Methods
Study site. This experiment was conducted at Yesheng Town (106°11′35″ E, 38°07′26″ N) in Wuzhong City, China. The temperate continental monsoon climate dominates the region, with a mean temperature of 9.4 °C and a mean annual precipitation of 192.9 mm. The soil is classified as anthropogenic alluvial soil, with a soil texture of 18.25% clay, 53.76% silt, and 27.99% sand. The top soil (0-20 cm) organic matter is 16.1 g kg −1 , the total N is 1.08 g kg −1 , and the soil bulk density is 1.33 g cm −3 .
Experimental design and rice management. Biochar was applied to the field plots at rates of 0 (C0), 4.5 (C1), 9.0 (C2) and 13.5 t ha −1 yr −1 (C3). Each treatment was performed in triplicate. A total of 12 plots (30 m × 20 m) were established, and each was separated by plastic film to 130 cm in depth, preventing water interchange between adjacent plots. Each plot was irrigated with an equal amount of water (approximately 14500 m 3 ha −1 yr −1 ). The biochar was produced by pyrolysis of wheat straw at 350-550 °C by the Sanli New Energy Company, Henan Province, China. The biochar had C, N, P and K contents of 65.7%, 0.49%, 0.1% and 1.6%, respectively, with a pH (H 2 O) of 7.78 35 . Urea was applied at 240 kg N ha −1 , of which 50% was applied as a base fertilizer before transplanting (26 May, 2014), 30% was applied at the tillering stage (6 June, 2014), and the remaining 20% was applied at the elongation stage (25 June, 2014). Double superphosphate and KCL were also applied as basal fertilizers before transplanting at rates of 90 kg P 2 O 5 ha −1 and 90 kg K 2 O ha −1 , respectively. Biochar and fertilizers were broadcast on the soil surface and incorporated into the soil by plowing to a depth of approximately 13 cm in May 2014. To maintain consistency, plowing was also performed for the plots without biochar. Rice (Oryza sativa L., cv. 96D10) was sown in a nursery bed on 1 May. Rice seedlings were transplanted on 28 May and harvested on 12 October 2014. Crop management was consistent across plots.
Measurement of the soil CH 4 and N 2 O fluxes. The soil CH 4 and N 2 O fluxes were measured using a static opaque chamber and gas chromatography, as described by Wang et al. 33 . Sampling of emitted gases was conducted between 8:00 and 10:00 in the morning. Fluxes were measured twice a week after irrigation and fertilization before the booting stage. Afterwards, the measurement frequency decreased to three times a month during the rice booting, filling and maturity stages and to two times a month during the fallow period. The gas fluxes were measured on 21 occasions during the observation period. The concentrations of CH 4 and N 2 O in the gas samples were simultaneously analyzed within 24 h using a gas chromatograph (Agilent 7890A, USA) equipped with a flame ionization detector (FID) and an electron capture detector (ECD). High-purity N 2 and H 2 were used as the carrier gas and fuel gas, respectively. The ECD and FID were heated to 350 °C and 200 °C, respectively, and the column oven was kept at 55 °C. The CH 4 and N 2 O fluxes were calculated based on the rate of change in concentration within the chamber, which was estimated as the linear or nonlinear regression slope between concentration and time 37 .
Scientific RepoRts | 7: 1592 | DOI:10.1038/s41598-017-01173-w Soil sampling and analysis. Soil samples (0-20 cm) were collected three times: tillering stage (16 June), filling stage (10 August) and harvest stage (12 October). Five soils from two diagonal lines through each plot were collected and pooled into one composite sample. Soils were sieved to 2-mm mesh size in the field and were then transported to the lab in a biological refrigerator. Soil samples were stored at −80 °C before analysis. Soil NH 4 + -N and NO 3 − -N were determined using a continuous-flow auto analyzer (Seal AA3, Germany). Soil bulk density was measured using a 100 cm 3 cylinder. The total N (TN) contents in the bulk soil were determined by dry combustion using the Kjeldahl method 38 .
Soil DNA was extracted from 0.5 g of soil using the Fast DNA ® SPIN Kit (Qbiogen Inc., Carlsbad, CA, USA) for soil following the manufacturer's instructions. The extracted DNA was checked on 1% agarose gel, and the DNA concentration was assessed using a Nanodrop ® D-1000 UV-Vis spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA). Tenfold diluted DNA was used in the PCR analysis. Primer pairs Arch-amoAF/Arch-amoAR 39 and amoA1F/amoA2R 40 were used for the qPCR of AOA and AOB amoA genes, as described by Wang et al. 33 . Product specificity was checked by melt curve analysis at the end of the PCR runs and by visualization via agarose gel electrophoresis. A known copy number of plasmid DNA for AOA or AOB was used to generate a standard curve. For all assays, the PCR efficiency was 90-100% and r 2 was 0.95-0.99.
Soil leachate sampling and analysis. Soil water samples used for the leaching calculations were collected from lysimeters, as described by Riley et al. 41 . Four PPR (polypropylene random) equilibrium-tension lysimeters (ETLs) (0.19 m 2 ) were installed at the desired depth (20, 60 and 100 cm) below the soil surface for each treatment condition. Soil leachate samples were collected using 100 ml plastic syringes and were transferred to a plastic tube and stored at 4 °C before analysis. Samples were taken on days 1, 3, 5, 7, and 9 after transplanting and topdressing; subsequent sampling was conducted at 10-day intervals. Soil leachate samples were collected 14 times during the observation period. The NH 4 + -N and NO 3 − -N leaching losses were calculated by multiplying the N concentration by the leachate volume.
Rice yield and N uptake. At rice maturity, rice aboveground biomass was estimated by manually harvesting three 0.5 m 2 areas. Rice straw and grain were oven-dried to a constant weight at 80 °C, weighed, finely ground, sieved, and analyzed for total N using the Kjeldahl method 38 . Total N uptake was calculated from the sum of the N mass in the straw and grain harvested from each plot. Statistical analyses. GWP is an index of the cumulative radiative forcing between the present and some chosen later time by a unit mass of gas emitted under specific conditions 42 . To compare net GWP after biochar amendment to soil, we calculated the CO 2 equivalents for CH 4 and N 2 O for a time horizon of 100 yr (assuming a GWP of 25 for CH 4 and 298 for N 2 O) using the following equation.
Repeated measures of analysis of variance (ANOVA) with the least significant difference (LSD) test were applied to examine the differences in N leaching, soil CH 4 and N 2 O fluxes, and soil microbial amoA gene copy numbers among the different treatments. Biochar amendment was set as a between-subjects factor, and the measurement period was selected as a within-subjects variable. We performed one-way ANOVA with an LSD test to evaluate the effects of biochar amendment on the soil properties, rice yield and N uptake. Linear regression analyses were used to examine the relationships among the soil CH 4 and N 2 O fluxes, the microbial amoA gene copy numbers and the soil inorganic N concentrations. All statistical analyses were conducted using the SPSS software (version 16.0), and differences were considered significant at P < 0.05, unless otherwise stated. All figures were drawn using SigmaPlot software (version 10.0).

Results
Soil properties, rice yield and N uptake. In the C0 treatment, the average concentration of soil NO 3 − -N was 26.52 mg kg −1 during the whole observation period ( Table 1). The soil NO 3 − -N concentration was reduced significantly by 15.10%, 32.13% and 51.02% in the C1, C2 and C3 treatments (Table 1). However, the soil NH 4 + -N concentration was only significantly decreased in the C1 treatment (Table 1). Biochar amendment significantly increased soil TN by 10.01~22.22% and significantly decreased bulk density by 4.51~7.81% compared with the treatment without biochar (Table 1). Biochar amendment tended to decrease soil pH (Table 1, P > 0.05).
In the control, the rice yield and grain N uptake were 8357 kg ha −1 and 87.2 kg ha −1 , respectively. Biochar amendment significantly increased the rice yield and grain N uptake compared with the C0 treatment for the C2 and C3 treatments (Fig. 1a,b). Straw N uptake increased with increasing biochar application rate. Moreover, the relative increase induced by biochar amendment on straw N uptake was 5.62%, 10.06% and 12.87% for the C1, C2 and C3 treatments, respectively (Fig. 1c). In addition, total N uptake was increased by 5.43%, 10.53% and 12.61% in C1, C2 andC3, respectively, compared withC0 (Fig. 1d). -N leaching peaks were dampened. There were no major changes in the NO 3 − -N and NH 4 + -N leaching during the final two months (Fig. 2). The average NO 3 − -N leaching increased with soil depth, whereas NH 4 + -N leaching decreased with soil depth (Table 2). Biochar application significantly reduced the mean NO 3 − -N and NH 4 + -N leaching ( Table 2). C1 had significantly decreased NO 3 − -N leaching at a depth of 100 cm and NH 4 + -N leaching at depths of 60 cm and 100 cm, while C2 and C3 showed significantly decreased NO 3 − -N and NH 4 + -N leaching throughout the soil profile (Table 2). Furthermore, significant interactions were found between the observation period and biochar treatment, except for the NO 3 − -N leaching at 100 cm (Table 2).
Biochar amendment consistently increased the net GWP, and C2 and C3 showed the largest increases (29.01% and 25.13%, respectively) ( Table 3). However, the increase in net GWP elicited by the C1 treatment was only 3.24% (Table 3).
Abundance of soil AOA and AOB communities. Soil AOA amoA copy numbers exhibited a slight seasonal variation (Fig. 4a, Table 3, P = 0.074), while soil AOB amoA copy numbers showed a clear pattern of seasonal changes (Fig. 4b, Table 3, P = 0.011). Biochar amendment significantly increased soil AOA amoA copy numbers (Table 3, P < 0.001). The highest AOA amoA gene copy numbers were observed in the C3 treatment(12.32% greater than that of C0) (Table 3). Conversely, there was no significant difference in soil AOB amoA copy numbers among the four treatment (Table 3, P = 0.349).

Relationships between soil fluxes and ammonia-oxidizer abundances and soil inorganic N
concentrations. The soil CH 4 fluxes were negatively linearly correlated with the soil NO 3 − -N concentration ( Table 4). The soil N 2 O fluxes were negatively correlated with the soil AOA abundance and positively correlated with the soil NO 3 − -N concentration ( Table 4). The soil AOA abundance was negatively correlated with the soil  -N leaching at various soil depths affected by different treatments over the entire experimental period. Data are mean ± SE. Lowercase letter in the same column represents significant differences among experimental treatments at the level of 0.05.

NO 3
− -N concentration, whereas the soil AOB abundance was not correlated with any of the measured soil fluxes or inorganic N concentrations (Table 4).     Table 2). These results confirmed our first hypothesis that biochar amendment reduces soil NO 3 − -N and NH 4 + -N leaching. Increases in N retention or absorption in soil 29,43 and stimulation of crop N uptake 44 have generally been hypothesized to be the primary causes of reduced N leaching after biochar application. In our study, biochar amendment significantly increased the total soil N concentration and rice yield (Table 1), consistent with the meta-analysis results reported by Biederman and Harpole 45 . Furthermore, biochar amendment increased AOA activity (Table 4), producing more available N for crop growth and increased rice N uptake (Fig. 5). The reduced soil NO 3 − -N and NH 4 + -N concentrations decreased the inorganic N pool for leaching (Table 1). Therefore, the reduced soil NO 3 − -N and NH 4 + -N concentrations induced by biochar application may be the main cause of the reduction in NO 3 − -N and NH 4 + -N leaching (Fig. 5). In addition, the increased soil water holding capacity due to the reduced soil bulk density (Table 1)  Effects of biochar amendment on the soil CH 4 and N 2 O fluxes. The net CH 4 flux from soil is the sum of production and oxidation. The effects of biochar amendment on the CH 4 flux were thus unclear. In agreement with previous studies 24,47 , the application of biochar at rates of 9 t ha −1 yr −1 and 13.5 t ha −1 yr −1 significantly increased the soil CH 4 flux by 35.16% and 40.62%, respectively ( Table 3). The promotion of the soil CH 4 flux was comparable to that of the Tai Lake plain in China for biochar amendment at rates of 10 t ha −1 and 40 t kg ha −1 23 . This was attributed to the following three aspects. First, the soil NH 4 + -N accumulation decreased soil CH 4 oxidation by altering the activity and composition of the methanotrophic community 48 . However, biochar amendment did not significantly change the soil NH 4 + -N accumulation (Table 1), and there was no significant relationship between the soil CH 4 flux and soil NH 4 + -N concentration (Table 4). Wang et al. 49 found that soil NO 3 − -N accumulation could significantly promote soil CH 4 uptake. The lower soil CH 4 uptake was due to the decreased soil NO 3 − -N concentration under biochar amendment (Table 1), which increased the soil CH 4 emissions in our study ( Table 3, Fig. 5). The significant negative relationship between the soil NO 3 − -N concentration and the soil CH 4 flux reflected this prediction (Table 4). Second, methanotrophs use the sorbed organic compounds in addition to CH 4 because methanotrophs can utilize a variety of substrates 50 . Therefore, biochar amendment reduced the net soil CH 4 oxidation 51 . Third, Knoblauch et al. 52 reported that the labile components of biochar increase the substrate supply and create a favorable environment for methanogens 53 . The lower pH in our biochar plots may have promoted methanogenic archaea, which have an optimal pH of 7 54 . Thus, a larger archaeal population may temporarily have increased CH 4 emissions in the biochar treatment 55 until the emissions declined due to the oxic environment. However, because we only measured the net effects, we could not distinguish between reduced soil methanotrophic activity and increased methanogenic activity in this study.
Suppression of soil N 2 O emissions following biochar amendment has been observed both under laboratory conditions 25,26 and in the field 23,47 . Enhanced soil aeration 27 , altered ammonia-oxidizer and denitrifier activity 56 , sorption of NH 4 + -N or NO 3 − -N by biochar 12 and the presence of inhibitory compounds such as ethylene 57 have been suggested as mechanisms to explain the reduction inN 2 O flux with biochar amendment. In anaerobic  Table 4. Correlation coefficients (R 2 ) for the relationships among soil CH 4   paddy soil, N 2 O production from denitrification is thought to be the dominant source. Baggs 58 suggested that decreased total N denitrification and enhanced reduction of N 2 O to N 2 can lead to lower N 2 O denitrification in soil. A reduction in NO 3 − -N availability would decrease the total denitrified N and would reduce the ratio of N 2 O/(N 2 + N 2 O) 17 . In our study, biochar amendment reduced the soil NO 3 − -N concentration by accelerating ammonia oxidation (Fig. 4, Table 3) and promoting total N uptake in rice (Fig. 1). The significant negative correlation between the soil NO 3 − -N concentration and soil AOA abundance provided direct evidence for this result (Table 4). Soil NO 3 − -N availability was positively correlated with soil N 2 O flux (Table 4), which could partially explain the reduction in soil N 2 O flux in the studied paddy soil (Fig. 5). In addition, evidence for the decreased N 2 O denitrification was provided by the decreased soil bulk density after biochar amendment (Table 1). Soil pH was not significantly changed after biochar amendment at our study site (Table 1). However, soil Eh was not monitored when the gas samples were taken in situ. Mechanisms associated with the oxidation and reduction of nitrogen species need to be studied under the alternating redox conditions on the surface of biochar.
However, our flux results were drawn from relatively few measurements. The gas fluxes were measured on 19 occasions during 150 days of rice growth, with an average sampling interval of 8 days. This frequency was slightly lower than the average 6.5 day interval of the reviewed rice paddy studies in China with single biochar amendment and biochar combined with NPK compound fertilizer (Supplementary Table S1). The effects of biochar amendment on the soil CH 4 flux are contradictory, including positive, negative, and neutral effects. Moreover, biochar amendment is mainly reported to decrease soil N 2 O flux. Increased soil CH 4 flux and decreased soil N 2 O flux were observed in our study. This evidence suggests that our conclusions can be drawn from relatively few measurements.
Biochar is considered to be an effective tool to mitigate GWP via carbon sequestration in soil and to influence carbon mineralization through priming effects. In the present study, the changes in net GWP were not significant between C1 and C0 treatments, while biochar amendment significantly increased paddy soil CH 4 emissions and decreased N 2 O emissions in the C2 and C3 treatments, leading to a significantly increased net GWP (Table 3). This result is in line with that of a rice paddy from Tai Lake plain, China 23 . Our results indicate that higher amounts of biochar amendment are associated with a risk of increased paddy net GWP in northwest China (Fig. 5). The stimulation effects of biochar on native soil organic carbon mineralization decreases with time due to the depletion of labile SOC from the initial positive priming and stabilization of native SOC via biochar-induced organ-mineral interactions during 5-year laboratory incubations 59 . In another 120-day incubation study, the time effects showed an initial increase followed by a decrease and stabilization, resulting in a significantly increased sequestration of carbon in the soil over the long term compared with conventional biowaste amendments 60 . In our study, improved rice yield and reduced soil nitrous oxide emissions contributed to the mitigation potential of biochar amendment. However, the response of native soil organic carbon mineralization to biochar amendment remains uncertain. Relative priming effects and cumulative CO 2 emissions studies are needed to evaluate the GWP of biochar amendment in our paddy soil.

Conclusions
The effects of biochar amendment on soil N leaching and soil CH 4 and N 2 O fluxes were investigated in paddy soil in northwest China. We found that biochar amendment significantly decreased soil NO 3 − -N and NH 4 + -N leaching, but that the C2 and C3 treatments significantly increased soil CH 4 emissions and reduced N 2 O emissions, leading to significantly increased net GWP. Biochar amendment significantly increased soil AOA abundance and rice N uptake. Soil NO 3 − -N availability can explain the responses of soil N leaching and soil CH 4 and N 2 O fluxes to biochar amendment. Our results indicated are commended dose of biochar amendment of 4.5 t ha −1 yr −1 with conventional N application in the study area. The responses of the soil CH 4 and N 2 O fluxes to biochar amendment were also influenced by the interannual variability in climate, temperature and precipitation. The long-term effects of biochar amendment on N leaching and net GWP in rice production required further investigation to identify the most cost-effective and environmentally friendly management practices for rice culture.