Response of N2O emissions to biochar amendment in a cultivated sandy loam soil during freeze-thaw cycles

In the last decade, an increasing number of studies have reported that soil nitrous oxide (N2O) emissions can be reduced by adding biochar. However, the effect of biochar amendment on soil N2O emissions during freeze-thaw cycle (FTC) is still unknown. In this laboratory study, biochar (0%, 2% and 4%, w/w) was added into a cultivated sandy loam soil and then treated with 15 times of FTC (each FTC consisted of freeze at −5/−10 °C for 24 h and thaw at 5/10 °C for 24 h), to test whether biochar can mitigate soil N2O emissions during FTC, and estimate the relationships between N2O emissions and soil inorganic nitrogen contents/microbial biomass content/enzyme activities. The results showed that biochar amendment suppressed soil N2O emissions by 19.9–69.9% as compared to soils without biochar amendment during FTC. However, N2O emissions were only significantly correlated to soil nitrate nitrogen (NO3−-N) contents, which decreased after biochar amendment, indicating that the decreased soil nitrification by adding biochar played an important role in mitigating N2O emissions during FTC. Further studies are needed to estimate the effectiveness of biochar amendment on reducing freeze-thaw induced N2O emissions from different soils under field conditions.

might induce significant increases in soil N mineralization 15,16 . A recent study by Case et al. 17 showed that adding biochar into soils could also stimulate soil N mineralization and nitrification, while suppressed cumulative production of N 2 O by 91%. Therefore, the relationship between soil inorganic N contents and N 2 O emissions may be complicated after biochar amendment during FTC.
Aside from the availability of substrates, soil microbial biomass and enzyme are two other important factors that influence soil N 2 O emissions because they involve in catalytic reactions and nutrient mineralization 18,19 . For instance, Wick et al. 20 reported that soil N 2 O emissions were positively correlated to soil microbial biomass N (SMBN) and β -glucosidase activities during a dry season; Bai et al. 21 found that soil urease activity was an indicator of N 2 O emission because of the close relationship between urease activity and nitrification. Previous studies have demonstrated that soil microbial biomass and enzyme activities can be changed by adding biochar [22][23][24] . However, the observation periods of these studies all focused on the growing season. During FTC, the dynamics of soil microbial biomass and enzyme activities as well as their relationships with soil N 2 O emission after adding biochar are poorly understood.
Soils in mid-high latitude regions are projected to experience higher frequencies and larger amplitudes of FTC in the context of climate change, which in turn release more N 2 O into the atmosphere 25 . Although biochar amendment is a potential amendment to mitigate soil N 2 O emissions, very limited information is available on the effect of biochar amendment on soil N 2 O emission during FTC. The objectives of this laboratory study were to: (1) investigate the effects of biochar amendment on soil N 2 O emission, inorganic N contents, microbial biomass and enzyme activities during FTC; (2) estimate the relationships between N 2 O emissions and soil inorganic N contents/soil microbial biomass contents/enzyme activities under the joint effects of FTC and biochar amendment. More specifically, we tested the hypothesis that the N 2 O bursts during FTC will be suppressed by biochar amendment.

Materials and Methods
Soil collection and analysis. In May 2015, soil samples (0-20 cm) were collected from a farmland (43°27′ N, 82°54′ E) cultivated with corn in the Ili River Valley, Xinjiang Uygur Autonomous Region, northwest China. The surface soils in this area usually experience seasonal freeze-thaw process during early spring. The soil was classified as Typic Haploboroll (USDA) with a sandy loam texture (4.2%, 23.2% and 72.6% for clay, silt and sand, respectively). Collected soil samples were air-dried, homogenized and grounded to pass through a 2 mm nylon fiber sieve before experimental use.
Soil organic C (SOC) was measured using the H 2 SO 4 -K 2 Cr 2 O 7 oxidation method, while soil total N (TN) was detected using an automatic azotometer (Kjeltec 8400, FOSS, Denmark) according to the Kjeldahl method. Soil ammonium N (NH 4 + -N) and nitrate N (NO 3 − -N) were determined using a continuous flow analyzer (AA3, SEAL Analytical, Germany) with 0.01 M CaCl 2 extracts (1:10, w/v) 26 . Soil pH and electric conductivity were measured in a volume ratio (H 2 O) of 1:5 (w/v) using a pH meter (SevenEasy, Mettler-Toledo, Switzerland) and an electric conductivity meter (DDSJ-308A, Rex, China), respectively. Soil texture was analyzed using a laser diffraction particle analyzer (Mastersizer 2000, Malvern, UK). SMBN was measured using the chloroform fumigation-K 2 SO 4 extraction method (1:4, w/v). The extracts were analyzed at 280 nm using an UV spectrophotometer (Cary 60, Agilent Technologies, USA) 27 . Activities of urease and protease were determined using the indigo colorimetric method and the ninhydrin colorimetric method with urea and casein as substrates, respectively 28,29 . Urease and protease activities were expressed as μ g NH 4 -N g −1 h −1 and μ g Tyr g −1 h −1 , respectively. The physicochemical properties of soil are shown in Table 1.
Biochar analysis. Biochar used for the experiment was made by the Seek Bio-Technology Company, Shanghai, China. It was produced using bamboo under a pyrolysis of 500-600 °C. The biochar was grounded to pass through a 2 mm nylon fiber and mixed thoroughly before experimental use.
Surface structure and elemental analysis of biochar were examined using a scanning electron microscope (Super 55VP, Zesis, Germany) associated with an energy dispersive X-ray spectroscopy (XFlash 5010, Bruker, Germany) (Fig. 1). The pH, electric conductivity, NH 4 + -N and NO 3 − -N of biochar were determined using the

Properties Soil Biochar
Organic C (g C kg −1 ) 11.0 ± 0. previously mentioned methods. To estimate ash content, 1.0 g of biochar was heated in a muffle furnace (LC-502, Koyo, Japan) at 500 °C for 8 h. The ash content was calculated from: ash content (%) = mass of ash/mass of biochar × 100 30 . The elemental (C, H, N and S) contents of biochar were measured using an elemental analyzer (vario MICRO cube, Elementar, Germany). The O content of biochar was determined by calculating the difference between 100% and the sum contents of ash, C, H, N and S. The physicochemical properties of biochar are also shown in Table 1.
Experimental design. In 250 mL Erlenmeyer flasks, 60.0 g (oven-dry basis) of soils were mixed with 0% (BC0), 2% (BC2) and 4% (BC4) (w/w) of biochar, and then wetted with deionized water to reach 60% of water holding capacity. Each flask was covered with parafilms with several small holes to allow gaseous exchange and reduce the loss of soil water. All flasks were pre-incubated at 25 °C in dark condition for seven days. After that, flasks of each application rate were separated into two equal groups to experience different FTC treatments. Taking the field surface temperatures during spring freeze-thaw periods into consideration 31 , we set up two experiments with one in a small amplitude (− 5 °C to 5 °C) of FTC (SFT) and the other in a large amplitude (− 10 °C to 10 °C) of FTC (LFT). There were fifteen times of FTC in total and each included freeze at − 5 or − 10 °C for 24 h and thaw at 5 or 10 °C for 24 h. Six treatments in the present study were established and after the 1st, 3rd, 5th, 10th and 15th FTC, triplicate flasks of each treatment were randomly selected and destructively sampled. Soils were used for determining NO 3 − -N, NH 4 + -N and SMBN contents together with urease and protease activities. Furthermore, deionized water was added into each flask at the end of every two FTCs to compensate for the lost soil water. Table 2 shows the details of treatment layout and properties of the mixtures after pre-incubation.
Gas sampling and analysis. Triplicate flasks for each treatment were sealed with rubber stoppers to collect gas samples at the end of 1st, 3rd, 5th, 10th and 15th FTC. In the middle of the stopper, a small hole was made and a plastic tube (0.2 cm in inner diameter, 10 cm in length) connected to a three-way stopcock was inserted into the hole. The gaps between stopper and tube were sealed with glue. The three-way stopcock was closed to make a gas-tight environment after covering. During a half hour closure period, a gas sample of approximately 2.5 mL was withdrawn using a gas-tight syringe at 0, 10, 20 and 30 min, respectively. The concentrations of N 2 O were detected within 3 days using a gas chromatograph (7890B, Agilent Technologies, USA), which was equipped with an electron capture detector. The carrier gas for N 2 O analysis was high-purity N 2 . N 2 O emissions were calculated using formula (1) 32 : where ρ is the density of N 2 O at 0 °C (1.963 g m −3 ), V (m 3 ) and W (kg) are the head space volume of the flask and the soil weight, respectively, Δ C is the change in N 2 O concentrations during the measurement period Δ t (h), and T is the absolute temperature. Cumulative emissions during the whole incubation were directly computed from the measured emissions and estimated by linear interpolation for days when no measurements were made.
Statistical analysis. Three-way ANOVA was used to examine the differences in soil N 2 O emissions, NO 3 − -N, NH 4 + -N, SMBN contents, urease and protease activities among FTC amplitudes, biochar addition rates and FTC frequencies. Differences in cumulative N 2 O emissions between FTC amplitudes were tested using independent-samples t test, while differences in cumulative N 2 O emissions among biochar addition rates were examined using one-way ANOVA. Data sets have gone through the normality and heterogeneity tests and were converted to log-transformation (base 10) when the variances were unequal before analyses. Pearson correlation was employed to examine the correlations among N 2 O emissions, NO 3 − -N, NH 4 + -N, SMBN contents as well as urease and protease activities. Differences and correlations were considered statistically significant if P < 0.05 and highly significant if P < 0.01.

NO 3
− -N, NH 4 + -N and SMBN contents. As shown in Fig. 2, soil NO 3 − -N contents of each treatment all increased during the incubation. As compared to the contents after pre-incubation, soil NO 3 − -N contents increased by 13.6-23.7% and 24.7-41.0% after 15 times of FTC under SFT and LFT, respectively. The largest increase was detected in LFT-BC0. Results of ANOVA analysis showed that soil NO 3 − -N content was significantly affected by FTC amplitude, biochar addition rate as well as FTC frequency (Table 3). Soil NH 4 + -N contents showed decreasing trends for all treatments in the first five FTCs and then slowly increased throughout the rest of FTCs (Fig. 2c,d). After the incubation, NH 4 + -N contents decreased by 29.1-54.6% and 25.4-49.0% under SFT and LFT compared with the contents after the pre-incubation, respectively. The effects of FTC amplitude and biochar addition rate on soil NH 4 + -N content were not significant when ignoring their interaction effects with FTC frequency (Table 3).
FTC amplitude and frequency had significant impacts on SMBN content ( Table 3). As compared to the contents after the pre-incubation, 15 times of FTC decreased SMBN contents by 0.1-7.7% under SFT, while the range of SMBN content was only 61.2 to 81.3 mg kg −1 during the whole incubation. By contrast, LFT showed stronger effects on decreasing SMBN contents than SFT. SMBN contents of all treatments under LFT generally decreased with the increase of FTC times (Fig. 2e,f). After the 15th FTC, SMBN contents of LFT-BC0, LFT-BC2 and LFT-BC4 decreased to 54.5, 51.9 and 53.1 mg kg −1 , respectively, which were the minimums of each treatment. However, biochar addition rate did not show a significant effect on SMBN content when ignoring its interaction effects with FTC amplitude and frequency (Table 3).
Enzyme activities. Similar to soil NO 3 − -N, soil urease activity was significantly affected by FTC amplitude, biochar addition rate and FTC frequency (Table 3). Under SFT, soil urease activities of each treatment decreased continuously from the 1st FTC to the 5th FTC, but reversed in the remaining FTCs (Fig. 3a). After the 15th FTC, soil urease activities of BC0, BC2 and BC4 increased by 7.5%, 14.3% and 7.5%, respectively, as compared to their activities after the pre-incubation. Soil urease activities under LFT varied from 21.9 to 30.1 μ g NH 4 -N g −1 h −1 and displayed an initial decrease but subsequently increased with the increase of FTC times. Furthermore, soils amended with biochar always showed higher urease activities than BC0 in both FTC conditions. The mean activities decreased in the following order: BC4 > BC2 > BC0.
In most treatments, soil protease activities decreased as FTCs proceeded (Fig. 3c,d). In comparison with the activities after the pre-incubation, soil protease activities decreased by 17.8-25.1% and 10.5-25.5% after 15 times of FTC under SFT and LFT, respectively, with the largest decrease found in LFT-BC0. Soil protease activity was also significantly affected by biochar addition rates (Table 3), and soil protease activities of BC2 and BC4 were higher than those of BC0 during most of the incubation time. However, neither FTC amplitude nor its interaction effects with other factors significantly influenced protease activities (Table 3). Table 3, soil N 2 O emission was significantly affected by FTC amplitude, biochar addition rate, FTC frequency, as well as their interaction effects. Under SFT, soil N 2 O emissions of all treatments were low with a range of 0.1 to 0.2 μ g N 2 O kg −1 h −1 after the 1st FTC. Thereafter, sharp increases in N 2 O emissions were observed for all treatments (Fig. 4). The emissions after the 3rd FTC were 3.3, 3.1 and 6.7 times higher than those after the 1st FTC for BC0, BC2 and BC4, respectively. During the rest of FTCs, N 2 O emissions of each treatment showed decreasing tendencies. The emissions of BC0 were generally higher than those of BC2 and BC4. Soil N 2 O emissions of each treatment also showed considerable changes with the increase of FTC times under LFT. For BC2 and BC4, soil N 2 O emissions fluctuated as the incubation continued and the peak emissions were observed after the 3rd FTC. By contrast, soil N 2 O emissions of BC0 first decreased between the 1st FTC and the 3rd FTC, then increased drastically after the 5th FTC, and finally showed a decreasing trend during the last 10 FTCs. Its peak emission was 2.5 μ g N 2 O kg −1 h −1 , which was about 2.0 and 2.1 times higher than peak emissions of BC2 and BC4, respectively.

N 2 O emissions. As shown in
Soil cumulative N 2 O emissions during the whole incubation showed significant differences among different biochar addition rates (Fig. 5). In comparison with BC0, biochar amendments decreased soil N 2 O emissions by 19.9% (BC2) and 37.3% (BC4) under SFT, and by 41.5% (BC2) and 69.9% (BC4) under LFT. Furthermore, LFT induced about 1.7 to 3.4 times higher soil N 2 O emissions than SFT when soils were treated with same biochar addition rates. The highest and lowest cumulative N 2 O emissions were observed in LFT-BC0 and SFT-BC4, respectively. N and SMBN 35,36 . Therefore, the decreased soil NH 4 + -N contents might be also attributed to the volatilization of NH 4 + -N because the soil pH was high in our study (Table 2). Although the effects of biochar on soil N dynamics have been widely investigated, information on how biochar affects soil inorganic N contents during FTC is still limited. Our results showed that biochar amendments significantly decreased soil NO 3 − -N content while had little effect on soil NH 4 + -N content as compared to BC0 during FTC, suggesting that soil nitrification may be inhibited by adding biochar when FTC occurs. In a laboratory study, Zhang et al. 37 found that both soil NH 4 + -N and NO 3 − -N contents decreased with adding biochar. They suggested that biochar had the ability to adsorb soil inorganic N, and then led to decreases in soil nitrification and net N mineralization. Christenson et al. 38 observed a significant negative relationship between net nitrification and soil C/N ratio. They suggested that the low gross NH 4 + -N production or higher NO 3 − -N consumption were the possible reasons for  this phenomenon. As shown in Table 2, soil C/N ratio increased from 9.1 to 12.4 or 15.5 by biochar amendments. Therefore, the increased soil C/N ratio might be another possible reason for the decreases in soil nitrification. Changes in soil microbial biomass can reflect the process of microbial growth, death, and the degradation of soil organic matter 19 . Similar to our results, previous studies have also reported that soil microbial biomass contents could be reduced by FTC 10,39 . Such decreases may be attributed to that FTC has a sterilization function, which kills soil microorganisms during freeze periods 40 . Moreover, our results indicated that bicohar amendments had little effects on SMBN contents during FTC. Although the effects of biochar on soil microbial biomass have been extensively investigated, the existing results are still disputable. Most of the related studies reported that soil microbial biomass could be increased by adding biochar 22,41 . Some studies demonstrated that biochar is a porous material, which has many pores, especially macropores (> 200 nm) on its surface 42,43 . These macropores may hold substrates and serve as favorable habitats for soil microorganisms 43 . However, contrary results were also reported. For example, Dempster et al. 23 pointed out that biochar could decrease soil microbial biomass C but not influence SMBN in a pot study. These differences may be explained in part by variations in biochar rate and type (e.g. biochar feedstock, pyrolysis temperature, etc.) along with soil types. Impacts of biochar amendment on soil enzyme activities during FTC. Soil enzymes play critical roles in maintaining nutrient availability. Their activities are "sensors" of microbial status and soil physicochemical conditions 18 . In this study, both urease and protease activities decreased during the first five FTCs, suggesting that FTC had a short-term effect on decreasing soil enzyme activities. Previous studies suggested that the decreased enzyme activities during FTC might be attributed to the decreased microbial activities because soil enzymes mainly originate from soil microorganisms 10,44 . As an example, Wang et al. 10 reported that soil enzyme activities were significantly correlated to soil microbial biomass C contents during FTC. Similarly, soil enzyme activities also showed significant correlations with SMBN in this study (Table 4), partially supporting the assertion. However, soil enzyme activities were quite stable or even increased after the 5th FTC, suggesting that soil enzymes or microorganisms had probably been adapted to the FTC conditions.

Impacts of biochar amendment on soil inorganic
During most of the incubation period, soil urease and protease activities of BC2 and BC4 were generally higher than those of BC0, indicating that biochar might help retaining soil enzyme activities during FTC. This result was in agreement with previous studies which also found that soil enzyme activities were increased by adding biochar 22,45 . The potential mechanisms of these increases may be: (1) the macropores of biochar serve as favorable habitats for soil microorganisms and protect them from being killed by FTCs; (2) the increased . Uppercase letters above the bars indicate significant differences among biochar addition rates while under a same FTC amplitude after one-way ANOVA with LSD test (n = 3, P < 0.05). Lowercase letters above the bars indicate significant differences between FTC amplitudes while under a same biochar addition rate after independent-samples t test (n = 3, P < 0.05)  substrates induced by FTC are fixed on the surface of biochar and can be used by soil microorganisms. However, contrary reports that biochar had no effects or even negative effects on increasing soil enzyme activities also existed 24,46 . Hence, more studies of biochar amendment on soil enzyme activities are needed to understand its effect as well as its underlying mechanisms.

Impacts of biochar amendment on soil N 2 O emissions during FTC.
It has been demonstrated that soil N 2 O emissions during freeze-thaw periods are an important part of the annual N 2 O budget. Our results showed that biochar amendment suppressed N 2 O emissions by 19.9-69.9% as compared to BC0 during FTC, suggesting that biochar amendment might be a potential way to mitigate soil N 2 O emissions during FTC. The results of Pearson correlation analysis showed that soil N 2 O emissions were significantly correlated to soil NO 3 − -N content ( Table 4). As illustrated above, soil nitrification, which converted NH 4 + -N to NO 3 − -N, might occur during the incubation. In addition, soil NO 3 − -N contents of BC2 and BC4 were significantly lower than those of BC0. Therefore, the suppression of N 2 O emissions might be related to the nitrification, which was inhibited by biochar amendments. Some studies indicated that biochar contains volatile organic compounds such as α -pinene and ethylene, which are known as nitrification inhibitors 47,48 . Similar to our results, Sarkhot et al. 49 pointed out that biochar amendments led to 68-75% and 26% reductions in net nitrification and N 2 O emission, respectively. They suggested that such reductions were a result of soil inorganic N adsorption. However, Case et al. 50 hypothesized that the decreased soil cumulative N 2 O productions by biochar were related to the biological or physical immobilization of NO 3 − -N, which removed large amounts of NO 3 − -N from the extractable pool. Therefore, the adsorption of soil inorganic N as well as the 15 N tracer experiments are suggested to be designed in the future to have a better understanding of the mechanisms of N 2 O suppression by biochar amendment during FTC.

Conclusion
Results of the present study showed that soil N 2 O emissions from a cultivated sandy loam soil could be suppressed by adding biochar during FTC. The decreased nitrification indicated by the lower soil NO 3 − -N contents in the biochar treatments was found to play an important role in such suppressions. Biochar amendments also had a positive effect on retaining soil urease and protease activities, while it did not affect NH 4 + -N and SMBN contents during FTC. However, soil NH 4 + -N and SMBN contents, urease and protease activities did not show significant correlations to soil N 2 O emissions. Our study indicates that biochar amendment can be a potential method to reduce soil N 2 O emissions during freeze-thaw periods. Although extrapolation of the findings from this short-term laboratory study to long-term field results should be conducted with caution, the results still gave an insight into how biochar affects soil N 2 O emissions during FTC. Further studies are needed to estimate the effectiveness of biochar amendment on reducing freeze-thaw induced N 2 O emissions from different soils under field conditions.