Nitrapyrin addition mitigates nitrous oxide emissions and raises nitrogen use efficiency in plastic-film-mulched drip-fertigated cotton field

Nitrification inhibitors (NIs) have been used extensively to reduce nitrogen losses and increase crop nitrogen nutrition. However, information is still scant regarding the influence of NIs on nitrogen transformation, nitrous oxide (N2O) emission and nitrogen utilization in plastic-film-mulched calcareous soil under high frequency drip-fertigated condition. Therefore, a field trial was conducted to evaluate the effect of nitrapyrin (2-chloro-6-(trichloromethyl)-pyridine) on soil mineral nitrogen (N) transformation, N2O emission and nitrogen use efficiency (NUE) in a drip-fertigated cotton-growing calcareous field. Three treatments were established: control (no N fertilizer), urea (225 kg N ha-1) and urea+nitrapyrin (225 kg N ha-1+2.25 kg nitrapyrin ha-1). Compared with urea alone, urea plus nitrapyrin decreased the average N2O emission fluxes by 6.6–21.8% in June, July and August significantly in a drip-fertigation cycle. Urea application increased the seasonal cumulative N2O emission by 2.4 kg N ha-1 compared with control, and nitrapyrin addition significantly mitigated the seasonal N2O emission by 14.3% compared with urea only. During the main growing season, the average soil ammonium nitrogen (NH4+-N) concentration was 28.0% greater and soil nitrate nitrogen (NO3--N) concentration was 13.8% less in the urea+nitrapyrin treatment than in the urea treatment. Soil NO3--N and water-filled pore space (WFPS) were more closely correlated than soil NH4+-N with soil N2O fluxes under drip-fertigated condition (P<0.001). Compared with urea alone, urea plus nitrapyrin reduced the seasonal N2O emission factor (EF) by 32.4% while increasing nitrogen use efficiency by 10.7%. The results demonstrated that nitrapyrin addition significantly inhibited soil nitrification and maintained more NH4+-N in soil, mitigated N2O losses and improved nitrogen use efficiency in plastic-film-mulched calcareous soil under high frequency drip-fertigated condition.

Introduction As a kind of potential greenhouse gas (GHG), nitrous oxide (N 2 O) causes climate change and indirectly contributes to stratospheric ozone depletion [1], and its global warming potential (GWP) is about 298 times that of carbon dioxide (CO 2 ) over a 100 yrs time horizon [2]. In China, agricultural intensification and heavy fertilizer use have led to an increase in N 2 O emission losses, the contribution of croplands to the national total of fertilizer N-induced N 2 O (FIE-N 2 O) increased from 79% to 92% only in the period of 1980-2000 [3]. Thus, reducing N 2 O emissions from agricultural sources constitutes a critical research challenge.
Agricultural N 2 O emission has long been studied, however, whether N 2 O is mainly derived from nitrification process under aerobic conditions, or from denitrification process under either anaerobic or aerobic conditions is still inconclusive. The relative contribution of each process to N 2 O is largely dependent upon soil environmental conditions (e.g. soil organic matter, pH, soil moisture, soil N status, etc.). For instance, several previous studies suggest that nitrification accounts for most N 2 O emission in well-aerated soil below 60% water-filled pore space (WFPS), whereas denitrification reaction accounts for most N 2 O emission when soil moisture exceeds 70% WFPS [4][5]. Recent studies show that denitrification prevails over nitrification during N 2 O production for irrigated dryland soils or more humid soils initially [6][7].
Many researches demonstrate that N 2 O emission can be substantially mitigated by nitrification inhibitors (NIs) through reducing the rate of ammonium nitrogen (NH 4 + ) oxidation into nitrate nitrogen (NO 3 -) during nitrification, and subsequently decreasing the substrate (NO 3 -) concentration for denitrificationin different agro-ecosystems [8][9][10][11]. Otherwise, NIs have been used to improve crop N status [8,12] and increase soil mineral N retention [13]. As one of the effective nitrification inhibitors, nitrapyrin (2-chloro-6-(trichloromethyl)-pyridine) exhibited some effectiveness on reducing nitrate leaching and N 2 O emission losses, improving yield and N retention in the grain and corn field or under the simulated experimental condition [13][14][15][16]. We previously showed that nitrapyrin was more effective than dicyandiamide (DCD) in controlling nitrification process, and nearly as effective as 3, 4-dimethylpyrazole phosphate (DMPP) in calcareous soil with sandy, loam and clay texture [17]. Although numerous studies have been undertaken on the influences of different NIs on N 2 O emission and nitrogen transformations, how repeated supply of nitrapyrin with urea via fertigation impacts soil N 2 O emissions, nitrogen transformation and utilization under drip-irrigated condition is still unknown. Therefore, it is necessary to clarify the influences of repeated addition of nitrapyrin on soil N 2 O emissions and its major affecting factors in plastic-film-mulched drip-fertigated cotton field.
Xinjiang, located in northwest of China, is an important agricultural area. Due to the arid climate with annual precipitation of 160.6 mm on average, conventional agriculture in this region depends heavily on irrigation. To save fresh water resource and improve water utilization efficiency, a combination of plastic film-mulching with drip irrigation has been extensively adopted in this region during the last two decades. In this system, N fertilizer is applied about 8-10 times coupling with water via the irrigation system during a growing season. As a consequence, the re-wetting of the topsoil frequently occurs, and soil N transformation process in drip irrigation condition substantially differs from that in conventional cultivation condition (i.e. flooding irrigation and single fertilizer application).
The objectives of this study were to explore the effects of nitrapyrin added to the urea solution on N 2 O emission, soil mineral nitrogen transformations and nitrogen use efficiency in the plastic-film-mulched and high frequency drip-fertigated cotton field, and ascertain the major factors impacting on N 2 O emissions in this system. The results will provide a basis for decreasing nitrogen losses and promoting efficient utilization of nitrogen fertilizer through adding nitrapyrin to urea solution in drip-fertigated cropland in arid regions.

Study site and soil characteristic
The field trial was set up at an experiment station of Agricultural College, Shihezi University, Xinjiang, China (44˚18 0 N, 86˚02 0 E) in April 2013. This site has a typical temperate continental climate and the annual precipitation varies between 125.0 and 207.7 mm.
The soil at the site is a Calcaric Fluvisals consisting of 24.7% sand, 20.1% silt and 48.8% clay. The initial physicochemical properties are: organic matter content, 16.2 g kg -1 ; total N content, 0.92 g kg -1 ; ammonium nitrogen content, 28.0 mg kg -1 ; nitrate nitrogen content, 40.9 mg kg -1 ; soil available phosphorus, 11.0 mg kg -1 and available potassium, 257.5 mg kg -1 . The soil is alkaline (pH 7.96) with bulk density of 1.42 g cm -3 (0-15 cm layer). The groundwater is generally used for irrigation in this region.

Experimental design and field management
This experiment consisted of a randomized block field experiment with three treatments: (i) control (no N fertilizer), (ii) urea and (iii) urea+nitrapyrin. Each treatment was replicated three times. The area of each plot was 45.9 m 2 (i.e. 5.4 m × 8.5 m), and 12 rows of cotton plants (Gossypium hirsutum cv. Xinluzao 45) were grown in each plot and the row spacing was 30 cm-50 cm-30 cm, and there was a 12 cm space between plants within a row (Fig 1). For each plot, a drip irrigation tape was placed in the middle of two rows, and six drip irrigation tapes were connected to a branch pipe controlled by a valve with a small fertilizing tank and a water meter installed to precisely control the irrigation volume and fertilizer rate.
A total of 465 mm irrigation water was applied to each plot in 10 split application times mainly in June, July, and August ( Table 1). The nitrogen, phosphorus, and potassium fertilizers (urea and potassium dihydrogen phosphate) were dissolved in the irrigation water and applied during the first eight irrigation events. No urea was added to the irrigation water in the control treatment. The urea and urea+nitrapyrin treatments both received a total of 225 kg N ha -1 , and all the treatments received a total of 90 kg ha -1 phosphorus pentoxide (P 2 O 5 ) and 60 kg ha -1 potassium oxide (K 2 O) during the cotton growing season. Nitrapyrin, which was supplied by Zhejiang Aofutuo Chemical Ltd, P. R. China, was mixed into the irrigation water in the urea+nitrapyrin treatment at a rate of 1% urea-N. More detailed information about the schedule and rate of irrigation and fertilization was shown in Table 1.

N 2 O sampling and measurement
Gas samples for N 2 O analysis were collected using static chambers consisting of top and middle chamber part, each being 60 cm×50 cm×55 cm for L×W×H in size to adapt to the height of the growing plants, and the final effective height reaching 110 cm. There were two chambers in each plot, the base of each chamber being inserted 10 cm deep into the soil to prevent air leakage, and each chamber having two fans to mix the air. The chambers were centred over two drip irrigation emitters and covered two rows of cotton plants (Fig 1). Gas samples were collected between 9:00 and 11:00 am on the 1, 2, 3, 4, 5 and 7 d after drip fertigation in June, July and August (fertilization season); four gas samples were collected from each chamber using 50 ml disposable syringes at 0, 10, 20 and 30 min after the chambers' coverage. Meanwhile, soil temperature was measured using soil thermometers inserted to a depth of 5 cm inside the chambers.
The N 2 O concentrations in the gas samples from the field were determined within 12 h of collection using an Agilent 7890A gas chromatograph (Agilent Technologies Ltd, USA) equipped with a 2-mm ID stainless steel column, 3-m long and packed with Porapak Q (80/ 100 mesh), fitted with an electron capture detector (ECD) set at 300˚C. The column temperature was maintained at 40˚C and the carrier gas was N 2 (99.999% purity) at a flow rate of 30 ml min -1 .

Soil sampling and measurement
The determination of the initial soil organic matter, total N, available phosphorus, available potassium and pH was based on Lu [18]. In parallel with the sampling of N 2 O gas, surface soil sampling (0-15 cm in depth) was collected for the determination of soil mineral N (NH 4 + -N and NO 3 --N) concentrations and soil moisture. The soil samples were collected at six points randomly and mixed together to form one composite sample for each plot. The soil was sieved (2 mm mesh) and 5.00 g of fresh soil was extracted with 50 ml 2 M KCl solution for 1 h by shaking on a reciprocating shaker, then the extracts were filtered through N-free quantitative

Nitrogen determination and yield measurement in cotton plants
Five cotton plants were randomly sampled from the harvested cotton plants and separated into stems, leaves and reproductive organs (buds, flowers, bell shells, seeds and fibres). The plant organs were oven-dried at 105˚C for half an hour, and then at 75˚C for another 72 hours prior to dry matter measurement. The dried samples were crushed and sieved (0.5 mm mesh) for N determination. The content of N was determined on Kjeltec 8400 Automatic Kjeldahl Nitrogen Determination Apparatus (FOSS Analytical Ltd, U.S.A.). Cotton yield was measured by weighing all of the cotton (fibres and seeds) in each plot at maturity, and the lint yield (only fibres) was calculated according to the cotton yield multiplied by lint percent.
Estimations N 2 O emission was calculated using Eq (1): , V is the mole volume (22.4 L mol −1 ) at 273 K and 1.013×10 5 Pa. H (m) is the effective chamber height (1.10 m in this experiment). P is air pressure in static chamber (Pa), P 0 is the ambient air pressure at the experimental site (1.013×10 5 Pa), P/P 0 %1. T is the temperature in the chamber (˚C). c (ppbv) is the volume concentration of N 2 O, t (h) is the time of chamber closure, and dc/dt (ppbv h -1 ) is the rate of change in the N 2 O volume concentration in the chamber enclosure [19]. Cumulative N 2 O emission was calculated by Eq (2): Where CE is N 2 O accumulative emission rate (kg N ha -1 ), F is N 2 O emission flux (μg N m -2 h -1 ), i represents the ith times of N 2 O determination, (t i+1 − t i ) is the time interval between two determinations, and n is the total determination times. The calculation of the N 2 O emission factor (EF, %) was according to Eq (3): WhereCE N 2 O ðtreatmentÞ is the cumulative emissions of N 2 O in the urea or urea+nitrapyrin treatment (kg N ha -1 ), CE N 2 O ðcontrolÞ is the cumulative emissions of N 2 O in the control treatment (kg N ha -1 ), and N applied is the total N amount applied to urea or urea+nitrapyrin treatment (225 kg N ha -1 in this experiment). Soil water-filled pore space (WFPS) was calculated based on Eq (4): Where SWC is the soil water content (%), D B is the soil bulk density (g cm -3 ) and D P is the soil particle density (g cm -3 ). The particle density is assumed to be 2.65 g cm -3 , and the soil bulk density in this experiment is 1.42 g cm -3 . The nitrogen use efficiency (NUE) was calculated according to Eq (5): Where N uptake (treament) is the N uptake in cotton plants based on total aboveground biomass in the urea or urea+nitrapyrin treatment, N uptake (control) is the N uptake in cotton plants based on total aboveground biomass in the control treatment, N input is fertilizer N application rate (225 kg N ha -1 in this experiment).

Statistical analysis
The

Results and discussion N 2 O emission in a drip-fertigated cycle
Under the drip-fertigated condition, some distinctive features occurring on N 2 O emissions were attributed to frequently re-wetting and repeatedly supplying of N fertilizer, thus leading to cycling of oxi-reduction and typical featuring of N 2 O dynamics (Fig 2). It was found that soil N 2 O fluxes increased to a maximum between 2 to 4 d after fertigation and then gradually declined within each fertigation cycle across all the three months (fertilizer season) tested, which is in line with the previous reports [20][21]. The greatest emissions of N 2 O occurred in July due to the majority of water and fertilizer N addition. More N application and lower N 2 O emission were observed in August than in June, suggesting that the nearly mature cotton plants absorbed N more quickly after fertigation in August, with the result such that less N was available for conversion to N 2 O [22][23].

Cumulative N 2 O emission during the main growing season of cotton plants
As shown in Table 2, cumulative N 2 O emissions in the urea treatment were significantly higher than in the control treatment (P<0.001). Compared with the urea treatment, application of urea with nitrapyrin (urea+nitrapyrin treatment) reduced cumulative N 2 O emissions by 7.1% (P<0.05), 21.9% (P<0.01) and 11.6% (P<0.01) in the three months tested accordingly. Urea application increased cumulative N 2 O emission by 2.4±0.1 kg N ha -1 during the main growing season, while nitrapyrin addition mitigated N 2 O emission by 14.3% (P<0.01) at the rate of 225 kg N ha -1 .
At the similar rate of N application, the rate of N 2 O emissions mitigated by nitrapyrin is distinctly affected by different crop ecosystems. For instance, Xiong et al. [24] reported that the N 2 O emissions decreased by 8.8, 21.0 and 24.3% with nitrapyrin mixed with fertilizer N at  the corresponding rate of 200, 300 and 400 kg N ha -1 in vegetable system, and a reduction rate of 50% was found by Sun et al. [25] in rice field treated with 180 and 240 kg N ha -1 . This difference also relates to the supply of water and fertilization. Although the reduction rate of N 2 O emission by nitrapyrin in our study was lower compared with some other reports [13][14][15][16]26], the differences in N 2 O emission between urea+nitrapyrin and urea treatment were still statistically significant ( Table 2). In addition, the relatively low rate of reducing N 2 O in this study might be associated with the lower N fertilizer rate (225 kg N ha -1 ) as the typical rate of N commonly used by local cotton farmers in northwest of China was above 300 kg N ha -1 . It is speculative that the effect of nitrapyrin in reducing N 2 O emission could be considerably higher in the condition of higher N input. In this study, the fertilizer was applied completely during the main growing season of cotton plants (June, July and August), as a consequence, higher N 2 O emissions occurred in these months than in other growing phase. Therefore, controlling N 2 O release during the fertilizer season will contribute to the reduction of annual N 2 O emissions, especially when other conditions (e.g. initial soil physical and chemical properties, water input, environment and agronomic management) are identical; annual emission characteristic and reducing rate of N 2 O can be deduced indirectly through analyzing the seasonal cumulative N 2 O emission. In this experiment, soil NH 4 + -N and NO 3 --N concentrations were significantly greater in both urea and urea+nitripyrin treatments than in the control (Fig 3). It was found that soil NH 4 + -N concentrations were consistently greater in the urea+nitrapyrin treatment than in the urea treatment while the reverse was true for the NO 3 --N concentrations (Fig 3). For example, the average soil NH 4 + -N concentrations in seven days after fertigation were 8.3% (P<0.05), 38.2% (P<0.01) and 34.8% (P<0.01) greater in the urea+nitrapyrin treatment than in the urea treatment in June, July and August, respectively, while soil NO 3 --N concentrations were 7.3% (P<0.05), 17.7% (P<0.01) and 15.8% (P<0.01) less in the urea+nitrapyrin treatment than in the urea treatment in June, July and August, respectively. During the main growing season, the average soil NH 4 + -N concentration was 28.0% greater and soil NO 3 --N concentration was 13.8% less in the urea+nitrapyrin treatment than in the urea treatment. These results indicate that nitrapyrin addition significantly inhibited soil nitrification, and shifted the soil inorganic nitrogen form from nitrate to ammonium. Earlier reports showed that soil NO 3 --N concentration was 23% less and N 2 O emissions were 26 to 49% less when nitrification inhibitors (i.e., DCD and DMPP) were added to ammonium sulfate-fertilized soil [8], while soil NH 4 + -N concentration was increased by 41% and NO 3 --N concentration was decreased by 39% in NH 4 + low soils after nitrapyrin was added [26]. Our findings were less than the above reports probably due to the differences in the ecological systems, N applied rate or NI types used, etc. Overall, nitrapyrin addition significantly inhibited nitrification, and kept the soil NO 3 concentration lower, consequently decreasing the concentration of substrate used for denitrification under high-frequency-irrigation condition in this study.

Driving factors affecting N 2 O emission
Many factors (e.g. soil temperature, WFPS, soil NH 4 + -N and soil NO 3 --N) influence soil N 2 O emission [27][28]. In this study, the factors driving N 2 O emissions are shown in Table 3    -N concentrations were also significantly correlated with N 2 O fluxes as previously reported in a rape field [29].
We can further determine the direct and indirect effects of the above factors on N 2 O emissions through path analysis (Table 3). In this study, soil NH 4 + -N concentration, soil and WFPS under this film-mulched drip-fertigated condition. In addition, it was more difficult to obtain a significant correlation between N 2 O and soil temperature probably due to low soil temperature (<25˚C) [30] and relatively narrow range of soil temperature (20-24˚C) in this test. N 2 O can be produced via both nitrification and denitrification when soil WFPS is at an intermediate levels (i.e. 30% -70%) [4][5]31]. It was discovered by Abalos et al. [27] that frequent irrigation (once per week) could maintain soil WFPS in the upper soil layer at a higher level than did conventional flood irrigation, which significantly stimulated denitrification. As a result, N 2 O emission was greater with frequent drip irrigation than with conventional irrigation. In our study, frequent rewetting with drip irrigation maintained soil WFPS between 31.2% and 67.0% (Fig 2), suggesting that both nitrification and denitrification could contribute to N 2 O emission. Combining with the strong relationship between N 2 O and soil NO 3 --N in Pearson correlation and path analysis, we could preliminarily speculate that more N 2 O was produced via denitrification in this plastic-film-mulched drip-fertigated system. However, more precise future work is needed to determine whether denitrification dominates nitrification during frequent rewetting of soil due to drip-fertigation in plastic-film-mulched calcareous soil.
The effect of nitrapyrin on the EF of soil N 2 O, N uptake, yield and NUE A wide range of N 2 O emission factors (0.01-3.7%) is shown in different countries and regions [32][33][34][35] due to the differences in soil, environment, temperature, crops and management practices. In a rice-winter wheat rotation system in southeast China, the fertilizer-induced emission factor for N 2 O averaged 1.02% during the rice season, 1.65% during the wheat season, and 1.25% during the whole annual cycle [36]. Scheer et al. [37] reported an average N 2 O emission factor of 1.48% in an irrigated cotton field in Uzbekistan (the arid deserts of Aral Sea Basin). Liu et al. [38][39] also showed that direct N 2 O emission factors in a cotton field varied between 0.9% and 2.2%. However, information is still lacking on soil N 2 O emission under plastic-filmmulched drip-fertigated condition. Similar to the previous results in China [36,[38][39][40], the EF of N 2 O in this study was 1.05% in the urea treatment during the main growing season of cotton plants, compared to 0.71% when nitrapyrin was added to the urea solution (Table 4); nitrapyrin addition decreased the EF of N 2 O by 32.4%. In view of the fertilizer N being applied completely in three months (the main growing seasons), it can be speculated that the annual EF of N 2 O would be higher than the seasonal EF based on the continued effectiveness of fertilizer N, and the difference between the seasonal and annual EF would not be too much on account of lower N input (225 kg N ha -1 ) in this experiment.
Compared with urea alone, the yield and nitrogen uptake in the urea+nitrapyrin treatment were increased by 4.1 and 3.6% (P>0.05), respectively. The yield in this study was lower than reported earlier [13,[41][42]. Our findings were more similar to the report by Crawford and Chalk [43] who found that there was no significant difference in yield between fertilizer N plus nitrapyrin and N applied only. Under the tested condition of drip-irrigated plot with input of 225 kg N ha -1 , the relatively high NUE (>50%) was gained, and more importantly, nitrapyrin addition increased NUE by 10.7% (P<0.05). Zhang et al. [44] observed that nitrapyrin raised NUE by 12.6% (P<0.05) in an intensively managed vegetable cropping system, and a similar finding was reported by Sun et al. [25] in a rice field. Clearly, these discrepancies were probably attributed to the differences indifferent crops tested as cotton crop tested in this study is less responsive to the enhanced ammonium nutrition caused by nitrificantion inhibitors, than wheat, corn or rice used in previous studies [13,[41][42].

Conclusions
In plastic-film-mulched and high frequency drip-fertigated system, repeated use of the nitrapyrin significantly inhibited nitrification, shifted the ratio of soil NH 4 + to NO 3 -, decreased N 2 O emission by 14.3% and EF of N 2 O by 32.4%, increased NUE by 10.7% in the calcareous cotton field amended with urea at the rate of 225 kg N ha -1 . NO 3 was the key factor impacting on N 2 O emissions in this system; nitrapyrin addition decreased the concentration of denitrifying substrate (NO 3 -) through inhibiting nitrification process and also reduced N 2 O emission from denitrification process. Taken together, the repeated application of nitrapyrin through drip fertigation is an efficient approach to reducing N losses and promoting nitrogen use efficiency in plastic-film-mulched drip-fertigated calcareous soil in arid areas.
Supporting information S1 Data. Data for the figures and tables. (XLSX)

Author Contributions
Data curation: TL.