Effects of different long-term crop straw management practices on ammonia volatilization from subtropical calcareous agricultural soil

ABSTRACT Ammonia (NH3) volatilized from agricultural production and its secondary aerosols contribute greatly to air pollution. Different long-term crop straw management practices may significantly affect the soil fertility and soil nitrogen cycle, however, the effect on NH3 volatilization has not been well studied. Therefore, a one-year field experiment was conducted to evaluate the effect of straw incorporation on NH3 volatilization from subtropical calcareous agricultural soil from a longterm perspective, including four treatments: synthetic fertilizer (CK); synthetic fertilizer incorporation with 100% or 50% of the previous season’s crop straw (SI1 and SI2, respectively); and synthetic fertilizer incorporation with 50% burned crop straw (SI2B). Soil NH3 volatilizations were monitored through a wheat–maize rotation year by using a dynamic chamber method. The results demonstrated that NH3 volatilization primarily occurred within 38 days and 7–10 days following nitrogen fertilization events for the wheat and maize seasons, respectively. Different crop straw management practices mainly impacted the NH3 flux of the basal fertilization rather than the topdressing fertilization; long-term crop straw incorporation effectively lowered NH3 loss (35.1% for SI1 and 16.1% for SI2 compared to CK; and the inhibiting effect increased with increasing straw amount, possibly contributed by the high straw carbon/nitrogen ratio, and enhanced microbial activity, which contributed to inorganic nitrogen immobilization and lower ammonium content in the topsoil. However, SI2B significantly increased (29.9%) the annual NH3 flux compared with SI2, indicating that long-term 100% straw incorporation could be a promising straw management practice for mitigating NH3 loss and increasing soil fertility. Graphical abstract


Introduction
Agricultural production, as the main anthropogenic source of ammonia (NH 3 ) emissions (Bouwman et al. 1997), contributes 80%-90% of total emissions (Kang et al. 2016). The volatilized NH 3 and its secondary inorganic aerosols may account for about 37% of the total PM 2.5 (fine particles with a diameter of 2.5 μm or less) mass (Tao et al. 2014), and result in serious air pollution (Gu et al. 2014). Furthermore, the emitted NH 3 can indirectly result in water eutrophication (Vitousek et al. 1997), soil acidification (Guo et al. 2010), and a reduction in plant diversity (Bobbink et al. 2010) through nitrogen (N) deposition . Recent studies have indicated that areas of high NH 3 emissions are concentrated in eastern and southwestern China (Kang et al. 2016), which are also areas of serious air pollution in China (Yang et al. 2011). Particularly, the unique topographic surroundings of the Sichuan Basin are conducive to PM 2.5 pollutant accumulation (Tao et al. 2014), so it is essential to reduce NH 3 fluxes in this areaespecially from calcareous arable soil, seen as NH 3 volatilization 'hotspots'.
Straw incorporation, as an environmentally friendly method, is widely used to maintain or increase soil organic carbon (SOC) storage in arable soils (Xia et al. 2018;Zhao et al. 2018). Because the carbon (C) and N biogeochemical cycles are tightly coupled (Cleveland and Liptzin 2007), straw incorporation may also influence N cycle processes in arable soil. Microbial activities in arable soils are generally C limited (Attard et al. 2016), and straw incorporation is favorable for the proliferation of microorganisms, increasing the microbial abundance and activity and consequently improving the immobilization of inorganic N by microbes (Marie-Anne et al. 2010) and thus decreasing the NH 3 volatilization substrate. Furthermore, the physicochemical properties of soil (e.g., pH, soil bulk density, soil aeration, soil water holding capacity, cation exchange capacity (CEC), etc.) are also affected by straw incorporation; in particular, in fields with long-term returning of straw, these factors will also affect NH 3 volatilization (Duan and Xiao 2000). Studies examining the effect of straw incorporation on NH 3 volatilization have reported negative effects (Wang et al. 2012;Yan et al. 2016;Cao et al. 2018;Li et al. 2019). However, some studies have reported that straw incorporation can significantly stimulate NH 3 volatilization (Wang et al. 2012;Sun et al. 2018;Xia et al. 2018). These opposite effects indicate a high level of variation in N migration and transformation after straw incorporation, which might depend on the quantity and quality (C/N ratio) of crop straw, soil properties (pH, CEC, soil bulk density), and climatic conditions, amongst other factors (Duan and Xiao 2000;Ju and Zhang 2017). However, few studies on the effects of the amount and method of crop straw incorporation have been conducted, and the NH 3 volatilization mechanism is still not adequately understood (Zhou et al. 2016). Moreover, from the long-term perspective, the effect of different straw management practices on NH 3 volatilization is influenced by both the incorporated crop straw itself and changes in soil physicochemical properties, which respond slowly to different straw management practices. Therefore, long-term straw incorporation experiments are more suitable for comprehensively elucidating the effects of crop straw incorporation on NH 3 volatilization, but such experiments have rarely been carried out. Straw burning used to be common practice in China (Qu et al. 2012;Tao et al. 2014), therefore, it is meaningful to study the long-term effects of straw burning on soil NH 3 volatilization, to further estimate historical NH 3 emissions.
An experiment was established in a typical wheatmaize rotation system in the subtropical area of China to study the effects of long-term straw incorporation on NH 3 volatilization from alkaline soil. Based on that experiment, the objectives of the present study were to (1) quantify the NH 3 volatilization dynamics and fluxes from long-term straw-incorporation soil; and (2) evaluate the effects of different straw incorporation quantities and methods on NH 3 volatilization.

Site description and experiment design
The field experiment was located in Yanting Agro-Ecological Station of the Chinese Academy of Sciences (31°16ʹN, 105°28ʹE) in Sichuan Province. The climate is classified as 'moderate subtropical monsoon', with annual mean precipitation of 826 mm and average annual air temperature of 17.3°C. The soil properties (0-20 cm) are listed in Table 1.
The long-term wheat-maize rotation experiment, initiated in 2004, was based on a random design including four treatments with three replicate plots (25 m × 2 m; slope: 6.5°) as follows: synthetic fertilizer (crop straw removed, CK); synthetic fertilizer incorporation with 100% or 50% of the previous season's crop straw (SI1 and SI2, respectively); and synthetic fertilizer incorporation with 50% burned crop straw (burned in-situ, SI2B). Before synthetic fertilizer application, the corresponding crop straw (chopped into 3-5-cm pieces) was broadcast evenly and the straw in the SI2B plots was burned to ashes and cooled. Then, ammonium carbonate (130 kg N ha −1 for the wheat season and 90 kg N ha −1 for the maize season), superphosphate (90 kg P 2 O 5 ha −1 (P 2 O 5 : phosphorus pentoxide)), and potassium chloride (36 kg K 2 O ha −1 (K 2 O: potassium oxide)) were broadcast as basal fertilization to all the plots, followed by mixing with surface soil (15 cm) using a rotary cultivator. Urea was surface-broadcast (60 kg N ha −1 ) in the maize season as topdressing after heavy rain during the flare opening Table 1. Soil pH, total nitrogen (TN), soil organic matter (SOM) content, alkali-hydrolysable nitrogen (AHN), available phosphorus (AP), available potassium (AK), and soil bulk density (BD) of topsoil (0-15 cm) (mean ± SE) for different treatments. stage. The mean crop straw C/N ratio used for the wheat season was 82.6 (maize straw), and for the maize season it was 75.0 (wheat straw). The pH of the burned ash was 12.68 for wheat straw and 12.71 for maize straw.

Measurement of NH 3 volatilization
The NH 3 volatilization was measured using a dynamic chamber method. The sampling system was composed of a vacuum pump, a cylindrical dynamic chamber (polymethyl methacrylate, with a height of 15 cm and inner diameter of 20 cm), and an acid trap (60 ml 0.05 mol L −1 sulfuric acid solution) to capture volatilized NH 3 (Cao et al. 2013). The chamber was inserted into the soil (1 cm) and ambient air located at 2.5 m above the soil was pumped to mix the inner air in the chamber, and then the NH 3 in the downstream air was trapped by acid. The background NH 3 in the ambient air (2.5 m) was also determined, using a sealed chamber without insertion into the soil. The NH 4 + -N concentration (NH 4 + : ammonium) in acid solution was measured using a continuous-flow analyzer (AA3 Bran + Luebbe, Norderstedt, Germany). The NH 3 volatilization was monitored twice per day (08:00-10:00 and 15:00-17:00) with a 15 L min −1 airflow rate (Hayashi, Nishimura, and Yagi 2006). Firstly, the NH 3 volatilization sampling frequency was continually measured for half a month following fertilization, and then the frequency was changed to twice per week throughout each season. The daily NH 3 volatilization flux was modified based on the equation published by Hayashi, Nishimura, and Yagi (2006): where F is the daily flux of NH 3 volatilization (kg N ha −1 d −1 ), c s is the NH 4 + -N concentration trapped in acid solution for soil plots (mg L −1 ), c b is the NH 4 + -N concentration trapped in acid solution for background ambient air (2.5 m) (mg L −1 ), v is the acid solution volume (L), r is the chamber radius (m), and t is the duration of NH 3 sampling (h). The seasonal cumulative NH 3 volatilization flux was determined by linear interpolation of the daily fluxes between the two closet days when observations were taken.

Environmental factors
Daily precipitation and air temperature were observed by an automatic meteorological station near the experimental plot. The soil temperature (5 cm depth) and soil moisture (at 0-6 cm) were simultaneously measured during NH 3 volatilization sampling throughout the experimental period by using a TidbiT v2 Temperature Data Logger (Part UTBI-001, Onset Corp., Pocasset, MA, USA) and a portable frequency domain reflectometry (FDR) moisture sensor (MPKit-B Hangzhou Tuopu Instrument Co. Ltd, China), respectively. The soil moisture measured by FDR was converted to water-filled pore space (WFPS) by the equation WFPS = soil volumetric water content/(1 − soil bulk density/2.65) × 100%, in which the assumed soil particle density was 2.65 g cm −3 . The soil heterotrophic respiration rate and soil inorganic N (NH 4 + and NO 3 − ) dynamic content (NO 3 − : nitrate) were simultaneously monitored during the NH 3 monitoring period. For more details on the materials and methods, please refer to the Supplementary Material.

Data processing and statistical analysis
Significant differences in means among the treatments were compared by the least significant difference at the P < 0.05 level with one-way ANOVA using the SPSS 19.0 software program (SPSS Inc., 2008). The relationships between NH 3 volatilization flux and environmental factors (soil temperature and WFPS) were analyzed using linear regression. Figures were prepared using the Origin 9.4 software (Origin Lab Corporation, Northampton, USA).

Related variables
Crop straw incorporation is considered a sustainable and environmentally friendly agricultural practice for improving soil biological and physical fertility and enhancing SOC storage (Blair et al. 2006). Also, more apparent effects are usually observed in long-term straw incorporation fields. The changes in soil fertility will significantly influence the migration and transformation processes of N following the application of N fertilizer (Wang et al. 2015), such as NH 3 volatilization etc., and this was further proved by our long-term crop straw incorporation experiment. For example, on average, the SI1 and SI2 treatments significantly increased soil total N (52.1%), soil organic matter content (55.1%), alkali-hydrolysable N (30.6%), and available potassium (70.6%) in the topsoil (0-15 cm) compared with CK. The above indexes in the burned crop straw treatment (SI2B) also showed a tendency to increase; however, no significant changes were observed in soil available phosphorus and soil pH among all the treatments (Table 1). The soil bulk density was reduced by 2.8%-7.1% under different straw incorporation treatments (Table 1). Similar results have also been observed by others in the same study region (Zhou et al. 2014;Wang et al. 2015). Moreover, soil heterotrophic respiration, as an important indicator for microbial activity (Liang et al. 2018), in our study, on average, enhanced by 41.6% and 20.8% for the SI1 and SI2 treatments, respectively, while it decreased by 14.5% for the SI2B treatment ( Figure S1). Figure 1 shows that all the treatments peaked on the first day following ammonium carbonate application in the basal period of both the wheat and maize seasons. The peak value of NH 3 flux in the SI1 and SI2 treatments decreased by 18.1% and 11.0% for the wheat season, and significantly decreased by 65.3% and 47.4% for the maize season (P < 0.05), respectively. The NH 3 peak value had a decreasing tendency with an increase in the straw incorporation rate, which is inconsistent with the results of Li et al. (2019). Meanwhile, the highest flux was observed on the third day after broadcast topdressing with urea in the maize season ( Figure 1). This lag effect has also been observed in some other studies after urea application, such as in the subtropical sugarcane production system (Dattamudi et al. 2016), rice-wheat rotation system , and typical vegetable system (Li 2013), mainly due to the time taken for the hydrolysis of urea into ammonium (Proctor, Koenig, and Johnston 2010). Proctor, Koenig, and Johnston (2010) reported that the increased urease activity facilitated by straw incorporation could also be a contributing factor to an earlier and higher peak value of NH 3 fluxes in a dryland Kentucky bluegrass seed production system; however, our study did not observe this phenomenon in the urea topdressing period. On the contrary, the SI1 and SI2 straw incorporation treatments lowered the peak value of NH 3 fluxes by 13.4% and 9.4%, respectively.

Temporal variations of NH 3 volatilization
The NH 3 volatilization lasted for 38 days in the wheat season following N application, while it only lasted for 8-10 days in the maize season, which might be attributable to the lower soil temperature inhibiting the processes of NH 3 diffusion and NH 4 + transformation , and which was further demonstrated by the positive correlation between daily NH 3 flux and soil temperature (P < 0.0001) (Figure 2) in the wheat season. Moreover, the soil temperature in the maize season was no longer a significant constraining factor (P > 0.05), which favored the process of nitrification ( Figure S2) (Clay, Malzer, and Anderson 1990) and decreased the substrate (NH 4 + -N) for NH 3 volatilization. The NH 3 volatilization characteristics indicated that long-term monitoring for the winter season and high-frequency monitoring for the summer season are needed to increase the flux accuracy.

Seasonal and annual cumulative NH 3 fluxes
As shown in Table 2, the annual cumulative NH 3 fluxes ranked as: SI2B > CK > SI2 > SI1. The NH 3 cumulative flux decreased with an increased straw incorporation rate. Compared with CK, straw incorporation (SI1 and SI2) significantly lowered the NH 3 flux by 48.6% and 27.5% (P < 0.05) in the wheat season, respectively. SI1 also evidently decreased the NH 3 flux by 48.5% in the basal fertilization of the maize season, and SI2 lowered it by 17.0% in this period. These findings are comparable with a previous study from Northeast China with different straw incorporation rates in the maize season (100% straw incorporation decreased the NH 3 flux by 24.6%) (Li et al. 2019). However, there was no significant difference among different treatments in the topdressing period (surface-broadcast urea) in the maize season, probably because the poor contact between urea and soil diminished the influence of the soil itself, or the effects of in-season crop straw management practices on NH 4 + transformation or immobilization. The mean NH 3 flux of the maize basal period was lower than that of the wheat season, probably because of the lower N application rate and higher nitrification rate ( Figure  S2) in the maize season (Duan and Xiao 2000), which ultimately decreased the substrate for NH 3 volatilization. The topdressing N rate accounted for 40% of the total N input in the maize season; however, the average NH 3 flux of topdressing accounted for 71.4% of the total NH 3 flux in maize season (Table 2), indicating that surface N application largely stimulates NH 3 volatilization (Nett et al. 2015). The highest cumulative NH 3 flux was observed in the SI2B treatment in the basal fertilization of both seasons, indicating that straw burning tends to stimulate NH 3 loss from ammonium carbonate application as the basal fertilizer. With respect to straw management practices, 50% straw burning and 50% straw incorporation in the basal fertilization increased NH 3 losses by 16.5%, on average, and decreased them by 22.3%, compared with CK, respectively (Table 2), indicating obvious disadvantages of straw burning in minimizing NH 3 loss with respect to straw management. Similarly, Phongpan and Mosier (2003) also found that straw burning could stimulate NH 3 loss in two field experiments.

The effects of long-term crop straw incorporation on NH 3 volatilization
Our result showed that, following basal fertilization, 100% and 50% crop straw incorporation led to a 48.6% and 27.5% reduction in NH 3 loss for the wheat season, and   reduced to a 48.5% and 17.0% NH 3 loss for the maize season, respectively. The inhibiting effect of straw incorporation on NH 3 volatilization might benefit from the higher soil microbial activity (soil heterotrophic respiration increased by 41.6% for SI1, on average, and by 20.8% for SI2 ( Figure S1)), which enhanced NH 4 + microbial immobilization from NH 4 + to labile organic-N or microbial N following the high C/N ratio of crop straw (ranging from 75.9 to 82.6) application and higher nitrification rate ( Figure S2) (Huang et al. 2004). As a result, compared with CK, the average NH 4 + content for the SI1 and SI2 treatments during the NH 3 monitoring period, respectively, decreased by 21.3% and 19.7% for the wheat season, and by 28.0% and 32.6% for the maize season ( Figure S2). Recently, Wang et al. (2015) also found that long-term straw incorporation increased the NH 4 + immobilization rate by 15.6 times (from 0.9 to 14.4 mg N kg −1 soil d −1 ) under a wheat-maize rotation system in the same region by using a complex 15 N tracing method. Furthermore, a 15 N labeling experiment in a paddy system showed that crop straw incorporation decreased NH 3 losses by 24%, promoted the immobilized soil inorganic N, and then improved the N mineralization in the later crop growing stage (Cao et al. 2018). Moreover, the present study showed long-term crop straw incorporation could improve soil fertility (Table 1), which might facilitate the absorption of N by crops, and then indirectly reduce the NH 3 loss. A global agroecosystem meta-analysis also indicated that crop straw incorporation can significantly increase the N uptake and N use efficiency of crops, by 10.9% and 15.0% respectively (Xia et al. 2018). On the contrary, previous studies have reported that straw incorporation might stimulate NH 3 loss from urea fertilizerfor example, in a rice-oilseed rape cropping system (Su et al. 2014) and rice-wheat rotation system (Wang et al. 2012;Sun et al. 2018)-due to the increased urease activity and catalyzed hydrolysis of urea into NH 4 + . Overall, the comprehensive effect of straw incorporation on NH 3 loss should be considered based on the cropping system, length of straw incorporation, fertilizer type, soil properties, and climatic conditions.

The effects of long-term crop straw burning on NH 3 volatilization
Crop straw residue, as a kind of obstacle to farming, was often burned to facilitate crop rotation, which had been proven to be a significant seasonal source of PM 2.5 (Qu et al. 2012). However, the effects of straw burning on NH 3 volatilization were not well studied. Our long-term straw burning experiment showed that straw burning stimulated NH 3 loss from the basal fertilization with ammonium carbonate, probably because straw burning decreased the soil microbial activity (Dattamudi et al. 2016) to assimilate NH 4 + (e.g., the processes of nitrification and immobilization). For example, the soil respiration result showed SI2B decreased soil heterotrophic respiration by 14.5% on average, compared to CK, throughout the NH 3 monitoring period ( Figure S1). Meanwhile, it increased the top-soil NH 4 + content by 28.0% and 54.8%, on average, in the wheat and maize season, respectively, which might also be attributable to this phenomenon. Moreover, the straw ash itself, with high pH (the pH of both wheat and maize straw averaged 12.7), might also contribute to the formation of free NH 3 in soil. Phongpan and Mosier (2003) also found that straw burning increased NH 3 loss by 5% in a rice cropping system with urea application. Fortunately, the Chinese government has implemented some economic and policy incentives to encourage straw return and prohibit straw burning (Zhao et al. 2018), which might have an extra benefit towards reducing NH 3 loss.

Conclusion
The present study showed that different long-term crop straw management practices significantly affected NH 3 volatilization. Crop straw incorporation effectively reduced annual NH 3 volatilization losses by 16.1% to 35.1%, and the inhibiting effect increased with the straw application rate, which might be due to enhanced microbial N immobilization and a higher nitrification rate. Meanwhile, straw burning resulted in a 29.9% annual NH 3 volatilization increase compared to the same amount of unburned straw incorporation, possibly attributable to reduced microbial activity and the high pH of straw ash. Therefore, proper crop straw management practices could be an environmentally friendly strategy for subtropical calcareous agriculture soils towards reducing NH 3 volatilization and increasing soil fertility.

Disclosure Statement
No potential conflict of interest was reported by the authors.