Airborne nitrogen load in Japanese and Chinese agroecosystems


K. HAYASHI, Carbon and Nutrient Cycles Division, National Institute for Agro-Environmental Sciences, 3-1-3, Kan-nondai, Tsukuba, Ibaraki 305-8604, Japan. Email:


The objective of this review is to make current knowledge on the nitrogen (N) load throughout the atmosphere (airborne N load) available to readers, with special emphasis on Japanese and Chinese agroecosystems. Key species of airborne N are ammonia, nitrogen dioxide, nitrogen oxide, nitric acid, nitrous acid and particulate ammonium and nitrate. Organic N also exists in the atmosphere. The main processes in terms of the airborne N load involve emission, atmospheric transportation and transformation, atmospheric deposition and environmental impacts. Agricultural activities are the largest emitters of ammonia through emissions mainly from livestock waste and field-applied N fertilizers. The ammonia emissions in China in 1995 from chemical fertilizers and organic fertilizers, such as animal excreta, were estimated to be 3.56 and 2.04 Tg N year−1, respectively, and the emissions in Japan were 0.059 and 0.069 Tg N year−1, respectively. The most fundamental causes of the airborne N load in relation to Japanese and Chinese agriculture were intensive livestock farming in Japan and over use of N fertilizers in China. However, agroecosystems are also a sink for airborne N. Atmospheric N deposition was up to 20 and 60 kg N ha−1 year−1 in Japan and China, respectively. The unrelenting load of airborne N continues in Japan and China. The development of a simple, but accurate method to determine the dry deposition flux that is applicable to simultaneous and multipoint observations would be valuable. The establishment of cross relationships among in situ observations, remote sensing and numerical modeling is also needed to cope with the issue by assessing the actual status, predicting the future status and working out effective measures.


Nitrogen (N) is indispensable for life. Although molecular nitrogen (N2) accounts for 79% of the earth’s atmosphere, this inert N2 cannot be utilized by life forms, excluding certain microorganisms with an extraordinary ability known as nitrogen fixation. In contrast, a very small portion of the atmosphere is composed of N compounds other than N2, collectively called reactive N (Nr) (Galloway and Cowling 2002). For example, the second highest concentration next to N2 is 319 ppb (v/v) (Intergovernmental Panel on Climate Change 2007) of nitrous oxide (N2O). The average air concentrations of other gaseous Nr are usually only several tens of ppb or less. However, Nr is involved in various biotic and abiotic processes in the earth’s system and has high reactivity and rapid circulation in the environment. Therefore, attention must be given to Nr even when its concentration in air is low.

Agriculture is a human activity that artificially enhances biological productivity to obtain foods and usable materials, and artificial N input is essential for agriculture because N is one of the most important macronutrients. Chemical and organic fertilizers are utilized to supply N to crop production. Artificially enhanced N inputs do not necessarily have serious environmental consequences if the amount of surplus N leaking from the farmlands is within the capacity of the surrounding environment. However, in Japan, nitrate concentrations in groundwater have increased steadily, particularly in some intensive agricultural areas (e.g. Kumazawa 2002). In contrast, in China, N pollution in rivers and lakes, particularly in the eastern agricultural regions, where heavy rotations of rice and other crops are observed, is more serious than that in Japan, and the situation is becoming worse (e.g. Xia and Yang 2003). One of the root causes of the pollution is the overflow of N from agroecosystems as a result of excessive inputs of chemical and organic fertilizers to farmlands.

Direct leaching and run-off of N from a field to a water body is a clear pathway of agricultural excessive N load to the environment. However, a more important pathway may be Nr emission–transport–deposition through the atmosphere (airborne N). Airborne N spreads broadly and rapidly. It flows off beyond a watershed and can even pollute an upstream area. In fact, some Nr can easily cover the distance between Japan and China (e.g. Takiguchi et al. 2008).

Agricultural activities act as a source of airborne N, meanwhile input of Nr from the atmosphere to farmlands also occurs as wet and dry deposition, collectively called atmospheric deposition. Thus, agroecosystems are both a source and a sink of airborne N. In addition, urban activities also act as a source of airborne N, for example, nitrogen oxides (NOX) from combustion and ammonia (NH3) from wastes. Motor vehicle exhaust contains not only NOX and particulate N, but also NH3 from the excessive effect of exhaust catalyst that reduces NOX into NH3 beyond N2 (e.g. Kean et al. 2009). Some of the airborne N that originates from urban activities contributes to the N load to agroecosystems and vise versa.

There are many excellent reviews on airborne N and on the relationships between air quality and agriculture (e.g. Aneja et al. 2001; Erisman et al. 2008; Mosier 2001; Sutton et al. 2008; Xiong et al. 2008). However, very little information on airborne N is available for Japanese and Chinese agroecosystems. The current status for both the emission and deposition of Nr between the agroecosystems and the atmosphere has not yet been elucidated in these countries; moreover, its future status is unclear.

The purpose of the present study is to make current knowledge on the airborne N load in Japanese and Chinese agroecosystems available to readers. Before doing so, however, we need to examine the characteristics of an airborne N load. Therefore, in this review, we begin with background information of relevant processes of airborne N and then look at the atmosphere–agroecosystems exchange of Nr. After that, we overview the current status of airborne N and its impacts in Japan and China and discuss future research needs.

Airborne nitrogen load

Key members

Ammonia (NH3), nitrogen dioxide (NO2), nitrogen oxide (NO), nitric acid (HNO3) and nitrous acid (HNO2) are the main inorganic gases and particulate ammonium (pNH4) and nitrate (pNO3) are the main inorganic particles in relation to airborne N. NH3 and pNH4 are the major reduced forms in the atmosphere and the other forms are oxidized forms. NO2 and NO are collectively called NOX. N2O, an important greenhouse gas, is excluded from airborne N because N2O is unlikely to contribute to atmospheric deposition. Organic N also exists in the atmosphere (Neff et al. 2002) and includes atmospheric bacteria, amines as a reduced form and nitric acid esters and organic nitrates, such as peroxiacetyl nitrate (PAN) as an oxidized form.

Main processes

The relevant processes of airborne N load are summarized in Fig. 1.

Figure 1.

 Schematic view of the airborne nitrogen load in and surrounding agroecosystems. Dep, deposition; Em, emission; NOX, nitrogen oxides, particularly nitrogen oxide and nitrogen dioxide; Nr, reactive nitrogen; OrgN, organic nitrogen; PM, particulate ammonium and nitrate; Trans, transportation.


The main types of gaseous Nr directly emitted from human activities are NH3, NO2 and NO. Photochemical formation in the atmosphere is the primal source of HNO3 and HNO2, in which NO2 and NO act as the precursors. Fine particles with diameters of <2.5 μm (PM2.5) are the main form of pNH4, whereas both fine and coarse particles contain pNO3 (e.g. Bardouki et al. 2003). Particulate ammonium and pNO3 in the fine fraction are secondarily formed in the atmosphere by condensation with gaseous substances. In contrast, coarse particles containing Nr can be directly emitted from human activities, such as combustion and incineration, which may contain organic N.

Transportation and transformation

Reactive N in the atmosphere is transported by advection and diffusion, collectively called atmospheric transportation. Advection is ruled by airflow and diffusion is governed by the gradient in concentration. Various chemical transformations, such as oxidation, gas-to-particle conversion and photochemical reactions, occur in the atmosphere in conjunction with transportation.

Gas-to-particle conversion is one of the most important atmospheric processes for NH3 and pNH4. The main counter anions of pNH4 are sulfate, nitrate (NO3) and chloride ions. Therefore, the formation of pNH4 also equals acid neutralization in the atmosphere. However, acid neutralization has no effect on the amount of the airborne N load. Furthermore, NH4-N acts as an acidifying substance after atmospheric deposition through nitrification and/or plant uptake. Fine particles generated from the condensations of NH3 and acids are themselves a pollutant that affects human health. Fine particles also result in long range, occasionally transboundary, transportation because their residence time in the atmosphere is longer than that of gases.

Atmospheric reactions in relation to oxidized N are more complex and various formation and removal processes occur in the atmosphere on the basis of photochemical reactions (e.g. Stutz et al. 2002). For example, NO is oxidized into NO2, and NO2 oxidizes O2 to form O3 and NO2 itself returns to NO. Hydroxyl radicals, an important photochemical species, oxidize NO and NO2 into HNO2 and HNO3, respectively, whereas the decomposition of HNO2 into NO and a hydroxyl radical is a key removal process of HNO2 and simultaneously a source of a hydroxyl radical.

Atmospheric deposition

According to Seinfeld and Pandis (2006), atmospheric deposition can be divided into two types, wet and dry.

Wet deposition is comprised of in-cloud and below-cloud scavenging. In-cloud scavenging is a process in which cloud droplets and ice crystals take in substances during their formation in a cloud. Below-cloud scavenging is a process in which rain droplets and snow flakes take in substances on their way to the ground surface. Wet deposition observed at the ground surface is a mixture of the two types. Most Nr in wet deposition is in inorganic ion forms, of which NH4+ and NO3 are the dominant species and nitrite ion (NO2) is occasionally found in very small fractions. Wet deposition also contains organic N, for example, 17% of the total N in a rural area in central Japan (Hayashi et al. 2007) and up to 22% in Beijing, China (Zheng et al. 2007). Wet deposition of Nr increases as a result of the effect of below-cloud scavenging when the air concentration of Nr is high (e.g. Hojito et al. 2006a), particularly for highly soluble gases, such as NH3 and HNO3.

Dry deposition is a process in which gases and particles in the atmosphere directly deposit onto the ground surface. Unlike wet deposition, dry deposition is ruled by the conditions near the ground surface alone, such as micrometeorological parameters and the air concentrations of Nr. Surface conditions strongly affect the rate of dry deposition in two ways. One way is the effect of surface roughness on micrometeorological parameters and the other is the affinity (absorbability) of the ground surface to each of the gaseous and particulate species. Thus, the rate of dry deposition differs among Nr species. Deposition velocity (Hicks et al. 1987), a type of mass transfer coefficient, is a usable scale to deal with dry deposition, in which the rate of dry deposition is expressed as the product of the air concentration at a given height and the deposition velocity at the same height. Highly soluble and reactive gases, such as NH3 and HNO3, have a relatively large deposition velocity, whereas fine particles have a generally small deposition velocity. Therefore, Nr that has a high air concentration and/or a large deposition velocity, typically NH3, NO2 and HNO3, tends to account for the major part of the total dry deposition.

Environmental consequences

An increase in N deposition enhances the primary productivity of terrestrial and estuary ecosystems. Solberg et al. (2004) reported a 25% increase in forest growth in southern Norway as a result of increased N deposition.

A number of Nr species, for example, NOX and PM2.5, are air pollutants. PM2.5 directly affects human health and NOX are precursors of tropospheric ozone.

Excess N in soils and water bodies leads to eutrophication if allowed to get worse. A lake or inner bay whose watershed has a high coverage of non-point sources of Nr eutrophicates easily. Furthermore, an airborne N load can even affect an upstream area by atmospheric transportation and deposition.

An incubation experiment using Swedish forest soils revealed that an increase in the N load resulted in enhanced net mineralization and nitrification, and then the limiting factor of nitrification shifted from ammonium availability to other factors, such as soil acidity (Månsson and Falkengren-Grerup 2003).

Butterbach-Bahl et al. (2002) reported that increases in N deposition in Scots pine forest soils in north-eastern Germany led to an increase in the emissions of NO and N2O owing to an increase in the volume of N circulation.

In contrast, the increased N deposition resulted in the inhibition of methane oxidation by forest soils (Butterbach-Bahl et al. 2002; Saari et al. 1997), in which the NH4-N input was, in particular, related to the inhibition of methane oxidation (Morishita et al. 2004).

An increase in N deposition may lead to changes in biodiversity. Long-term and chronic N deposition significantly reduced plant species richness in oligotrophic grasslands in the United Kingdom (Stevens et al. 2004); the effect was serious with a reduction of one species per 4-m2 quadrat for every 2.5 kg N ha−1 year−1 of chronic N deposition.

Atmospheric deposition of HNO3 and NH3 has the potential to result in acidification; the latter results from proton production through nitrification after deposition to the ground surface. The risk of acidification in Asia is worrisome. Hicks et al. (2008) pointed out a large risk of future depletion in the acid-neutralizing capacity of soils in Asia, including parts of southern China, on the basis of a pessimistic, but not improbable scenario.

Acidity provided by acidifying Nr, such as HNO3, causes the corrosion of structures, particularly those made of stone and metal.

Agroecosystems as a source and sink of airborne nitrogen

Modern agriculture is the largest emitter of NH3 in the earth’s system (Galloway et al. 2004). Livestock wastes, including slurry, manure and compost, are a major NH3 emitter in agriculture. Synthetic chemical N fertilizer is another major NH3 emitter. Plant bodies, typically their stomata, become an emitter of NH3 when N nutrition is excessive (e.g. Hayashi et al. 2008a,b). Decomposed organic matter, such as crop residues (Milford et al. 2000) and field burning (Andreae and Merlet 2001), is another source of NH3.

Fossil fuel combustion by agricultural machines is a source of NOX and particulate N. Farmlands can emit NO as a result of denitrification. Some agricultural practices, such as cultivation, can stir up particles from bare croplands, including inorganic and organic N. Furthermore, the incineration of crop residues generates NH3, NOX, HNO2 and particulate N.

In contrast, agroecosystems receive both wet and dry deposition. It is indeed difficult to quantify the annual budget of N dry deposition at farmlands because the surface conditions of farmlands vary significantly with crop growth and agricultural practices that strongly affect the deposition velocity. Hayashi et al. (2009a) reported that an upland field with bare soils was a sink of NH3, except for the period just after the application of manure. Dry deposition of NOX is also important. In particular, NO2 has a higher air concentration than other airborne N species, including NO, and has a larger deposition velocity than NO. Therefore, dry deposition of NO2 to farmlands seems important next to that of NH3.

Current status in japanese and chinese agroecosystems

Emissions of reactive nitrogen

Double cropping of rice and other crops is a general practice in the warm regions of Japan, particularly western Japan. Double or triple cropping of leaf vegetables is also practiced in eastern Japan. However, single cropping is the general practice in most of Japan because of the cold winters and snow. Therefore, the fallow period tends to be long. In the case of rice cropping in eastern Japan, the fallow period is approximately 6 months. The application rate of N fertilizers is usually small, <100 kg N ha−1 per cropping for paddy rice. This small application rate is an environmentally friendly practice and simultaneously optimizes the palatability of rice. In contrast, double cropping with a short fallow period or no fallow period at all is practiced in many parts of China. Furthermore, the application rate of N fertilizers is large, and is typically approximately 200–300 kg N ha−1 per cropping for paddy rice and second crops (e.g. maize, wheat, oil seed rape and cotton). Therefore, there is a large difference in the annual application rates of N fertilizers in rice cropping areas, with <100 kg N ha−1 year−1 in Japan and approximately 400–600 kg N ha−1 year−1 in China. The application rate of N fertilizers is larger for leaf vegetable cropping. For example, the mean application rates of N fertilizers to vegetable fields in Nanjing and Wuxi in China reached 1,862 and 2,749 kg N ha−1 for 2 years, respectively (Wang et al. 2008a). Therefore, Chinese agricultural areas have a high potential for N pollution, including NH3 volatilization loss.

Although Japan strongly depends on imports for food and feed, the self-sufficiency ratio of livestock products in 2006 accounted for 67% on a caloric basis (Ministry of Agriculture, Forestry and Fisheries 2008); however, imported feed accounted for 76% of the domestic production. An important feature of Japanese livestock production is its intensity. Intensive livestock production is accompanied by vast amounts of livestock waste within a narrow area, which may become a strong NH3 emitter. The N load accompanying the treatment and disposal of livestock waste is a serious problem in Japan. In contrast, the N load per area of livestock production in China remains smaller than that in Japan (e.g. Shindo et al. 2006).

The amount of Nr emitted from agricultural activities in Japan and China must be considered. Yan et al. (2003) compiled an inventory of NH3 and NO emissions from croplands in East Asian countries in 1995. The estimated NH3 emissions in China from chemical fertilizers and organic fertilizers, such as animal excreta, were 3.56 and 2.04 Tg N, respectively, whereas those in Japan were 0.059 and 0.069 Tg N, respectively, in which Yan et al. (2003) determined the emission factors for urea and ammonium bicarbonate, the two most consumed N fertilizers in China, through experiments conducted in China. In contrast, the estimated NO emissions in China from chemical fertilizers, crop residues and organic fertilizers were 0.106, 0.005 and 0.058 Tg N, respectively, whereas those in Japan were 0.003, 0.0002 and 0.001 Tg, respectively (Yan et al. 2003). Several studies have examined NH3 emission in Japan. Kannari et al. (2001) showed that the total NH3 emission in Japan in 1994 was 0.43 Tg N, of which 0.26 and 0.023 Tg were emitted from livestock waste and chemical N fertilizers, respectively. Furthermore, Murano and Oishi (2000) showed, regarding NH3 emissions in Japan in 1991, that 0.19 and 0.050 Tg N of NH3 were emitted from livestock waste and chemical N fertilizers, respectively. However, these two studies used NH3 emission factors determined in Europe. In all the three studies, the authors noted that the emission factors were insufficient and did not completely reflect the status in Japan and China. Attention should be given to the development of Nr emission factors specific to Japanese and Chinese agriculture. Yan et al. (2006) also compiled an emission inventory of various substances in relation to biomass burning in China in 2000, in which the total amount of field-burned crop residues was 122 Tg, accounting for 19.4% of the total crop residues. The emissions of NOX, NH3 and PM2.5 from crop residue used as fuel were estimated to be 0.043 and 0.061 Tg N and 1.1 Tg, respectively, and those from field-burned crop residues were estimated to 0.14 and 0.049 Tg N and 0.48 Tg, respectively (Yan et al. 2006). However, Freney et al. (2008) noticed that the estimated annual emission of NH3 in Asia from agricultural practices varied greatly among research groups, ranging from 11.7 to 19.1 Tg N year−1 from 1990 to 2001, whereas the values for NOX emission fell within a narrow range of 0.58–0.66 Tg N year−1. The emission inventory of NH3 has room for improvement.

The actual rate of Nr emission from agroecosystems varies largely with different agricultural practices and their magnitude, including the application rate of N fertilizer. Therefore, a ‘ratio’, for example, NH3 volatilization loss to applied N amount, is a convenient scale to explain the effects of agricultural activities on Nr emissions from agroecosystems. It is, however, noteworthy that the ratio varies depending on the conditions.

Case studies of NH3 emission from farmlands in Japan and China are summarized in Table 1. Case studies at a paddy field in central Japan (Hayashi et al. 2006a, 2008a) showed that NH3 volatilization loss as a result of the application of urea from the surface of the paddy field, excluding rice plants, averaged 1.4%, whereas loss from the whole paddy field was 8.2% (Table 1). In particular, the first supplemental fertilization with a rate of 30 kg N ha−1 resulted in high NH3 volatilization loss (21%), which indicated that rice plants also acted as emitters of NH3 (Hayashi et al. 2008a). Furthermore, Hayashi et al. (2008b) found temporal and rapid increases in the NH4+ concentration in the xylem sap of rice with the application of urea. Thus, NH3 emission from rice plants with excessive N nutrition is highly possible. There are a number of case studies of NH3 emission from Chinese paddy fields. The application of urea by surface incorporation at puddling results in relatively small volatilization losses, 8.8% (Cai et al. 1986) and 10.6% (Cai et al. 1992a) (Table 1). However, supplemental fertilization with urea by surface application boosted the total NH3 volatilization loss to 28.7 and 29.9% (Fan et al. 2006) (Table 1). Furthermore, alkaline soils with a high pH (8.8) showed large volatilization losses of 30 and 39% from applied urea and ammonium bicarbonate, respectively (Zhu et al. 1989) (Table 1). Although acidic conditions generally inhibit NH3 volatilization, a paddy field with a pH of 5.5 showed a large volatilization loss of 40% from applied urea (Cai et al. 1992b) (Table 1). Cai et al. (1992b) attributed this rather extreme result to the high temperature, more than 30°C in many cases, and the shallow floodwater with a high pH (>7), despite the low soil pH.

Table 1. Case studies of ammonia volatilization loss from farmlands in Japan and China
LocationSoil typeSoil pH (H2O)CropFert.Appl. methodRate
(kg N ha−1)
Total rate
(kg N ha−1)
Meas. methodNH3 loss ratio (%) toSource
  1. Plants were excluded from the measurements. Source: 1, Hayashi et al. (2006a); 2, Hayashi et al. (2008a); 3, Hou et al. (2007); 4, Cai et al. (1986); 5, Fan et al. (2006); 6, Li et al. (2008); 7, Cai et al. (1992a); 8, Cai et al. (1992b); 9, Zhu et al. (1989); 10, Hayashi et al. (2009a); 11, Hayashi et al. (2009b);12, Cai et al. (2002); 13, Zhang et al. (1992); 14, Fan et al. (2005); 15, Matsunaka et al. (2008). Fert.: fertilizer; AB, ammonium bicarbonate; ADS, anaerobically digested slurry; AP, ammonium phosphates; AS, ammonium sulfate; CCM, composted cattle manure; CM, cattle manure; CP, pelleted cattle manure; CS, cattle slurry; PM, poultry manure; PP, pelleted poultry manure; U, urea. Appl. method: application method; B, broadcast (surface application); B+Ir, broadcast followed by irrigation; In, surface incorporation; Pl, placement. Meas. method: measurement method; AT, acid trap method; DC, dynamic chamber method; MG, micrometeorological method – gradient method; MM, micrometeorological method – mass balance method; WT, wind tunnel method.

Paddy field
 TokyoFluvisol5.8Forage riceAPIn/B/B100 × 3300300DC12.212.23
ADSIn/B/B100 × 3300103DC51.7150.83
ADSIn/B/B150 × 3450154DC37.7110.03
ADSIn/B × 575 × 6450154DC29.686.33
 JiangsuLacustrine clay5.4RiceUIn909090MM8.88.84
 JiangsuAnthrosol7.2RiceUB × 341/53/41135135MM28.728.75
UB × 381/108/81270270MM29.929.95
 HubeiYellow sediment6.8RiceUB225225225AT20.620.66
 ZhejiangAlluvial- lacustrine deposit6.0RiceUIn606060MM10.610.67
 JiangxiRed clay5.5RiceUIn909090MM40408
 HenanCalcareous sandy loam8.8RiceUB909090MM30.130.19
Upland field
 IbarakiAndosol5.9Pac choiPMIn1201207.3DC0.00.711
 HenanCalcareous sandy loam8.4–8.8MaizeUB+Ir757575MM181812
 HenanCalcareous sandy loam8.8MaizeUPl808080MM121213
 HokkaidoGray upland soil6.4TimothyCSB276276120DC81815

Volcanic ash soils are typical in Japanese upland fields. A relatively low pH and a high cation exchange capacity are key features of volcanic ash soils, which effectively inhibit NH3 volatilization. For example, a case study at an upland field of volcanic ash soils in northern Japan showed that the application of composted cattle manure by surface incorporation to bare soil surface resulted in a very small NH3 volatilization loss (0.8% to total N), and the subsequent application of chemical fertilizers as a mixture of ammonium sulfate (AS) and ammonium phosphates by placement (3 cm depth) resulted in negligible loss (Hayashi et al. 2009a) (Table 1). Moreover, case studies at an upland field of light-colored volcanic ash soils in central Japan also showed negligible NH3 volatilization losses from surface-incorporated manure and AS (Hayashi et al. 2009b) (Table 1). However, Matsunaka et al. (2008) reported a relatively large NH3 loss in grassland on grey upland soils in northern Japan, where surface applications of dairy cattle slurry resulted in the loss of 8–17% to total N (or 18–35% to NH4-N) (Table 1). Comprehensive studies in an upland field of calcareous soils with a high pH of 8.8 used for rotations of maize and wheat in the North China Plain showed that the NH3 volatilization losses from applied urea to maize as summer cropping were 26–48%, 18% and 11–12% for surface application, broadcast followed by irrigation and deep placement, respectively (Cai et al. 2002; Zhang et al. 1992) (Table 1). Despite the strong inhibition effect of deep placement, 11–12% of volatilization loss still occurred. To a certain extent this may be ascribed to the high soil pH of 8.8. Fertilizer application at farmlands with a high soil pH presents the problem of N use efficiency as well as the emission of airborne N. In contrast, growing wheat as a winter crop in the same field resulted in relatively small volatilization losses of 2.0–15%, 0.6% and 2.3% for surface application, broadcast followed by irrigation and deep placement, respectively (Cai et al. 2002) (Table 1). The small losses after wheat cropping may result from the low temperature at fertilization in autumn and the lack of canopy emission without plants at fertilization.

Emission of NO from farmlands results from denitrification. Therefore, soil conditions appear to be the strongest factors controlling NO emission. Unfortunately, only a limited number of studies have been conducted, and these studies form part of N2O related studies (Table 2). Case studies at an upland field of light-colored volcanic ash soils in central Japan showed a range of NO emissions to applied total N (0.02–1.19%) (Akiyama and Tsuruta 2002, 2003; Akiyama et al. 2000) (Table 2), with a mean value of 0.48%. Katayanagi et al. (2008) reported a similar result in maize fields in northern Japan (0.44%) (Table 2). They also reported 0.24–0.28% of emissions from grassland and pasture (Table 2). However, Kusa et al. (2006) recorded an extremely high emission of NO (1.64–3.98%) at an upland field of volcanic ash soils with poor drainage in northern Japan (Table 2). Highly activated denitrification may result in a high NO emission in addition to N2O.

Table 2. Case studies of nitrogen oxide emission from farmlands in Japan
LocationSoil typeSoil pH (H2O)CropFertilizerTotal rate
(kg N ha−1) as T-N
Meas. methodNO emission
ratio (%) to T-N
  1. Source: 1, Kusa et al. (2006); 2, Katayanagi et al. (2008); 3, Akiyama et al. (2000); 4, Akiyama and Tsuruta (2002); 5, Akiyama and Tsuruta (2003). AM, a mixture of ammonium sulfate and urea in a 1:2 ratio with 0.42% nitrification inhibitor (2-amino-4-chloro-6-methyl pyrimidine); AP, ammonium phosphates; CF+M, chemical fertilizer and manure; CCN, controlled release calcium nitrate; CN, calcium nitrate; CU, controlled release urea; PM, poultry manure; SM, swine manure; U, urea; UAS, a mixture of ammonium sulfate and urea in a 1:2 ratio. Meas. method: measurement method; CC, closed chamber method.

 HokkaidoAndosol (poor drainage)5.6MaizeCF+M 1128CC1.641
CF+M 2128CC3.981
 HokkaidoAndosol and Histosol5.2–5.3MaizeCF+M220CC0.442
Andosol (partly Histosol)5.0–6.1GrasslandCF+M143CC0.282
 IbarakiAndosol5.9Pac choiCN 1150CC0.324
CN 2150CC0.164
CCN 1150CC0.224
CCN 2150CC0.024
CU 1150CC0.994
CU 2150CC0.234
 IbarakiAndosol5.9Pac choiPM150CC0.075

Atmospheric deposition of reactive nitrogen

The air concentration provides important information for identifying the degree of air pollution and for evaluating the rate of dry deposition. However, the rate of wet deposition also increases at a high air concentration owing to below-cloud scavenging (e.g. Hojito et al. 2006b). The annual means of the air concentration of Nr in Japan and China are shown in Table 3. Particularly high concentrations of NH3 are found at several sites in Japan and China (Table 3); these sites were affected by neighboring livestock facilities (Hojito et al. 2006a; Hu et al. 2007). Freney et al. (2008) also pointed out that the background concentrations of NH3 in Chinese agroecosystems were high, that is, 20–25 μg N m−3, with extreme values exceeding 45 μg N m−3 and no values below 5 μg N m−3. Although the amount of data is limited, the air concentrations of NOX and particles tend to be higher in China than in Japan (Table 3).

Table 3. Annual mean values of air concentration of reactive nitrogen in Japan and China
Site locationLatitudeLongitudeAlt. (m)ClassFreq.MethodYearAnnual mean concentration (μg N m−3, 0°C, 1013 hPa)Source
  1. Source: 1, Network center of EANET; 2, Hayashi et al. (2007); 3, Hojito et al. (2006a); 4, Carmichael et al. (2003); 5, Aas et al. (2007); 6, Hu et al. (2007). Re, remote site; Ru, rural site; U, urban site. Freq.: measurement frequency; B, biweekly; C, continuous; M, monthly; NS, not specified; W, weekly. Method: measurement method; AT, automatic monitor; FP, filter pack; FT, filter trap; GT, gas trap; PS, passive sampler. NOX: nitrogen oxides. pNH4 and pNO3, particulate ammonium and nitrate, respectively.

 HokkaidoRishiri45°07′N141°12′E40ReBFP, AT2002–20070.
 AomoriTappi41°15′N141°21′E105ReBFP, AT2003–20070.
 NiigataSado-seki38°14′N138°24′E136ReBFP, AT2003–20070.
 TochigiNILGS 136°55′N139°57′E350RuMPS20043.23
 TochigiNILGS 236°55′N139°56′E310RuMPS20049.23
 TokyoOgasawara27°05′N142°13′E230ReBFP, AT2003–20070.
 NaganoHappo36°42′N137°48′E1850ReBFP, AT2003–20070.
 GifuIjira35°34′N136°41′E140RuBFP, AT2003–20070.
 ShimaneOki36°17′N133°11′E90ReBFP, AT2002–20070.
 ShimaneBanryu34°41′N131°48′E53UBFP, AT2003–20070.
 KochiYusuhara32°22′N132°56′E790ReBFP, AT2003–20070.
 OkinawaHedo26°52′N128°15′E60ReBFP, AT2003–20070.
 JinagxiYingtan (Forest Micro-Met. Sub-Station)ReNSAT, GT, FT2003–200467.65.26.512.36
 XiamenHongwen24°28′N118°08′E50UBFP20074.90.6 3.11.41
 HongkongCape D’Aequier22°12′N114°15′E1ReMPS1999–20000.64

The atmospheric deposition of Nr as the sum of the wet and dry deposition at 10 sites in Japan, eight of which were remote sites, ranged from 3.1 to 18.2 kg N ha−1 year−1 as a 5-year mean from 2003 to 2007 (Ministry of the Environment 2009). The largest deposition was recorded at a rural site, Ijira, in central Japan, and the smallest depositions were recorded at clean remote islands, such as Rishiri in northernmost Japan and Ogasawara in the Pacific Ocean. According to earlier nationwide monitoring implemented by the Ministry of the Environment, Japan, during 1983–2002, wet deposition of Nr (as bulk precipitation during 1983–1988 and wet-only precipitation after that) at 23, 29 and 29 remote, rural and urban sites was 7.0, 7.9 and 8.2 kg N ha−1 year−1, respectively. Wet deposition of Nr was greatest at the urban sites along the central part of the Japan Sea coastal regions, with a mean value of 11.4 kg N ha−1 year−1; however, in addition to this region, wet deposition of NO3 was also higher on the Pacific coast and the Southwestern Islands regions, and wet deposition of NH4+ was also higher on the Pacific coast and the East China Sea coastal regions (Hayashi et al. 2006b). In contrast, Lü and Tian (2007), who attempted to evaluate the wet deposition of NH4+ and NO3 and the dry deposition of NO2 across China from 1990 to 2003, demonstrated that atmospheric deposition peaked over south central China, with a mean value of 12.9 kg N ha−1 year−1 and a maximum value of 63.5 kg N ha−1 year−1.

Table 4 is a summary of the recent data of annual atmospheric deposition in Japan and China. The rates of wet deposition of NH4+ were on par with those of NO3 in Japan in many cases; however, the contribution of NH4+ was greater than that of NO3 at some sites. In particular, the rates of wet deposition of NH4+ outweighed those of NO3 at the Hinoki and Aoki sites in an intensive dairy farming area in Tochigi Prefecture, central Japan (Hojito et al. 2006b) (Table 4); however, these data were obtained by open bulk samplers that involved dry deposition to some extent. In contrast, the maximum level of wet deposition of inorganic N (17.7 kg N ha−1 year−1) was observed at the Kajikawa site in Niigata Prefecture, central Japan, a remote forest site in the central part of the Japan Sea coastal region (Kamisako et al. 2008) (Table 4). In addition, Sase et al. (2008) used filtering-type bulk samplers to prevent evaporation of the collected samples or deterioration as a result of reactions with gases or aerosols in the atmosphere. Although limited data are available, a range of 1.3–5.5 kg N ha−1 year−1 of wet deposition of organic N was reported in central Japan (Hayashi et al. 2007; Hojito et al. 2006b). The rates of wet deposition of NH4+ were also on par with those of NO3 in China; however, NH4+ showed larger contributions in many cases (Table 4). The larger NH4+ wet deposition was mainly ascribed to the urban and agricultural sources of NH3 in the urban and non-urban sites, respectively. The rates of wet deposition of inorganic N at remote sites in China were similar to those in Japan, whereas those at other sites in China (22.7 kg N ha−1 year−1 as a mean value of Table 4 involving open bulk data) were several-fold greater than those in Japan. The rates of wet deposition of organic N in China ranged from 0.7 to 26.4 kg N ha−1 year−1 (Table 4) and these rates are also likely to be greater than the rates in Japan, excluding inland remote sites.

Table 4. Annual wet and dry deposition of reactive nitrogen in Japan and China
Site locationLatitudeLongitudeAlt. (m)ClassFreq.MethodYearAnnual wet deposition
(kg N ha−1 year−1)
Annual dry deposition
(kg N ha−1 year−1)
  1. Source: 1, Hayashi et al. (2006b); 2, Network center of EANET; 3, Ministry of the Environment (2009); 4, Hayashi et al. (2009c); 5, Kamisako et al. (2008); 6, Hayashi et al. (2007); 7, Hojito et al. (2006b); 8, Zhang et al. (2008a); 9, Liu et al. (2006); 10, Zhang et al. (2008b); 11, Zhang (2006); 12, Wang et al. (2004); 13, Xie et al. (2008); 14, Aas et al. (2007); 15, Hu et al. (2007); 16, Wang et al. (2008b); 17, Du et al. (2008); 18, Fang et al. (2009); 19, Zhang et al. (2007). Re, remote site; Ru, rural site; SU, suburban site; U, urban site. Freq.: measurement frequency; B, biweekly; E, rain event; M, monthly; NS, not specified; W, weekly. Method: measurement method; Inf, inferential method (resistance model); OB, open-bulk; NS, not specified WOB, wet only (bulk sampling was conducted during 1983–1988); WO, wet-only. OrgN, organic nitrogen; pNH4, particulate ammonium; pNO3, particulate nitrate.

 All Japan23 remote sites   ReD,W,B,MWOB1983–20023.73.41
 All Japan29 rural sites   RuD,W,B,MWOB1983–20024.53.41
 All Japan29 urban sites   UD,W,B,MWOB1983–20024.73.61
 TochigiNILGS 136°55′N139°57′E350RuMOB20047.07.11.87
 TochigiNILGS 236°55′N139°56′E310RuMOB20047.87.12.27
 LiaoningDalian   RuDOB20068.110.310.98
 BeijingChina Agric. Univ.39°57′N116°18′E SUEOB1999–200419.89.59
 BeijingDongbeiwang40°03′N116°17′E SUEOB2003–200424.110.39
 BeijingChangping   RuDOB2005–200620.113.34.68
 BeijingChina Agric. Univ.   SUDOB2005–200617.410.510.38
 BeijingDongbeiwang   SUDOB2005–200627.012.59.58
 BeijingShunyi (Yanhe)   RuDOB2005–200616.18.53.18
 BeijingFangshan   RuEOB2005–200618.311.310
 BeijingBeijing Academy of Agro-Forestry Sciences   UEOB2005–200618.710.310
 BeijingShunyi (Liangshan)   RuEOB2005–200619.48.710
 HebeiBaoding   RuDOB2005–200614.18.64.68
 HebeiQuzhou   RuDOB2005–200610.55.65.58
 HebeiWuqiao   RuDOB2005–200610.15.96.68
 ShandongHuimin   RuDOB2005–200619.111.08.68
 ShandongQingdao   RuDOB20067.95.57.48
 Shanghai44 urban, suburban, and rural sites    NSNS1998–200326.631.511
 ChangshuChangshu31°33′N120°42′E RuEOB2001–200312.99.54.712
 ChangshuChangshu31°32′N120°42′E RuEWO2003–200516.911.013
 ChangshuWuxi31°36′N120°28′E RuEWO2003–200515.610.713
 SichuanGonggashan   ReDOB2005–20064.72.626.48
 ZhejiangFenghua   RuDOB2004–200520.010.723.18
 JiangxiYingtan (Forest Micro-Met. Sub-Station)   ReMOBInf2003–20042003–20045.212.439.
 JiangxiYingtan28°12′N117°00′E RuMWOInf2004–20052004–200512.314.
 ZhuhaiXiang Zhou22°16′N113°34′E40UDWO2000–200710.84.12
 ZhuhaiZhuxian Cavern22°12′N113°31′E45UDWO2000–20059.45.22
 GuangdongDinghushan23°10′N112°10′E ReNSOB200323.210.918
 Inner MongoliaDuolun   ReDOB200613.23.50.7 8
 XingjiangUrumchi   RuDOB20063.61.31.58
 TibetLinzhi   ReDOB2005–20062.40.05.68
 Yellow SeaQianliyan Island   ReEWO2000–20030.
 East China SeaShengsi Archipelago   ReEWO2000–20034.

In contrast to the large quantity of data for wet deposition, limited information has been available for dry deposition in Japan and China. Case studies in Japan have shown smaller rates of dry than wet deposition (Table 4). However, dry deposition accounted for 40% at maximum (Ministry of the Environment 2009) and 51% (Hayashi et al. 2007) of the atmospheric deposition of inorganic N. The actual contribution of dry deposition to total deposition appears greater because these studies excluded the dry deposition of NO2. Two case studies in China have shown an extremely large rate of dry deposition of NH3, 22.0 and 39.0 kg N ha−1 year−1 (Hu et al. 2007; Wang et al. 2008b) (Table 4). The large deposition resulted from the high air concentrations of NH3 arising from neighboring livestock facilities, such as piggeries. Local sources strongly affect dry deposition of NH3. All case studies shown in Table 4 adopted the inferential method (or resistance model) (e.g. Hicks et al. 1987) to estimate the dry deposition flux. By definition, the resistance model infers a deposition velocity on the basis of micrometeorological data and parameterizations of the effects of surface conditions, such as vegetation and soil. However, the parameterizations depend largely on the empirical equations derived from early studies in Europe and the USA. Although methodological improvements have been progressing (e.g. Matsuda 2008; Zhang et al. 2003), the development of an improved inferential method that takes the key features of Japanese and Chinese agroecosystems into consideration is still required.

Environmental concerns

Shindo et al. (2005) revealed that long-term N deposition is an important factor to determine the N concentration in stream water in natural ecosystems on the basis of 1 km × 1 km grid estimation for the whole of Japan. Xie et al. (2007) also recognized the importance of atmospheric deposition as a source of N pollution of water bodies, although these researchers held to the view that N pollution in the Taihu Lake region in south-eastern China should be primarily ascribed to the direct discharges of urban sewage and rural human and animal excreta without treatment.

In 2002–2007, in a mountainous forested catchment in Niigata, central Japan, which received the highest level of N deposition, leaching of N to stream water occurred even during the growing season, and the rate of N deposition certainly contributed to the high NO3 concentrations in the stream water (Kamisako et al. 2008). Matsubara et al. (2009) reported the current serious status of the river water chemistry in Niigata and Gifu, central Japan, which is characterized by a significant long-term decline in pH (Niigata, 1986–2003; Gifu, 1988–2003). These researchers pointed out the possible impact of acid deposition, including inorganic N species, on the acidification of river water because of the higher levels of deposition load than those in Europe and the USA.

Earnest studies have also been conducted in China in relation to the impact of N deposition on forested catchments. Du et al. (2008) concluded that the Shaoshan Forest, a subtropical evergreen mixed forest in south-central China, was still far from N saturation, despite the high N deposition. In contrast, Fang et al. (2009) found large and rapid N leaching from a mature monsoon broadleaf forest in southern China, which accounted for 80% of N deposition, perhaps as a result of long-term N accumulation by the high N deposition of >30 kg N ha−1 year−1 over the course of 15 years. An increase in N deposition generally enhances the decomposition of organic matter because of the improvement in N nutrition (e.g. Månsson and Falkengren-Grerup 2003). However, further additions of N to subtropical and tropical forests in southern China, which have already been exposed to high N deposition, have resulted in significant reductions in litter decomposition and soil respiration (Fang et al. 2007; Mo et al. 2006, 2007, 2008). Mo et al. (2007) also reported that the reduction occurred primarily in the hot and wet growing season. More attention should be given to the diverse ecosystems in East Asia and to their various responses against N load.

Future research needs

Airborne N has been imposing unrelenting N loads on Japan and China and this situation is likely to deteriorate in the future. Japanese and Chinese agroecosystems are both a source and a sink of airborne N. The types of studies that could best address this issue need to be determined.

Methodological development

A well-developed in situ observation of airborne N regarding its emission, air concentration and atmospheric deposition is needed. Of these three issues, the observation of dry deposition is particularly problematic because of the methodology required. Furthermore, dry deposition denotes only the downward flux of the atmosphere–land exchange. The inverse flux, emission, can occur simultaneously and can occasionally become quite large, in particular for NH3 in agroecosystems. Therefore, bidirectional fluxes as exchange should be determined. For this purpose, micrometeorological techniques, such as eddy covariance and gradient methods, are suitable (e.g. Fowler et al. 2001). However, it is practically difficult to deploy micrometeorological observations widely and simultaneously in Japan and China. Thus, a simple, but quantitative method should be developed. The inferential method (Hicks et al. 1987) is a candidate, but problems remain regarding the accuracy of parameterization. It is, therefore, a good way to develop an improved inferential method verified by micrometeorological techniques (e.g., Hole et al. 2008; Spindler et al. 2001) for each type of typical land use and crop calendar in the Japanese and Chinese agroecosystems.

Data quality

The quantification of uncertainties is also important. For example, a passive sampler is advantageous because it requires no electricity and is inexpensive, thus allowing simultaneous monitoring over a wide area (e.g. Namieśnik et al. 2005). The disadvantage of this system is the lack of accuracy and low time resolution. When using such a simple method, a sufficient number of samplers for each site are needed to statistically evaluate the errors in measurement. Quality control and quality assurance should be carried out for analysis with due caution.

However, the most important uncertainty is that the exchange flux may involve a systematic error because of the correlation between the air concentrations and the micrometeorological parameters. In fact, the flux of a substance that has a clear diurnal pattern in concentration will inevitably incur a degree of error if the time resolution of the measurements is low, that is, daily or less. For example, NH3 fluxes determined by the gradient method on the basis of a daily mean resulted in a 30% underestimation compared with those determined on the basis of day/night separated means (Hayashi et al. 2009d). This systematic error cannot be eliminated by technical improvements. Enhancing the time resolution of the measurements or the development of an ingenious method of measurement that discriminates between day and night is required.

Wide-scale assessment

A regional-scale to national-scale observation is essential to assess the actual status of an airborne N load, in which not only emission and deposition, but also many other variables can be targeted, for example, spatial distributions and temporal changes in polluted air mass, vegetation, farmlands, meteorological factors and soil conditions. In situ observations are also required for site representativeness in the target area to be applied to a large-scale assessment (e.g. Marner and Harrison 2004). In contrast, the concept of a ‘hot spot’ (Sutton et al. 2009) is important with respect to a substance, typically NH3, that is strongly affected by local sources. Area sources, such as farmlands, may show daily and seasonally large changes in the emission and deposition of Nr, the actual status of which in Japan and China needs to be elucidated in detail, including the effects of meteorological factors and soil and plant conditions on the exchange fluxes. Observation techniques for emission and dry deposition of Nr are still being developed in Japan and China; however, a simple method applicable to network observation with sufficient accuracy and time resolution is necessary, or at least a method that would provide quantitative information of uncertainty. Long-term observations are also quite important. Long-term multipoint data are highly valuable for the assessment of temporal trends and for numerical modeling of airborne N load. An efficient and effective measure is to reinforce the existing network, for example, the Acid Deposition Monitoring Network in East Asia (EANET). The EANET is also making a considerable effort to relocate sites and estimate dry deposition flux (Network Center of EANET 2006).

Numerical models with respect to airborne N load (e.g. Hertel et al. 2006; Kangas and Syri 2002; Zhang et al. 2002) are indispensable for spatio-temporally extrapolated estimations. For example, the source–receptor relationships of airborne N that atmospheric transportation models can give are fundamental for discussions of effective countermeasures. However, the gaps between the observation and the model, for example, bias in site selection and differences in temporal scales, need to be bridged. Modeling the interface between the land and the atmosphere has not progressed because relevant variables, such as land use, vegetation and soil conditions, are various and fluctuant. It is, therefore, to be expected that a numerical model of atmosphere–land exchange will have a large degree of uncertainty. Sensitivity analysis and model intercomparison are needed. Furthermore, in addition to data input from observation to modeling, feedback input from modeling to observation is important regarding the necessary variables and their spatio-temporal resolutions.

The most essential measure against airborne N load in relation to agricultural activities is to reduce Nr emissions at a farm level. Along the lines of Sutton et al. (2009), measures need to consider a multi-effect (human health, climate change, acidification, eutrophication and related biodiversity loss), multi-media (air, water and soil) and multi-scale (points, regional, Asian and global) framework. These measures also have to consider possible trade-offs. For example, a measure to reduce gaseous Nr emissions loses its value when it leads to an increase in nitrate leaching. Erisman et al. (2008) reported the true nature of the issue and suggested that policies should be directed to decreasing the production of Nr from N2, or to increasing the conversion of Nr back into N2, for example, by denitrification.


Agroecosystems are both a source and a sink of Nr. Intensive livestock farming in Japan and the overuse of N fertilizers in China are likely the most fundamental causes of the airborne N load in relation to Japanese and Chinese agriculture. Estimated NH3 emissions in China in 1995 from chemical fertilizers and organic fertilizers, such as animal excreta, were 3.56 and 2.04 Tg N, respectively, whereas those in Japan were 0.059 and 0.069 Tg N, respectively.

Emitted Nr is widely transported in the atmosphere and then deposited onto the ground surface. Atmospheric N deposition was up to approximately 20 and 60 kg N ha−1 year−1 in Japan and China, respectively. The contribution of dry deposition is on par with that of wet deposition. The contribution of NH3 dry deposition becomes particularly high in an area that is affected by livestock farming.

In future studies, a simple, but accurate method to determine dry deposition flux, which is applicable to simultaneous and multipoint observations, needs to be developed. Cross relationships among in situ observation, remote sensing and numerical modeling also need to be established.


This study is part of the bilateral program between Japan and China provided by the Japan Science and Technology Agency as the Strategic International Cooperative Program entitled “Comparative Study of Nitrogen Cycling and Its Impact on Water Quality in Agricultural Watersheds in Japan and China” and the Chinese National Natural Science Foundation (Grant No. 40721140018). This study was also supported in part by a Grant-in-Aid for Scientific Research (A), No.19201008, provided by the Ministry of Education, Culture, Sports, Science, and Technology, Japan.