Effect of chemical fertilizer and manure application on N2O emission from reed canary grassland in Hokkaido, Japan

Authors


T. JIN, Soil Science Laboratory, Graduate School of Agriculture, Hokkaido University, Kita 9 Nishi 9, Kita-ku, Sapporo, Hokkaido 060-8589, Japan. Email: jintao@chem.agr.hokudai.ac.jp

Abstract

We evaluated the effect of chemical fertilizer and manure applications on N2O emission from a managed grassland by establishing three treatment plots of chemical N fertilizer (chemical fertilizer), manure combined with chemical N fertilizer (manure) and no N fertilizer (control) at the Shizunai Experimental Livestock Farm in southern Hokkaido, Japan. The N2O fluxes from the soils were measured using a closed-chamber method from May 2005 to April 2008. Soil denitrifying enzyme activity (DEA) in the root-mat layer (0–2.5 cm) and in the mineral soil layer (2.5–5 cm) of each treatment plot was measured using an acetylene inhibition method after treatment with NO3-N addition, glucose addition, both NO3-N and glucose addition or neither NO3-N nor glucose addition. Annual N2O emission ranged from 0.6 to 4.9 kg N2O-N ha−1year−1, with the highest emission observed in the manure plot and the lowest in the control plot. The chemical fertilizer induced emission factor (EF) (range: 0.85–1.32%) was significantly higher than the manure-induced EF (range: 0.35–0.85%). The denitrification potential of the soil horizons was measured with the addition of both NO3-N and glucose, and was significantly higher in root-mat soil than in mineral soil. Soil DEA in the root mat with the addition of NO3-N with and without the addition of glucose had a significantly positive correlation with soil pH (P < 0.05). Soil pH was significantly influenced by N source, which was significantly lower in the chemical fertilizer plot than in the control and manure plots. For a fixed quantity of available N, the application of manure could result in higher N2O emission compared with chemical fertilizer, owing to higher soil pH values under manure application than under chemical fertilizer application.

Introduction

Nitrous oxide (N2O) is one of the most important radiatively active trace gases in the atmosphere and contributes at least 5% to the observed global warming at present (Myhre et al. 1998). The atmospheric concentration of N2O has increased from a pre-industrial value of approximately 270–319 p.p.b. in 2005 and continues to increase as a result of human activities (Intergovernmental Panel on Climate Change 2007). Agriculture as a whole (i.e. animal excreta, denitrification of leached nitrate) contributes approximately 80% of anthropogenic N2O emissions (Brown et al. 2001). Direct and indirect emissions from agricultural systems are now thought to contribute approximately 6.2 Tg N2O-N year−1 to the total global source strength of 17.7 Tg N2O-N year−1 (Kroeze et al. 1999). Approximately 57% of the global atmospheric sources of N2O are estimated to be related to emissions from soils (Mosier et al. 1998).

Nitrous oxide is produced in soils primarily by the microbial processes of nitrification and denitrification (Conrad 1996; Tiedje 1988). Nitrification is the biological oxidation of ammonium to nitrite or nitrate under aerobic conditions, and denitrification is the reduction of nitrate to N2O and N2 when the supply of oxygen is limited. Increasing soil N availability as a result of increased N inputs by the application of chemical fertilizer and manure and atmospheric deposition have greatly enhanced N2O emissions from soils (Kroeze et al. 1999) by influencing nitrifying and denitrifying enzyme activity. Chemical fertilizer and animal waste are the two most important sources of direct N2O emissions from agricultural soils (Mosier et al. 1998). The default Intergovernmental Panel on Climate Change (IPCC) emission factor, that is, the percentage of applied N emitted as N2O, is 1.0%, regardless of the fertilizer type (Intergovernmental Panel on Climate Change 2006). However, the type of N inputs to the fields may affect the N2O emission rate in different ways, leading to different patterns of N2O emissions from inorganic and organic N fertilizers. The addition of inorganic N increases N2O emission by affecting the process of nitrification and denitrification by increasing the available NH4+-N and NO3-N substrates. Organic fertilizers with a high and easily mineralizable organic C content stimulate microbial activity and thus N2O emissions (Chadwick et al. 2000). The application of chemical N fertilizer results in short-lived N2O peaks (Dobbie and Smith 2003). However, N2O losses from manure plots extended over a longer period of time and were greater in magnitude than the losses from chemical N fertilization (Jones et al. 2007). Higher N2O fluxes from manure and sewage applications compared with the flux from chemical N fertilizers have also been observed in other studies (Scott et al. 2000). High N2O fluxes from manure treatments can be partly explained by the higher total N input than chemical fertilizer treatments, providing more available N that can be mineralized over a longer period of time. Another reason for the increased N2O losses could be the addition of organic C with the manure, which is known to stimulate denitrification. McTaggart et al. (1997) reported that the carbon supply from slurry spread onto grasslands stimulated N2O production, resulting in a fourfold larger N2O loss compared with the application of NH4-NO3, although the total rate of N application was similar in both treatments.

Grassland is an important ecosystem to support the production of herbivorous livestock (Soussana et al. 2007). The application of chemical fertilizer and animal manure to grasslands has been conducted to increase grass production, particularly in developed countries where grassland-based livestock production is important (Bouwman et al. 2002). However, N application to grasslands also poses a risk of N loss to the environment in the form of N2O emission. The objective of the present study was to clarify the effect of fertilizer and manure application on N2O emission and to identify the factors controlling N2O emission from the grassland.

Materials and methods

Study site

The present study was conducted on a managed grassland located at the Shizunai Experimental Livestock Farm at the Field Science Center for Northern Biosphere, Hokkaido University, Southern Hokkaido, Japan (42°26′N, 142°29′E). The site is characterized by a humid continental climate with cold winters and cool summers. From 1979 to 2000, the mean annual precipitation and air temperature for this region were 1365 mm and 7.9°C, respectively. The soil is derived from Tarumae (b) volcanic ash and is classified as Thaptic Melanudands (Soil Survey Staff 2006; Mollic Andosol [IUSS Working Group WRB 2006]). A layer of 3-cm thick root mat was found on the top and a 21-cm thick Ap-layer was found under the root mat in a survey conducted in August 2004 (Shimizu et al. 2009). The C and N contents in the Ap-layer were 3.7% and 0.33%, respectively, and the C : N ratio was 11:1. The dominant grass species at this site were reed canary grass (Phalaris arundinacea L.) and foxtail grass (Alopecurus pratensis L.). Harvesting of the grass was carried out twice per year (21 June and 11 August in 2005, 27 June and 23 August in 2006 and 18 June and 18 August in 2007) in accordance with the local practice.

Experimental setup

Three experimental plots were set up on the study site: one for treatment with chemical fertilizer (chemical fertilizer plot), another for treatment with beef cattle manure and chemical fertilizer (manure plot) and the final plot with no N fertilizer or manure (control plot). The treatments were established in the spring of 2005. Eighteen subplots (5 m × 4 m) were established for the chemical fertilizer, manure and control plots with six and four replications from May 2005 to April 2007 and from May 2007 to April 2008, respectively.

Table 1 shows the date of application and the application rates of the chemical fertilizer and manure. The N application rates in the chemical fertilizer plot were at the recommended level for this site on the basis of soil tests, and were 164 kg N ha−1 year−1 in 2005 and 184 kg N ha−1 year−1 in 2006 as ammonium sulfate and ammonium phosphate (Table 1). To determine the N2O emission from grassland soil under the local N fertilization level, N fertilization decreased to 74 kg N ha−1 year−1 in the chemical fertilizer plot according to recommendations by the Shizunai Experimental Livestock Farm staff. The pH values of the manure that was used in 2005, 2006 and 2007 were 8.3, 8.8 and 9.1, respectively. The manure application rates were the optimum rates used by farmers in the region, and were based on adequate amounts of potassium (K) application to the fields. Beef cattle manure with bedding litter (bark) was applied to the manure plot, and the application rates were 44 Mg fresh matter (FM) ha−1 (236 kg N ha−1 and 5.8 Mg C ha−1) in May 2005, 43 Mg FM ha−1 (310 kg N ha−1 and 6.0 Mg C ha−1) in May 2006, and 43 Mg FM ha−1 (331 kg N ha−1 and 7.7 Mg C ha−1) in May 2007 (Table 1). In the manure plot, the N supply rates from the manure were estimated by multiplying the application rates by the N mineralization rate, and the differences between the supply rates in manure and the application rates in the fertilizer plot were supplied by chemical fertilizer. The N mineralization rates were estimated based on Uchida’s model (Shiga et al. 1985), which was developed in Japan, and were 13.2, 7.0 and 5.5%, respectively, in the first, second and third years after application. The mineralization rates of P and K from the manure were estimated based on the handbook of animal waste management and utilization in Hokkaido 2004 (Anon. 2004). The P mineralization rate was 20, 10 and 0% and the K mineralization rate was 70, 10 and 0% in the first, second and third years after application, respectively.

Table 1.   Date of application and the application rates of chemical fertilizer and manure over the study period
TreatmentDateFertilizer typeApplication rates (kg ha−1)
CNP2O5K2O
  1. aChemical fertilizer is comprised of ammonium sulfate, ammonium phosphate, potassium sulfate and potassium magnesium sulfate. bBeef cattle manure with bedding litter was applied in the manure plot.

Control2005/5/11Chemical fertilizera0000
2005/7/4Chemical fertilizera0000
2006/5/9Chemical fertilizera0000
2006/7/10Chemical fertilizera0000
2007/5/12Chemical fertilizera001473
2007/7/5Chemical fertilizera00737
Chemical fertilizer2005/5/11Chemical fertilizera010323168
2005/7/4Chemical fertilizera0612397
2006/5/9Chemical fertilizera012450177
2006/7/10Chemical fertilizera0591897
2007/5/12Chemical fertilizera0491473
2007/7/5Chemical fertilizera025737
Manure2005/5/11Manureb5833236191266
2005/7/4Chemical fertilizera0133770
2006/5/9Manureb5958310212167
2006/5/9Chemical fertilizera071033
2006/7/10Chemical fertilizera059697
2007/5/12Manureb7714331342336
2007/7/5Chemical fertilizera02100

N2O and NO fluxes

We defined the crop growing season as a 7-day moving average of daily air temperature above 5°C and the non-growing season as the remaining time (Shimizu et al. 2009). The growing season was 215 days in 2005 (from 10 April to 10 November), 218 days in 2006 (from 15 April to 18 November) and 220 days in 2007 (from 13 April to 18 November). The N2O and NO fluxes from the soil to the atmosphere were measured using the static closed chamber method on the control, fertilizer and manure plots (Shimizu et al. 2009). Flux measurements were conducted in 2–28 day intervals during the crop growing season and 10–30 day intervals during the non-growing season, and between 08.00 and 11.00 hours on each measuring day to minimize the effect of diurnal temperature variation. The stainless steel chambers were 40 cm in diameter and 30 cm high in the chemical fertilizer and manure plots, and 20 cm in diameter and 25 cm high in the control plots. The chambers were placed directly into the soil to a depth of approximately 3 cm, 12 h before the measurement of each subplot, and contained no above-ground biomass in the chemical fertilizer, manure and control plots. Before closing the chamber, a 250 mL gas sample from the headspace of each chamber was extracted into a Tedlar bag for NO analysis, and a 20 mL gas sample was injected into an evacuated vial (10 mL) for N2O analysis. This measurement was regarded as time 0 min. After 20 min or 30 min under closed-chamber conditions, a 250 mL headspace gas sample was extracted from each chamber into a bag and 20 mL was injected into a vial. From these bag samples, NO gas concentrations were determined in a laboratory within 16 h using a Chemiluminescence N Oxide Analyzer (Model 265P; Kimoto Electric, Osaka, Japan). The N2O gas concentrations were determined in a laboratory within 1 month using an electron capture detector (ECD) gas chromatograph (model GC-14B; Shimadzu, Kyoto, Japan) from the samples in the vials.

Gas fluxes were calculated from the change in gas concentration in the chamber against closure time:

image

where F is the gas flux (μg N m−2 h−1 for N2O), ρ is the gas density (N2O-N = 1.26 × 109 μg m−3), h is the height of the chamber from the soil surface (m), Δc/Δt is the change in gas concentration inside the chamber during the sampling period (m3 m−3 h−1) and T is the average air temperature during the sampling period (°C). A positive flux denotes emission from the soil and a negative flux denotes uptake from the atmosphere.

The cumulative gas flux was calculated as follows:

image

where Ri is the mean gas flux (mg m−2 hour−1) of the two successive sampling dates, Di is the number of days in the sampling interval and n is the number of sampling times. The cumulative periods of 2005, 2006 and 2007 were calculated from 10 April 2005 to 14 April 2006, from 15 April 2006 to 18 April 2007, and from 19 April 2007 to 4 April 2008, respectively.

Emission factor

The N2O emission factor (EF) for chemical fertilizer and manure (kg N2O-N (kg N input)−1) was calculated as follows:

image

Environmental variables

Daily precipitation was obtained from the Sasayama AMeDAS (Automated Meteorological Data Acquisition System) station by the Japan Meteorological Agency. Air temperature and soil temperature at a depth of 5 cm were measured at the same time as the flux measurements were taken using a thermistor thermometer (CT220; CUSTOM, Tokyo, Japan), and the soil moisture content at a depth of 0–6 cm was measured using the frequency domain reflectometry (FDR) method (DIK-311A; Daiki, Saitama, Japan). Soil core samples (14 cm diameter, 13 cm height) were collected in April 2007 and calibration curves were made to calculate water-filled pore space (WFPS) from the FDR device reading (m3 m−3) and the percentage total porosity (Linn and Doran 1984). The percentage total porosity was measured using a 100 mL soil core collected in April 2007 and was regarded as constant throughout the study period because there was no tillage.

Soil chemical analyses

Three replicate soil samples at a depth of 0–5 cm from the ground surface were collected from April to November in all treatment plots. Within 48 h of sampling, the soil samples were sieved through a 2 mm sieve and any stones and roots were removed. The soil samples were then immediately extracted in deionized water (1:5) and in 2 mol L–1 KCl (1:10), and the extracts were stored at 4°C until analysis for dissolved nutrients after being filtered through 0.2-μm membrane filters. The water-soluble organic carbon (DOC) content in the deionized-water-extract solution was analyzed using a total organic carbon (TOC) analyzer (TOC 5000A; Shimadzu, Japan). The concentrations of NO2-N and NO3-N in the deionized-water-extract solution were analyzed by ion chromatography (Dionex QIC Analyzer; Dionex Japan, Osaka, Japan). The concentration of NH4+-N in the 2 mol L−1 KCl extracted solution was determined using the indophenol-blue method (UV mini 1240; Shimadzu). Soil pH was measured in the deionized-water-extract solution with a combined electrode pH meter (F-8 pH meter; Horiba, Kyoto, Japan).

Measurement of soil denitrifying enzyme activity

To measure soil denitrifying enzyme activity (DEA), three replicate soil samples were taken from all treatment plots in the root-mat layer (0–2.5cm depth) and mineral soils layer (2.5–5 cm depth) in April, June and August 2007. The root-mat soil samples were cut into small pieces (diameter 1 cm) and any stones or roots were removed from the mineral soil samples by passing the soil through a 2 mm sieve within 48 h after sampling. The three replicate soil samples were then mixed and kept in a refrigerator at 4°C until analysis. The DEA was determined by an acetylene block technique, which inhibits the final conversion of N2O-N2 gas (Tiedje 1994). Soil samples were incubated under anaerobic conditions at 25°C with a solution treated with: (1) chloramphenicol (1 g L−1) (Chl), (2) chloramphenicol (1 g L−1) and NO3-N (200 mg N L−1 as KNO3) (Chl+N), (3) chloramphenicol (1 g L−1) and organic C (2 g C L−1 as glucose) (Chl+C), (4) chloramphenicol (1g L−1), NO3-N (200 mg N L−1 as KNO3) and organic C (2 g C L−1 as glucose) (Chl+N+C). Fresh soil (15 g) was placed into a 100 mL conical flask and 15 mL of the treated solution was added to the flask. The flasks were evacuated and flushed four times with N2 to ensure anaerobic conditions, and acetylene (C2H2) gas was added to a final concentration of 10% (10 kPa) in the headspace. The headspace gas was sampled using a syringe after 2 h and 4 h and denitrification rates were calculated from the linear increment of N2O production against time. The denitrification potential of the soil horizons was measured with the addition of both NO3-N and soluble C sources as proposed by D’Haene et al. (2003).

Statistical analyses

ANOVA and Pearson correlation analyses were performed using SPSS 13.0 (SPSS Inc., Chicago, US). Linear regression and other statistical analyses were carried out using Excel 2003 (Microsoft corp., Singapore). Two-way ANOVAs and Tukey’s tests were used to compare the mean difference (P < 0.05) of a given variable between treatment plots and years. Three-way ANOVAs and Tukey’s tests were used to compare the mean difference (P < 0.05) in soil N2O fluxes among the seasons, treatment plots in the field experiment and years; and in soil DEA among the treatment plots in the field experiment, soil layers and incubation treatments.

Results

Soil temperature and moisture

Daily precipitation is shown in Fig. 1a. Annual precipitation was 1176 mm from mid April 2005 to mid April 2006, 1047 mm from mid April 2006 to mid April 2007, and 879 mm from mid April 2007 to the beginning of April 2008. These values are smaller than the mean annual precipitation (1365 ± 215 mm) recorded from 1989 to 2000.

Figure 1.

 Seasonal patterns in (a) precipitation, (b) water-filled pore space (WFPS) at a depth of 6 cm and (c) soil temperature at a depth of 5 cm. Data of WFPS and soil temperature are mean ± standard deviation (n = 4–6).

Soil moisture content expressed as WFPS at a depth of 0–6 cm is shown in Fig. 1b. The WFPS from April to November was influenced by precipitation and low soil moisture was observed with low precipitation in August 2006 and June 2007. From December to March, soil moisture was not observed because of soil freezing. In the winter of 2006/2007, soil freezing began at the beginning of December 2006, reaching a maximum depth of 17.75 cm on 9 March 2007, and then thawed in early April 2007. In the winter of 2007/2008, soil freezing also began at the beginning of December 2007, but reached a maximum of 29 cm on 12 March 2008, and then thawed in early April 2008. The soil freezing depth was not observed in the winter of 2005/2006.

The soil temperature at a depth of 5 cm is shown in Fig. 1c. The soil temperature increased from April, reaching a maximum from July through to August, and then decreased gradually. The soil temperature was approximately 0°C from December to March. There was no difference in soil temperature between the chemical fertilizer and manure plots, but the soil temperature was higher in the control plot than in the chemical fertilizer and manure plots (P < 0.05).

N2O fluxes

Seasonal patterns in the N2O fluxes were mainly driven by seasonal variation in air and soil temperatures, which were higher in summer and lower in winter, but were also influenced by fertilization (Fig. 2a). The N2O fluxes in the chemical fertilizer and manure plots increased after the application of manure or chemical fertilizer. These fluxes remained at a higher level than in the control plot until the beginning of September (Fig. 2a). A three-way ANOVA showed that there was a significant difference in N2O fluxes between the non-growing and growing seasons (P < 0.001) and between each treatment (P < 0.05). A significant interaction between season and treatment in N2O fluxes was also observed (P < 0.05) (Table 2). In the growing season, the mean N2O fluxes in 2005, 2006 and 2007 were 12.3, 12.9 and 16.7 ug N2O-N m–2h−1 for the control plot, 85.3, 83.9 and 36.5 ug N2O-N m–2h−1 for the chemical fertilizer plot and 101.9, 187.6 and 50.6 ug N2O-N m–2h−1 for the manure plot, respectively (Table 3). The N2O fluxes in the growing season were significantly higher in the chemical fertilizer and manure plots than in the control plot (P < 0.01), but there was no significant difference between the chemical fertilizer and manure plots. In the non-growing season, the N2O fluxes were lower and stable with mean values of 0.3, 2.6 and 4.7 ug N2O-N m–2h−1 in the control plot, 2.7, 6.0 and 7.4 ug N2O-N m–2h−1 in the chemical fertilizer plot and 2.7, 3.6 and 3.5 ug N2O-N m–2h−1 in the manure plot in 2005, 2006 and 2007, respectively (Table 3). There was no significant difference in the mean N2O flux between each treatment plot.

Figure 2.

 Seasonal patterns in (a) soil N2O fluxes and (b) soil NO fluxes. Data represent mean ± standard deviation (n = 4–6). Full arrows indicate the dates of the application of chemical fertilizer and the dotted arrows indicate the date of manure application. The growing season was 215 days in 2005 (from 10 April 2005 to 10 November 2005), 218 days in 2006 (from 15 April 2006 to 18 November 2006) and 220 days in 2007 (from 13 April 2007 to 18 November 2007).

Table 2.   ANOVA results for mean N2O fluxes
Sourced.f.Mean squareFP-value
Season1179,974.0417.5210
Year213,990.751.3620.258
Treatment239,922.533.8870.022
Season × year215,307.661.490.228
Season × treatment238,549.473.7530.025
Year × treatment47,659.250.7460.562
Season × year × treatment47,454.570.7260.575
Error21010,271.71  
Table 3.   Mean N2O fluxes from the control, chemical fertilizer and manure plots
SeasonTreatmentMean N2O fluxes (ug N2O-N m−2 h−1)
200520062007
  1. Data are mean (standard deviation) (n = 4–6). We defined the crop-growing season as a 7-day moving average of daily air temperature above 5°C and the non-growing season as the remaining time (Shimizu et al. 2009). The growing season was 215 days in 2005 (from 10 April 2005 to 10 November 2005), 218 days in 2006 (from 15 April 2006 to 18 November 2006) and 220 days in 2007 (from 13 April 2007 to 18 November 2007).

Growing seasonControl12.3 (12.7)12.9 (15.23)16.7 (27.6)
Chemical fertilizer85.3 (94.1)83.9 (66.3)36.5 (53.2)
Manure101.9 (102.1)187.6 (301.6)50.6 (48.9)
Non-growing seasonControl0.3 (0.6)2.6 (6.0)4.7 (3.1)
Chemical fertilizer2.7 (2.3)6.0 (11.4)7.4 (13.7)
Manure2.7 (3.9)3.6 (5.1)3.5 (8.0)

The annual N2O emission in the control, chemical fertilizer and manure plots ranged from 0.6 to 0.7, from 1.4 to 3.0 and from 2.1 to 4.9 kg N2O-N ha−1year−1 during 2005, 2006 and 2007, respectively (Table 4). Application of both chemical fertilizer and manure stimulated annual cumulative N2O emissions, and the highest significant annual cumulative N2O emission was observed in the manure plot, followed by the chemical fertilizer plot. Application of chemical fertilizer contributed to 76.9, 79.2 and 47.2% of the total N2O emission from the chemical fertilizer plot in 2005, 2006 and 2007, respectively. In the manure plot, N2O emission from the applied chemical fertilizer and manure contributed to 81.8, 87.4 and 67.6% of the total N2O emission in 2005, 2006 and 2007, respectively. The chemical fertilizer induced EF was 1.32, 1.30 and 0.85% in 2005, 2006, and 2007, respectively. The manure-induced EF was significantly lower than the fertilizer-induced EF (P < 0.001), and was 0.51, 0.85 and 0.35% in 2005, 2006 and 2007, respectively (Table 5).

Table 4.   Cumulative N2O emissions from the control, chemical fertilizer and manure plots
TreatmentCumulative N2O emissions (kg N2O-N ha−1 year−1)
200520062007
  1. Data are mean (standard deviation) (n = 4–6) and different small letters denote significant differences at P < 0.05 between each treatment.

Controla0.7 (0.4)0.6 (0.3)0.7 (0.5)
Chemical fertilizerb2.8 (0.7)3.0 (0.8)1.4 (0.5)
Manurec3.6 (1.2)4.9 (2.8)2.1 (0.6)
Table 5.   N2O emission factors for the chemical fertilizer and manure (kg N2O-N [kg N input]−1)
 N2O emission factor (%)
200520062007
  1. Data represent mean (standard deviation) (n = 4–6).

Chemical fertilizer1.32 (0.43)1.30 (0.44)0.85 (0.97)
Manure0.51 (0.42)0.85 (0.89)0.35 (0.23)

The NO fluxes showed a seasonal variation that was similar to the seasonal pattern of N2O fluxes, that is, higher in summer and lower in winter, and were also influenced by fertilization (Fig. 2b). The NO fluxes ranged from −1.2 to 91.3 ug NO-N m−2h−1; these values were smaller than the N2O fluxes (−3.6 to 1290.7 ug N2O-N m−2h−1). Large NO fluxes were observed mainly after the application of manure and chemical fertilizer (Fig. 2b). Most of the N2O/NO ratio values were distributed from 1 to 100, and a significant positive correlation was found (P < 0.01) between the N2O/NO ratio and the N2O fluxes (Fig. 3).

Figure 3.

 Relationships between N2O fluxes and the ratio of N2O : NO.

Soil chemical properties

From 2005 to 2007, the soil pH in the chemical fertilizer plot was obviously lower than that in the control and manure plots (Fig. 4a). The application of chemical fertilizer could lead to a drop in soil pH, not only in the chemical fertilizer plot, but also in the manure plot (Fig. 4a). The mean soil pH values from 2005 to 2007 in the control, chemical fertilizer and manure plots were 5.2, 4.6 and 5.1, respectively. The soil pH in the chemical fertilizer plot was significantly lower than that in the manure and control plots (P < 0.001). There was no significant difference in soil pH between the manure and control plots.

Figure 4.

 Seasonal patterns in (a) soil pH, (b) soil NH4+-N, (c) soil NO3N and (d) soil water-soluble organic carbon (DOC) at a depth of 0–5 cm. Data represent mean ± standard deviation (n = 3). Full arrows indicate the dates of the application of chemical fertilizer and the dotted arrows indicate the date of manure application.

The mean NH4+-N concentrations from 2005 to 2007 in the control, chemical fertilizer and manure plots were 4.4, 23.7 and 17.5 mg kg−1, respectively. Soil NH4+-N concentration varied widely in the chemical fertilizer and manure plots (0.4–245 mg kg−1) (Fig. 4b). In contrast, the soil NH4+-N concentration in the control plot was stable, and was always below 12 mg kg−1. The pattern of soil NH4+-N concentration was not influenced by the application of manure only; however, it was influenced by the application of chemical fertilizer (Fig. 4b). Soil NH4+-N concentration in the chemical fertilizer and manure plots increased rapidly right after the application of chemical fertilizer, but then decreased within a few days (Fig. 4b). In 2005 and 2006, peak concentrations of soil NH4+-N were always observed in both the chemical fertilizer and manure plots after the application of chemical fertilizer. However, in 2007, only one small peak was observed in the chemical fertilizer plot following the base fertilizer application in May. After chemical fertilizer application in July 2007, the soil NH4+-N concentration increased very small than last two years (Fig. 4b) may be lead by two reasons: first is the application rate was lower than last two years (Table 1), second is low precipitation and soil moisture with high temperature (Fig. 1) may lead the fertilizer volatilization.

The mean NO3-N concentrations from 2005 to 2007 were 1.4, 2.7 and 2.2 mg kg−1 in the control, chemical fertilizer and manure plots, respectively. The pattern of soil NO3-N concentration was also influenced by the application of N and peaks were observed slightly later than of the peaks for soil NH4+-N concentration (Fig. 4c).

Over the study period the soil DOC concentration ranged from 48 to 121 mg kg−1 in the control plot, from 23 to 116 mg kg−1 in the chemical fertilizer plot, and from 43 to 199 mg kg−1 in the manure plot (Fig. 4d). Mean soil DOC concentrations in the control, chemical fertilizer and manure plots were 73.3, 59.4 and 97.8 mg kg−1, respectively. The soil DOC concentration in the manure plot was significantly higher than that in the control and chemical fertilizer plots (P < 0.01), but the application of chemical fertilizer had no significant influence on the soil DOC concentration compared with the control plot. Three continuous years of manure application significantly increased the soil DOC concentration, which was significantly higher in 2007 than in 2005 and 2006 (P < 0.01).

A Pearson correlation analysis showed that instantaneous N2O flux had a strong positive correlation with soil temperature (P < 0.01), soil NO3-N concentration (P < 0.01) and soil NH4+-N concentration (P < 0.01) (Table 6).

Table 6.   Relationships (Pearson correlation coefficient, r) between instantaneous N2O fluxes and environmental factors using the whole dataset
 N2OSoil T.WFPSpHNO3NH4+DOC
  1. *P < 0.01; **P < 0.05. DOC, water-soluble organic carbon; Soil T., soil temperature; WFPS, water-filled pore space.

N2O1.0 – – – – – –
Soil T.0.330*1.0 – – – – –
WFPS0.020−0.288*1.0 – – – –
pH−0.117−0.275*0.378*1.0 – – –
NO30.307*0.378*−0.061−0.201**1.0 – –
NH4+0.329*0.156**−0.048−0.273*0.380*1.0 –
DOC−0.104−0.173**−0.204**0.341*−0.282*−0.0401.0

Soil denitrifying enzyme activity

Table 7 shows the results for soil DEA. A three-way ANOVA shows that there was a significant difference in soil DEA among the soil layers (root-mat and mineral layers) (P < 0.001) and treatments (with and without NO3 and glucose) (P < 0.001), but there was no significant difference in soil DEA among plots (control, chemical fertilizer and manure) (P = 0.058) (Table 8). However, there was a significant interaction between the soil layer and the treatments (P < 0.001). In the root-mat layer, soil DEA was significantly increased by the addition of NO3-N with (P < 0.001) or without (P < 0.001) the addition of glucose. However, there was no significant effect on soil DEA with the addition of only glucose. There was no significant difference in soil DEA between the treatments Chl+N and Chl+N+C. There was no significant effect of a single addition of NO3-N (Chl+N) or glucose (Chl+C) on soil DEA in the mineral soil. However, a combination of NO3-N and glucose addition (Chl+N+C) increased the soil DEA significantly (P < 0.05). The soil DEA with the addition of both NO3-N and glucose in the root-mat soil was significantly higher than that in the mineral soil (P < 0.001). The soil DEA in the root-mat soil with the addition of NO3-N and both NO3-N and glucose had a significantly positive correlation with soil pH (P < 0.05) (Fig. 5).

Table 7.   Denitrification enzyme activity of the soil samples from the three (control, chemical fertilizer and manure) treatment plots
DateSoil layerTreatmentDEA (mg N2O-N kg−1 h−1)
Control plotChemical fertilizer plotManure plot
  1. Data represent mean (standard deviation) (n = 3). Chl, chloramphenicol (1 g L−1); Chl+C, chloramphenicol (1 g L−1) and organic C (2 g C L−1 as glucose); Chl+N, chloramphenicol (1 g L−1) and NO3-N (200 mg N L−1 as KNO3); Chl+N+C, chloramphenicol (1g L−1), NO3-N (200 mg N L−1 as KNO3) and organic C (2 g C L−1 as glucose); DEA, denitrification enzyme activity.

2007/4/29Root matChl0.04 (0.04)0.04 (0.01)0.04 (0.02)
Chl+N9.99 (1.07)3.66 (0.50)16.30 (2.98)
Chl+C0.06 (0.01)0.07 (0.02)0.14 (0.07)
Chl+N+C10.98 (6.71)3.08 (0.49)13.56 (3.16)
MineralChl2.68 (3.84)0.16 (0.11)0.58 (0.34)
Chl+N0.52 (0.64)0.16 (0.03)1.56 (0.25)
Chl+C0.02 (0.12)0.22 (0.01)0.07 (0.14)
Chl+N+C1.00 (0.20)2.21 (1.50)1.97 (0.36)
2007/6/11Root matChl0.15 (0.02)0.01 (0.00)0.10 (0.05)
Chl+N14.05 (5.03)2.63 (0.68)15.05 (2.28)
Chl+C0.08 (0.02)0.02 (0.01)0.13 (0.03)
Chl+N+C9.47 (0.33)4.66 (0.40)17.35 (0.81)
MineralChl0.09 (0.03)0.07 (0.02)0.71 (0.09)
Chl+N1.43 (0.00)0.35 (0.10)1.24 (0.55)
Chl+C0.03 (0.00)0.04 (0.01)0.05 (0.14)
Chl+N+C2.14 (0.21)0.54 (0.15)1.71 (2.15)
2007/8/20Root matChl0.12 (0.05)0.20 (0.06)0.04 (0.01)
Chl+N27.84 (3.87)9.80 (0.18)13.58 (1.83)
Chl+C0.13 (0.06)0.14 (0.05)0.03 (0.02)
Chl+N+C21.33 (0.23)35.19 (37.29)19.91 (1.73)
MineralChl1.14 (0.31)0.76 (0.37)3.01 (0.20)
Chl+N1.34 (0.17)0.55 (0.09)2.49 (1.83)
Chl+C0.37 (0.03)0.64 (0.33)1.03 (1.22)
Chl+N+C2.92 (0.23)0.83 (0.33)7.65 (0.87)
Table 8.   ANOVA results for soil denitrification enzyme activity
Sourced.f.Mean squareFP-value
Plot294,796,6983.3510.058
Soil layer1991,237,94035.0400.000
Treatment3832,401,85929.4260.000
Plot × soil layer29,030,3520.3190.390
Plot × treatment653,775,7751.9010.174
Soil layer × treatment3451,061,07615.9450.000
Plot × soil layer × treatment525,358,8330.8960.258
Error3728,288,420  
Figure 5.

 Relationship between soil denitrifying enzyme activity (DEA) and soil pH in the root-mat soil.

Discussion

Seasonal pattern in N2O emission

Soil N2O fluxes were significantly higher in the growing season than in the non-growing season. This is attributed to the high soil temperature (Table 6) in the growing season. Granli and Bøckman (1994) found an increased rate of N2O production with an increase in soil temperature up to 20–40°C. High peaks of N2O fluxes were usually observed in both the chemical fertilizer and manure plots within a few weeks after the application of manure or chemical fertilizer in our study (Fig. 2a) as a result of rapid increases in soil NH4+-N and NO3-N concentrations immediately after the application of fertilizer, which decreased within a few days (Fig 4). It has been well established that the rate of N2O emission usually increases with an increase in soil available N (Sehy et al. 2003; Skiba and Smith 2000). Several studies have reported that N2O fluxes significantly increased after the application of N fertilizers. Mu et al. (2008) found that N2O fluxes increased rapidly to higher emission levels in soils cultivated with wheat (from 242 to 433 μg N m−2 h−1) and onion (from 47.2 to 157 μg N m−2 h−1) after N fertilization and that the fluxes lasted for approximately 3 weeks. Schils et al. (2008) also reported that high N2O fluxes occurred in the first week after the application of chemical fertilizer or cattle slurry.

The bacterial processes of nitrification and denitrification are the most important sources of N2O in soil (Granli and Bøckman 1994). According to Davidson (1992) and Skiba et al. (1993), nitrification produces more NO than N2O; conversely, dentrification produces more N2O than NO. The ratio of N2O-N : NO-N is the index of N2O production from nitrification or denitrification (Lipschultz et al. 1981). Lipschultz et al. (1981) reported that the ratio of production of N2O-N : NO-N ranged from 0.2 to 1.0 in nitrification and 100 in denitrification. A significant positive correlation between the N2O fluxes and the ratio of N2O-N : NO-N was found (Fig. 3) in our study (P < 0.01), indicating that the high N2O emissions primarily result from denitrification.

Cumulative N2O emission

Chemical fertilizer and animal waste are the two most important sources of direct N2O emissions from agricultural soils (Mosier et al. 1998). Increasing soil N availability associated with the application of N by chemical fertilizer and manure has greatly increased N2O emissions from agricultural soils (Kroeze et al. 1999). Meng et al. (2005) found that chemical fertilizer and manure contributed to 74–82% of the total N2O emissions. Mori et al. (2008) also reported the N2O emission was predominantly derived from the application of manure and chemical fertilizer N on a volcanic grassland soil in Nasu, Japan. In general, emissions of N2O increase with an increase in N application rates (Granli and Bøckman 1994; MacKenzie et al. 1997). In our study, the N2O emission from applied chemical fertilizer and manure contributed to 77–85% of the total N2O emission in 2005 and 2006. The contribution of chemical fertilizer and manure to N2O emission in 2007 decreased to 47–65% owing to the lower application rates in this year compared with 2005 and 2006. The chemical fertilizer induced EF ranged from 0.85 to 1.32%, which was comparable to the IPCC default value of 1% (Intergovernmental Panel on Climate Change 2006), but was higher than that reported by Akiyama and Tsuruta (2003) from Japanese Andisols amended with chemical fertilizer (ranging from 0.06 to 0.29%). The manure-induced EF of our study ranged from 0.35 to 0.85%, which was significantly lower than the chemical fertilizer induced EF and the IPCC default value, but was close to that reported by Akiyama and Tsuruta (2003) (i.e. 0.55%).

Soil denitrifying enzyme activity

The soil DEA with the addition of NO3-N and glucose in the root-mat soil was significantly higher than that in the mineral soil, indicating that the soil denitrification potential in the root-mat soil was significantly higher than that in the mineral soil. Microbial activities in surface soil are reported to be higher than the activities in deeper soil (Higashida and Takao 1985; Speir et al. 1984). Parkin and Meisinger (1989) reported that total viable bacteria and numbers of denitrifying bacteria were found to decrease exponentially with an increase in soil depth on a well-drained silt loam soil.

Soil DEA in the root-mat soil significantly increased with the addition of NO3-N with (P < 0.001) or without (P < 0.001) the addition of glucose, indicating that the availability of soil NO3-N could be the major limiting factor for soil DEA in our study grassland. In the mineral soil, the addition of NO3-N or glucose alone did not increase the soil DEA, but the addition of NO3-N and glucose together increased the soil DEA, that is, both NO3-N and carbon are limiting factors for soil DEA in mineral soil. The soil DEA in the root-mat soil with the addition of NO3-N and both NO3-N and glucose had a significantly positive correlation with soil pH (P < 0.05; Fig. 5). Soil pH is assumed to be a major variable of soil, controlling the microbial community in general and the community of denitrifiers in particular (Simek and Hopkins 1999). Simek and Hopkins (1999) detected an optimum pH value (7–8) for denitrification in soils. Simek and Cooper (2002) reported that both the overall rates of denitrification under field conditions (i.e. the formation of N2O, N2 and NO and their subsequent emissions) and soil DEA were influenced by soil pH, and that they were reduced in acidic soils compared with neutral or slightly alkaline soils. Ellis et al. (1998) observed in an incubation experiment that the production of N2O decreased with decreasing pH under anaerobic conditions. These results suggest that the highest N2O emission in the manure plot in our study resulted from soil DEA that could have been controlled by the soil pH.

The application of chemical fertilizer significantly decreased the soil pH in the chemical fertilizer plot compared with the control plot. However, the soil pH in the manure plot was not significantly different from that in the control plot, possibly because the higher pH of the manure (8.3–9.1) decreased the effect of the application of chemical fertilizer on the soil pH. Soil acidity is controlled by the amount of H+ and Al3+, which is either contained in or generated by the soil and soil components. According to Kirikae et al. (2001), nitrification is a source of H+ through two nitrification pathways of NH4+ origin and organic N origin. These researchers found that the ratio of H+ to NO3 was two in the pathway of NH4+ origin and one in the pathway of organic N origin. Meanwhile, NO3 uptake by vegetation was a sink of H+. Therefore, organic N has less effect on H+ production than NH4+-N. In contrast, the application of manure increased the cation exchange capacity (CEC) compared with that of chemical fertilizer (Bulluck et al. 2002). Soils with a high CEC have a greater capacity to contain or generate sources of acidity. Several studies (Bulluck et al. 2002; Gil et al. 2008) have shown that soil pH is higher in soils with manure than in soils with chemical fertilizer.

Conclusions

The application of both chemical fertilizer and manure to grassland stimulated annual N2O emission. The chemical fertilizer induced EF (range: 0.85–1.32%) was significantly higher than the manure-induced EF (range: 0.35–0.85%); however, annual N2O emission was significantly higher in the manure plot than in the chemical fertilizer plot. Soil DEA in the NO3-abundant root-mat layer significantly decreased with a decrease in soil pH. Moreover, the application of chemical fertilizer could significantly decrease soil pH, but manure application had no significant effect on soil pH. Therefore, for a fixed quantity of available N, the application of manure could result in higher N2O emission compared with the application of chemical fertilizer owing to the higher soil pH values under manure application than under chemical fertilizer application.

Acknowledgments

We would like to thank the technical staff of the Shizunai Livestock Farm for their help with the field measurements. This study was partly supported by a research grant provided by the project entitled “Establishment of good practices to mitigate greenhouse gas emissions from Japanese grasslands” funded by the Racing and Livestock Association.

Ancillary