| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Landcare Research, Private Bag 3127, Hamilton, New Zealand
* Corresponding author (lbarton{at}agric.uwa.edu.au)
Received for publication December 5, 2000.
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Abbreviations: DFE, dairy farm effluent • WFP, water-filled porosity
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
For national reporting of greenhouse gas emissions, current methods for calculating N2O emissions from agricultural soils are based on anthropogenic N inputs into agricultural systems (Mosier et al., 1998). While the type of inorganic N fertilizer may affect the magnitude of N2O emissions (Eichner, 1990; Clayton et al., 1997; Velthof et al., 1997), the relative importance of organic (e.g., animal slurries) versus inorganic N fertilizers is less well understood as only a few comparative studies have been conducted (Mosier et al., 1996). Where comparisons have been made, fertilizer types have been applied at different N and hydraulic loading rates, making strict comparisons difficult.
Nitrous oxide emissions may also vary between soil types under the same land use. In many countries, peat soils represent a small but significant proportion of their agricultural soils (Clough et al., 1996; Kasimir-Klemedtsson et al., 1997). Nitrous oxide emissions from peat soils generally increase following their development for agriculture, as a result of drainage and N fertilization (Kasimir-Klemedtsson et al., 1997; Augustin et al., 1998; Regina et al., 1998). European studies have shown that N2O emissions from agricultural peat soils are greater than emissions from agricultural mineral soils (Velthof and Oenema, 1995), and these enhanced emissions are most likely due to their higher organic matter and soil water contents.
In New Zealand, irrigation of dairy farm effluent (DFE) onto grazed pastures is increasingly used to treat effluent. Dung and urine deposited on the milking shed floor are washed out and directly irrigated onto nearby pastures, which avoids direct discharge to surface waters and provides an organic source of N fertilizer. Our objectives were to (i) determine if DFE irrigation increased short-term N2O emissions from a surface-drained peat soil and a freely drained mineral soil, and (ii) establish if this increase could be attributed to increased N availability, to increased soil water content, or to a combination of both factors.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Experimental Design
In spring 1999 and autumn 2000, N2O emissions were measured after applying different N treatments to peat and mineral soils. At each soil site, three blocks (1 x 9 m) were established approximately 15 m apart. Each block was evenly divided into nine plots: five treatment plots with a buffer plot between each treatment. Each treatment plot received one of the following: 20 mm of DFE (50 kg N ha-1, DFE treatment); 20 mm of ammonium chloride solution (50 kg N ha-1, +N+H2O treatment); ammonium chloride powder (50 kg N ha-1, +N treatment); 20 mm of water (0 kg N ha-1, + H2O treatment); and no treatment (control). The DFE-nitrogen loading rate chosen was the maximum rate allowed by the local regulatory authority's current ruling, while the hydraulic loading rate was 5 mm less than the maximum rate. On two occasions (mineral soil, spring, after 24 and 48 h), rain was excluded from the experiment to ensure that hydraulic loadings did not vary between experiments. Rainfall was excluded by loosely placing plastic sheeting across the plots, which was removed after the rainfall event. Each soil type was located on a different dairy farm, and examined 2 wk apart. Both sites had a history of DFE irrigation, and were grazed by a dairy herd until 10 d before each trial. Effluent (120 L) used in the trial was collected each season, analyzed for total N by the method described below, and stored for up to 30 d at 4°C before being used.
In Situ Nitrous Oxide and Carbon Dioxide Measurements and Analysis
Nitrous oxide emissions were measured using closed flux chambers (250 mm in diameter x 130 mm in height) inserted 30 mm into the soil (Mosier, 1989). The stainless steel chambers were insulated (rubber foam covered with self-adhesive aluminum foil) to prevent temperature changes in the chamber, and contained two sample ports (20 mm in diameter) that were sealed 30 min after inserting chambers. Chambers were sealed 30 min after insertion to minimize any possible effects of chamber insertion (i.e., soil disturbance) on N2O emissions. A preliminary study showed that chamber temperatures did not vary from air temperature by more than 1°C. Two chambers were inserted in each plot up to 30 min after applying treatments (referred to as 0 h), and re-inserted in the same position at 2 to 4, 24, 48, and 72 h after applying treatments. A preliminary 10-d study showed that N2O emissions peaked within 24 h of applying DFE to both soil types (data not shown). The headspace of the chambers was sampled 0 and 2 h after closing the chambers, and stored in 10-mL evacuated Vacutainers (Becton Dickinson, Franklin Lakes, NJ) for N2O and CO2 analysis. The temperature of the headspace was recorded at 0 and 2 h, and was found not to differ from outside air temperatures by more than 1°C. A preliminary study showed that N2O fluxes increased linearly (r2 > 0.90 for 85% of the covers) for up to 2 h after applying DFE to peat and mineral soils (data not shown). Lower correlation coefficients were often associated with low surface emissions that were difficult to detect.
Nitrous oxide was measured using a Phillip's (Cambridge, England) PU4410 gas chromatograph fitted with an electron capture detector (350°C). Gases were separated using a packed column (HayeSep D; Alltech, Auckland, New Zealand) at 60°C and at an injector port temperature of 160°C. The carrier gas (10% methane in argon) had a flow rate of 32 mL min-1, and was passed though an oxygen and water trap. Carbon dioxide was measured using an infrared gas analyzer (Model LI-6252; Licor, Lincoln, NE).
Soil Analysis
Duplicate, intact soil samples (38 mm in diameter by 75 mm in depth) were collected from the surface of each treatment plot at each sample date, and analyzed separately for soil nitrate, ammonium, and water content. Samples were collected from each treatment plot 1 h after chambers had been sealed. Duplicate cores were analyzed separately and not composited. Soil temperature was measured from the surface 75 mm of effluent and control plots, 1 h after sealing the chambers. Soil water content was determined gravimetrically after drying subsamples at 104°C for 24 h. Soil nitrate and ammonium were determined by adding 100 mL of 2 M potassium chloride to 10 g of field-moist soil and extracting for 1 h. The filtered solution was frozen until analysis for nitrate and ammonium using a modified hydrazine reduction method (Downes, 1978). Particle density values of 2.25 Mg m-3 for the mineral soil and 1.56 Mg m-3 for the peat soil were used to calculate water-filled porosities. These values were obtained from a national database, where particle density was measured using a modified method of Blake and Hartge (1986) using oven-dried soils.
Effluent Analysis
Total N content of 12 effluent subsamples was measured in spring and autumn following digestion of 1.5 mL of effluent with 4.5 mL of 0.165 M sodium persulfate for 30 min at 120°C (Sparling et al., 1996). The digested sample was analyzed for ammonium N and nitrate N as described above. The total N content of the digested effluent sample was calculated by summing ammonium N and nitrate N. The dry matter content of the DFE was determined gravimetrically after drying subsamples for 24 h at 104°C. The pH was measured using a glass electrode, while nitrate and ammonium were determined after filtering DFE (Whatman [Maidstone, UK] No. 42) and analyzing as above. Total C was determined by dry combustion of dried samples using a LECO (St. Joseph, MI) 2000 CNS analyzer.
The total N content of the effluent was 260 mg N L-1 in spring and 280 mg N L-1 in autumn, and did not vary between seasons (P < 0.05). The DFE had a pH of 7.9, dry matter content of 0.4%, total C content of 0.15%, ammonium content of 99 mg N L-1, and nitrate content of 0.1 mg N L-1.
Data Analysis
Nitrous oxide flux rates were calculated from the slope of the temporal change of the concentration within the chamber, and corrected for temperature and chamber volume to surface area ratio (Flessa et al., 1996). Treatment differences in daily N2O fluxes for each soil type and season were analyzed by fitting linear mixed effects models in S-Plus 2000 (MathSoft, 2000). Likelihood ratio tests were used to compare nested models in order to test for treatment, time, and interaction effects. All models included a random effect for sampling unit, and the autocorrelation between repeated measures on the same unit was modeled as an AR1 process (autoregressive process of order 1). Dependent variables were transformed to ensure constant residual variance, either using the natural logarithm or a power transformation (e.g., square root) when there were zeros in the data. Post-hoc pairwise comparisons of treatment means were made using Bonferroni-adjusted P values. Nitrous oxide emissions [ln(N2O + 1)] were related to soil and environmental properties using correlation (Spearman) and stepwise (backwards) multiple regression using SYSTAT (SYSTAT, 1992). Correlation coefficients between independent variables were calculated before conducting multiple regression analyses to ensure multicollinearity did not occur. Water-filled porosity (WFP, %) was calculated by dividing volumetric water content by total porosity and then multiplying by 100 (Linn and Doran, 1984). Volumetric water contents were calculated by multiplying gravimetric water content by bulk density. Total porosity was calculated as 1 - (bulk density/particle density).
|
RESULTS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
By contrast, in the mineral soil during autumn, and in the peat soil in the spring, N2O emissions following DFE irrigation were greater than all other treatments. This result suggests that another component of DFE irrigation, such as its C content, may have enhanced N2O emissions. Increasing soil C availability can enhance denitrification directly, by providing C to denitrifiers, or indirectly, by increasing soil respiration and clogging soil pores, decreasing soil aeration (Christensen, 1985; Groffman et al., 1987; Petersen, 1999), with both direct and indirect effects probably occurring simultaneously. Increased soil C availability following DFE irrigation in comparison with other treatments was evident in our field study as CO2 emissions increased following irrigation in spring at the peat soil, and again in autumn at the mineral soil.
Other researchers have also attributed increased N2O emissions to the C content of animal slurries. Chadwick et al. (2000) found that dairy cow slurry increased soil N2O emissions more than pig slurry. They proposed that greater C addition in the dairy-cow slurry treatment may have enhanced denitrification to a greater extent than the pig slurry treatment, as there were no differences in soil water-filled porosity or mineral N content between treatments. Consequently, by lowering the metabolizable content of an animal slurry, Petersen (1999) was able to decrease N2O emissions from a slurry-amended soil. Furthermore, applying digested slurries to soils produced N2O emissions similar to those from soils fertilized with inorganic N (Petersen, 1999). These findings, plus the findings from our study, suggested that the proportion of N applied to the soil and emitted as N2O may be greater for organic N fertilizers than inorganic N fertilizers, particularly if the organic N fertilizer contains sufficient available C to enhance denitrification.
Measuring nitrous oxide emissions is often labor intensive, and can require intensive sampling techniques over an extended period. Developing models that predict emissions from more easily measurable soil properties provides an alternative to measuring actual emissions. Successfully predicting N2O emissions from DFE-irrigated soils using soil properties and simple regression techniques was difficult in our study. Although we were able to explain a higher proportion (i.e., up to 72%) of our N2O emissions than other researchers comparing N2O emissions from animal slurry–amended soils (Petersen, 1999), regression analysis is not sufficient for predictive modeling. Furthermore, our regression analyses explained <61% of the variation in N2O emissions when combining seasonal data for each soil type, indicating that a single regression equation is unlikely to successfully predict N2O emissions from DFE-irrigated soils throughout the year. Poor correlation between N2O and soil properties is common (e.g., Clayton et al., 1997), and has been attributed to our inability to measure soil properties at the soil microsite scale (Davidson and Hackler, 1994), and the existence of threshold values for soil factors affecting denitrification and nitrification (Clayton et al., 1997). Furthermore, linear regression may be limited if the various effects of carbon availability on N2O emissions occur at the same time. This could cause the relationship between N2O and carbon availability to be nonlinear and possibly synergistic. More sophisticated approaches (e.g., simulation models) should, therefore, be used when modeling N2O emissions from soil. It is useful to note that incorporating CO2 emissions in our regression analyses improved the correlation between N2O emissions and independent variables, emphasizing the importance of understanding the relationship between soil C dynamics and N2O emissions.
Our findings indicate that DFE irrigation is a source of N2O from New Zealand dairying soils, and that this is due to increases in soil nitrogen, water content, and available carbon. Two approaches for decreasing N2O emissions from DFE-irrigated soils include lowering the nitrogen and/or carbon content of the applied effluent. A variety of strategies for lowering the N content of stock urine and dung have been proposed (Castillo et al., 2000), but the effect of these strategies on N2O emissions from soils treated with dairy effluents has not been widely investigated. Chadwick (1997) reported lower denitrification rates from pig-manured soils after decreasing the N content of the manure by lowering the crude protein content of the stock feed by 25%. However, N2O emissions were not significantly altered. Lowering the C content of animal slurries has also decreased N2O emissions in a limited number of studies (Petersen, 1999). For example, using current IPCC guidelines, Petersen (1999) estimated that anaerobically digesting land-applied slurry could potentially lower N2O emissions from Danish agriculture by 1.2 to 2.5%. Other approaches to lowering N2O emissions from DFE-irrigated soils include decreasing DFE nitrogen loading to soil, matching DFE nitrogen application rates to crop demand (Chadwick, 1997), and incorporating nitrification inhibitors, which suppress the formation of NO3 from NH4, with applied DFE. However, while nitrification inhibitors have successfully lowered N2O emissions from soils fertilized with NH4–based fertilizers in the Northern Hemisphere (Mosier et al., 1996; McTaggart et al., 1997), the effectiveness of nitrification inhibitors may be less in northern areas of New Zealand, where warm soil temperatures may cause inhibitors to decompose rapidly (Williamson et al., 1996).
In conclusion, the application of DFE to two contrasting soil types greatly enhanced N2O emissions, and at times more than the application of inorganic N at the same N and hydraulic loading rate. Increased N2O emissions following DFE were attributed to increased soil N and soil water contents in two instances, but also to increased C availability in another two cases. These findings suggest that the proportion of N applied to the soil and emitted as N2O may at times be greater for organic N fertilizers than inorganic N fertilizers, particularly if the organic N fertilizer contains sufficient available C to enhance denitrification.
|
ACKNOWLEDGMENTS |
|---|
-Vukovi, and Annie Lloyd-Jones are thanked for technical support in the field and laboratory. Comments made by three anonymous reviewers improved the manuscript. This research was funded by the New Zealand Foundation for Research, Science, and Technology C09802. |
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  T. J. Clough and F. M. Kelliher Dairy Farm Effluent Effects on Urine Patch Nitrous Oxide and Carbon Dioxide Emissions J. Environ. Qual., May 11, 2005; 34(3): 979 - 986. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Ginting and B. Eghball Nitrous Oxide Emission from No-Till Irrigated Corn: Temporal Fluctuation and Wheel Traffic Effects Soil Sci. Soc. Am. J., May 6, 2005; 69(3): 915 - 924. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. J. Calderon, G. W. McCarty, J. A. S. Van Kessel, and J. B. Reeves III Carbon and Nitrogen Dynamics During Incubation of Manured Soil Soil Sci. Soc. Am. J., September 1, 2004; 68(5): 1592 - 1599. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. R. Sharpe, H. H. Schomberg, L. A. Harper, D. M. Endale, M. B. Jenkins, and A. J. Franzluebbers Ammonia Volatilization from Surface-Applied Poultry Litter under Conservation Tillage Management Practices J. Environ. Qual., July 1, 2004; 33(4): 1183 - 1188. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Vadose Zone Journal | Journal of Plant Registrations | ||||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal |
