JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (4)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Petersen, S. O.
Right arrow Articles by Ambus, P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Petersen, S. O.
Right arrow Articles by Ambus, P.
Agricola
Right arrow Articles by Petersen, S. O.
Right arrow Articles by Ambus, P.
Related Collections
Right arrow Animal Waste
Right arrow Nutrients
Right arrow Spatial Variability
Right arrow Infiltration
Right arrow Isotopes
Published in J. Environ. Qual. 32:2399-2409 (2003).
© ASA, CSSA, SSSA
677 S. Segoe Rd., Madison, WI 53711 USA

TECHNICAL REPORTS

Waste Management

Redistribution of Slurry Components as Influenced by Injection Method, Soil, and Slurry Properties

Søren O. Petersen*,a, Henrik H. Nissenb, Ivar Lundc and Per Ambusd

a Dep. of Agroecology, Danish Inst. of Agric. Sci., P.O. Box 50, DK-8830 Tjele
b Aalborg Univ., Dep. of Civil Engineering, Sohngaardsholmsvej 57, DK-9000 Aalborg
c Dep. Automation and System Engineering, Danish Inst. of Agric. Sci., P.O. Box 536, DK-8700 Horsens
d Risø National Lab., Dep. of Plant Res., P.O. Box 49, DK-4000 Roskilde, Denmark

* Corresponding author (soren.o.petersen{at}agrsci.dk).

Received for publication August 21, 2002.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The distribution of moisture, degradable C, and N after direct injection of slurry can affect the turnover and plant availability of slurry N. This study examined effects of injection method, soil conditions, and slurry properties on the infiltration of several slurry components under practical conditions. The water retention capacity of 22 pig and cattle slurries was quantified by dialysis at -0.016, -0.047, and -0.100 MPa. All slurries followed the relationship: relative water loss = 1/(1 + aVS[volatile solids]), indicating that retention of liquids in the slurry injection zone can be predicted from slurry VS and soil water potential. Two-disc injection and harrow tine injection were simulated (no slurry applied) in five trials. Two trials indicated that disc injection resulted in higher permeability compared with harrow tine injection. In a separate experiment, soil moisture and dissolved ions were monitored in and around injection slits amended with pig or cattle slurry. Moisture gradients, which were recorded with small printed-circuit-board (PCB) time-domain-reflectometry (TDR) probes, were temporally stable and reestablished following rainfall. Slit sections with pig and cattle slurry containing 13C-acetate and 15N-ammonium showed a shift in the 13C to 15N ratio of the injection zone within 24 h, which was explained by removal of dissolved C and/or retention of NH+4. Cattle slurry was more concentrated around the injection slit than pig slurry, and greater contact between slurry and soil was obtained with harrow tine injection. The heterogeneity of slurry C and N distribution after direct injection should be accounted for in models describing slurry N turnover.

Abbreviations: {theta}, volumetric soil moisture • CEC, cation exchange capacity • ECa, apparent electrical conductivity • K, relative dielectric permittivity • PCB, printed circuit board • PEG, polyethylene glycol • TDR, time domain reflectometry • VS, volatile solids


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
WITHIN AGRICULTURE, field application of slurry (liquid manure) by direct injection is widely used. The technique can minimize NH3 volatilization, thereby increasing N use efficiency (Sommer and Hutchings, 2001) while also protecting natural ecosystems against excessive N loads via deposition of airborne NH3 (Pearson and Stewart, 1993). Furthermore, this application method, when properly performed, enables slurry amendment to established crops and grasslands with minimum risk of leaf damage or dispersal of pathogens (Chen and Tessier, 2001).

In the past, studies of slurry infiltration have mostly been motivated by the need to prevent ground water pollution from slurry lagoons or feedlots, or from fields irrigated with diluted slurry, and the focus has therefore been on steady state infiltration rates (DeTar, 1979; Parker et al., 1999; McCullough et al., 2001). Application of slurry as an organic fertilizer for crop production represents a different situation, where mass flow is dominated by matric potential gradients that will cease after a few hours. Although the initial placement of directly injected slurry is relatively well defined, the subsequent redistribution can affect a much larger soil volume (Comfort et al., 1988). Laboratory studies have indicated that the equilibrium distribution of slurry liquids is a function of both slurry organic matter content and soil water potential (Petersen and Andersen, 1996; Olesen et al., 1997a,b). Dissolved compounds and suspended particles will be carried along with the aqueous phase, but slurry components may interact with the soil during this transport. Ammonium ions can adsorb to negatively charged surfaces, metabolizable C can be taken up by soil microorganisms, and suspended micrometer-scale particles in the slurry may become trapped in the soil matrix. The importance of such mechanisms for the C and N turnover of injected slurry will depend on both slurry and soil properties.

The redistribution of slurry liquids also depends on the physical properties of the injection slit and, hence, the injection method. Two basically different methods are (i) disc injection, which gives a sideward displacement of the soil and typically results in a well-defined and smooth-walled slit; and (ii) harrow tine injection, where pressure on the soil is concentrated at the bottom of the injection slit and soil structure around the injection slit is broken up. The potential for compaction or smearing of soil around the injection slit is likely to depend on soil texture and soil moisture at the time of application (Baker et al., 1996).

This study aimed to characterize the injection slit environment and the infiltration of several slurry components under practical conditions. We hypothesized that both slurry organic matter content and injection method would influence the distribution of slurry components, and these questions were addressed in experiments that involved two contrasting slurries, two injection methods, and two soil types. Existing methods were adapted to quantify slurry water retention capacity, injection slit permeability, and water redistribution while slurry samples enriched with stable isotopes were used to track the redistribution of dissolved C and N.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Soils and Slurries
Two different sites were used in this study, henceforth referred to as Bygholm and Foulum (Fig. 1) . Geographical locations of these sites in the western part of Denmark were 55°52' N, 9°49' E (Bygholm) and 56°29' N, 9°34' E (Foulum). The Bygholm soil was a Glossic Phaeozem with a mineral composition of 72% sand, 15% silt, and 14% clay. It contained 15 g kg-1 total C and 1.4 g kg-1 total N. The pHKCl was 5.8 and the cation exchange capacity (CEC) was 12.3 cmol kg-1. The Foulum soil was a Typic Hapludult with a mineral composition of 79% sand, 13% silt, and 8% clay. It contained 19 g kg-1 total C and 1.8 g kg-1 total N. The pHKCl was 5.9 and the CEC was 13.1 cmol kg-1. The experiments were performed on recently tilled fallow soil that had been used for growth of small-grain crops the previous year.



View larger version (27K):
[in this window]
[in a new window]
  		Fig. 1. A map of Denmark in northern Europe, showing the locations of the two experimental sites at 55°52' N, 9°49' E (Bygholm) and 56°29' N, 9°34' E (Foulum).

 
To quantify the relationship between slurry organic matter content and water retention, a total of 22 different slurries were collected at farms near the Foulum site in February 2000. The slurries were sampled from dairy cattle, beef cattle, piglet, and slaughter pig production facilities. Slurry was collected from either a prestorage tank or the main storage tank, and the slurries thus varied in age between a few days and several months. Four of the slurries had been anaerobically digested before storage. Volatile solids were quantified as a measure of organic matter content, while two selected slurries to be used for a field experiment (see below) were further characterized (Table 1).


View this table:
[in this window]
[in a new window]
  		Table 1. Selected characteristics of the two slurries used in the field experiment.

 
Slurry Water Retention Characteristics
Water retention capacity of the 22 slurries was quantified at three selected water potentials by osmosis (Zur, 1966). Lengths (15 cm) of dialysis membrane (Spectra/Por 3, Spectrum Lab., Rancho Dominguez, CA) with a molecular weight cutoff value of 3500 Da were soaked in demineralized water for 30 min, wiped dry with tissue, and closed at one end with a plastic clip. Slurry was then added to the membrane via a funnel and the membrane closed with a second clip. Duplicate samples of the 22 slurries were prepared for each water potential.

The samples were incubated in solutions (0.2 L per sample) of polyethylene glycol (PEG) (Fluka, Copenhagen) containing 30, 60, and 90 g PEG L-1, corresponding to osmotic potentials of -0.016, -0.047, and -0.100 MPa (Waldron and Manbeian, 1970). The PEG molecular weight was 5000 to 7000 Da. Since the extraction of slurry water would dilute the PEG solutions and thus change their osmotic potential, the samples were transferred to fresh solutions after 48 h. The incubation was continued for a total of 7 d, at which time the slurry samples were at equilibrium with the PEG solutions, as determined in a preliminary experiment (Fig. 2) . At the end of incubation, samples were cut open and the slurry residue was sampled for determination of water content; it was assumed that VS was fully retained by the membrane and was not degraded during the incubation period.



View larger version (14K):
[in this window]
[in a new window]
  		Fig. 2. An example of the time course of water loss from slurry due to a water potential gradient. The slurry was enclosed in a dialysis membrane and submerged in a solution of polyethylene glycol (see text).

 
Permeability of Injection Slits
The effect of injection method on infiltration capacity was assessed by measurements of air permeability (Roseberg and McCoy, 1990). Air permeability was measured with soil cores taken in stainless steel cylinders (diam., 32 mm; length, 50 mm; wall thickness, 0.5 mm). A cylinder was mounted in a test rig (Fig. 3A) , and then a pressure gradient was established across the soil sample which induced a mass flow through the soil column. The flow through a porous medium can be described by Darcy's law:

[1]
where q is flow rate (m3 s-1) and ka is the permeability (m2) for the specific material, {eta} is the dynamic viscosity of the mobile phase (kg m-1 s-1), p is the pressure (kg m-1 s-2), and x is the distance (m). Equation [1] is only valid under isothermal conditions and at laminar flow, but applies to both liquid and air flow. Hence, air permeability measurements should also reflect the potential for infiltration of slurry liquids.



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 3. The experimental setup used for quantifying the air permeability of soil cores sampled from the wall of simulated injection slits. A constant head pressure of 1 kPa was used for the field measurements.

 
The air permeability was calculated by integrating Eq. [1] (Kirkham, 1947):

[2]

Here, Q is the volumetric flow rate (m3 s-1), {Delta}P is the pressure gradient across the sample (kg m-1 s-2), as is the cross-sectional area of the sample (m2), and Ls is the length of the soil sample (m).

A standard method for determination of air permeability was downscaled to accommodate the experimental situation, and so it was first tested under laboratory conditions with soil from the Bygholm site. The soil water content was adjusted to 135, 164, or 186 g kg-1 by partial drying at 50°C, and then the soil was packed in boxes of 16 x 26 x 14 cm (l x w x h) and compacted with uniform pressures of 6, 12, or 18 kN m-2. Eleven soil cores were taken in a fixed pattern (three rows with four, three, and four samples) for measurements of air permeability, which were performed at head pressures ranging from 0.5 to 4 kPa.

In the field, direct injection was simulated by mounting pairs of either two-disc or harrow tine injection units on a tractor with a three-point linkage and then injecting to a depth of 6 to 7 cm, but without slurry application during this operation. Randomly selected slit sections were excavated so as to leave one side of the injection slit undisturbed, and soil cores were then taken in the bottom part of the opposite slit wall (Fig. 3B). Measurements were performed within 4 h of injection. Five different trials with simulated injections were conducted, two at the Foulum site (Trials 1,2) and three at the Bygholm site (Trials 3–5). In one of the trials at Bygholm, the interaction with soil moisture was examined by simulating 30-mm rainfall before injection. The air permeability data reported from field trials were obtained at a head pressure of 1 kPa, and the bulk density of soil cores was subsequently determined.

Distribution of Slurry Components
The last part of this study was a field experiment which examined (i) the distribution of the slurry liquid phase and dissolved ions and (ii) the distribution of 15N-labeled ammonium and 13C-labeled acetate (CH3COO-) as a function of slurry type and injection method. The experiment was based on simulated direct injection, as described above, followed by manual slurry application to sections of the injection slits as specified below.

The experiment was initiated at the Foulum site on 22 June 2000. Lanes, each consisting of two slits, were prepared by simulated full-scale disc injection and harrow tine injection as described above. In each slit, four randomly selected sections were delimited by insertion of steel plates to about the 10-cm soil depth. These four sections per slit were amended with untreated cattle slurry, untreated pig slurry, labeled cattle slurry, or labeled pig slurry within 1 h of the simulated injection. The application rates of cattle and pig slurry corresponded to 50 and 30 Mg ha-1, respectively, and the slits were covered with soil following slurry application to minimize NH3 losses. The two slits of each lane represented the replicates of each treatment, and the order of treatments within a slit was fixed because of practical limitations; hence, the experimental design was not completely randomized.

Approximately 24 h after slurry application, measurements of volumetric water content ({theta}) and apparent soil electrical conductivity (ECa) were initiated in and around injection slits with untreated slurry. All four combinations of injection method and slurry type were monitored and each combination duplicated. Small metal-coated PCB-TDR probes (Nissen et al., 2003; Fig. 4A) were installed in a vertical cross-section of each slit. In contrast to conventional straight TDR probes, the PCB-TDR probe features a carrier material (the circuit board) on which a serpentine waveguide (probe rods) is positioned. This reduces the outer dimensions of the probe significantly while maintaining a sufficient waveguide length to obtain precise near point estimates of volumetric soil moisture, {theta}. The probes were positioned in an array of four probes per slit, thus enabling measurements of {theta} and ECa in both horizontal and vertical direction relative to the bottom of the injection slit (Fig. 4B). Probe positions were located with a plastic template, and probe slits were made with an undersized (relative to the PCB-TDR probes) metal poking instrument to prevent probe damage during installation. The TDR probes were connected to a fully automated system including a Tektronix (Beaverton, OR) 1502 B cable tester, a laptop computer, and two multiplexers and TACQ software (Vadoze Zone; Houston, TX). Upon probe installation, the soil profiles were covered with insulating material and soil. Time domain reflectometry traces were acquired every half hour during a 2-wk period and stored for later processing.



View larger version (12K):
[in this window]
[in a new window]
  		Fig. 4. Soil moisture and apparent electrical conductivity in and around the injection slit was measured with a small printed circuit-board time-domain-reflectometry probe. (A) Design of the probe, with dimensions in mm. (B) Positions A–D of the four probes installed in each profile; probes were positioned relative to the bottom of the injection slit.

 
The presence of the circuit board within the sample volume of the PCB-TDR probe will influence the measured relative dielectric permittivity (K). For determination of {theta}, this was compensated for by creating a relationship between K measured by the PCB-TDR probe (includes contributions from both the medium surrounding the probe and the circuit board) and K of the medium surrounding the probe. The relationship was established in a calibration experiment with immersion of the PCB-TDR probes in air and fluids having well-known K values; the calibration procedure is described in detail by Nissen et al. (2002). In the present study, we used an empirical {theta}–K relationship for the Foulum soil (Jacobsen and Schjønning, 1993). Apparent soil EC determination with the PCB-TDR probes followed the procedure used for conventional TDR probes (Nadler et al., 1991). Further, the probe cell constant, Kp, was determined for each probe before they were used in the field.

To investigate the redistribution and possible retention of specific slurry components, portions of the two slurries had been spiked with 13CH3COONa (Cambridge Isotope Laboratories, Andover, MA) at a rate of 0.8 g kg-1 slurry, with NaBr (included as an inert reference) at a rate of 1 g kg-1 slurry, and with (15NH4)2SO4 (Icon Services, Summit, NJ) at rates of 0.95 and 0.45 g kg-1 for pig and cattle slurry, respectively. Slurry portions with tracers were shaken for 16 h at 2°C for equilibration. Injection slit sections with labeled slurry were sampled after about 24 h. Blocks of soil, 14 x 4 x 8 cm (l x w x h), were taken with a rectangular steel corer and sectioned as indicated in Fig. 5 ; corresponding sections were pooled. Soil unaffected by slurry amendment was also sampled for tracer analyses. Field-moist soil samples were subsampled for determination of Br- and soil moisture, and the remainder of each sample was then split in two parts for isotope analyses. To counteract volatilization of NH3 and CH3COOH during air-drying, soil for 15N analysis was wetted with NaH2PO4 (0.5 M, pH 4.3), and soil for 13C analysis with Na2HPO4 (0.5 M, pH 8.8).



View larger version (15K):
[in this window]
[in a new window]
  		Fig. 5. Injection slits amended with labeled pig and cattle slurry were sampled after 24 h with a rectangular corer (14 x 4 x 8 cm, l x w x h). The blocks were sectioned as indicated.

 
Analytical Techniques
Dry matter was determined after drying at 105°C for 24 h, and VS after an additional 3 h at 450°C (Broadbent, 1965). Dissolved organic C was determined with persulfate digestion of diluted slurry after filtration (<0.45 µm) (Paul et al., 1999).

Soil texture was determined after dispersion with sodium pyrophosphate. Clay and silt was quantified by sedimentation in a hydrometer, and sand fractions by wet-sieving (Gee and Bauder, 1986). The size distribution of slurry particles was measured with a water-jet sieving device (Møller et al., 2002). Manure samples (0.5–1.0 kg) were added on top of six nested sieves of decreasing mesh size (1000, 500, 250, 100, 50, and 25 µm). A spraying arm with 34 nozzles rotated over each sieve, and the liquid was collected in a tray below the 25-µm sieve. The amount of slurry dry matter in each fraction was determined after drying (48 h at 65°C).

Bromide was determined with a Metrohm (Herisau, Switzerland) 690 ion chromatograph. Inorganic N was quantified by standard colorimetric methods (Keeney and Nelson, 1982). For isotope analyses, dried soil samples were ball-milled, and subsamples of 10 to 40 mg packed in tin capsules. Total contents of C and N, as well as 13C/12C and 15N/14N isotopic ratios, were measured on a Carlo Erba (Milan, Italy) EA 1110 elemental analyzer coupled in continuous flow mode to a Finnigan MAT Delta PLUS isotope ratio mass spectrometer (Thermo-Finnigan, Kungens Kurva, Sweden).

Statistical Analyses
Effects of injection method on the air permeability of injection slits were tested for each soil type by a generalized linear model followed by Tukey's HSD multiple comparisons test. Similar tests were performed for concentrations of Br-, 13C, and 15N around slurry-amended injection slits, although the distribution of treatments was not fully randomized. Each of the eight sections were analyzed separately.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Slurry Water Retention
The water retention of 22 slurries with VS contents ranging from 3 to 93 g kg-1 was quantified at three different water potentials with a method originally developed for soils (Zur, 1966). The particular advantage of using solutions with defined osmotic potentials for adjustment of slurry water potential was the fact that equilibration occurred under submerged conditions, where the limited oxygen diffusion (104 times slower than in air) and high oxygen demand of the slurry (Petersen et al., 1996) would have ensured largely anaerobic conditions, thereby restricting the degradation of particulate organic matter during the 7-d equilibration period.

Figure 6 shows slurry water retention as a function of VS content at three different water potentials. At -0.047 and -0.100 MPa, cattle and pig slurry clearly followed the same relationship between water retention and VS, whereas the results at -0.016 MPa were more scattered. The four anaerobically digested slurries followed the same relationships as untreated slurry. Hence, water retention characteristics of slurry VS appeared to be independent of slurry origin or treatment before storage, a finding which could facilitate prediction of slurry infiltration characteristics on field application. At all three water potentials the relationship between VS and water retention could be expressed as:

[3]
where the parameter a is a function of water potential. Values of a (shown in Fig. 6) were determined by curve fitting, and disregarding the two outliers (both pig slurry) at -0.016 MPa.



View larger version (22K):
[in this window]
[in a new window]
  		Fig. 6. The water loss from 22 cattle (white), pig (black), and anaerobically digested (gray) slurries as a function of slurry organic matter content for three selected water potentials. Water was extracted by dialysis with solutions of polyethylene glycol (30, 60, and 90 g L-1). Except for two outliers (square symbols), all measurements accommodated the relationship relative water loss = 1/(1 + aVS); VS = volatile solids, values of a (a function of water potential) (± SE) are inserted in the three plots.

 
Equation [3] provides an estimate of the fraction of slurry liquids that will become absorbed by the soil matrix at equilibrium. In practice, dissolved and suspended organic matter will follow the liquid into the soil, and the final distribution of VS and liquids must be determined by an iterative procedure. Transformations of C and N in slurry liquids absorbed by the soil will take place in a well-aerated environment, whereas C and N in liquid retained in the injection slit will be metabolized in an environment that is dominated by slurry with higher moisture content and biological activity, and therefore lower oxygen status, compared with the bulk soil (Rice et al., 1989; Petersen et al., 1996; Clemens and Huschka, 2001). It is likely that the description of N transformations and environmental losses after slurry injection could be improved if the inherent heterogeneity and strong concentration gradients of this system were taken into account. For example, Olesen et al. (1997a) studied N transformations around a slurry layer and found that neglecting diffusion of ammonium from the slurry layer changed net nitrification rates at 1-cm distance from the soil–slurry interface by up to 50% depending on slurry type and soil moisture.

Permeability of Injection Slits
The infiltration rate of injected slurry will depend on the potential for mass flow, or permeability, across the injection slit wall. A modified method for measuring air permeability was tested under controlled conditions, and Fig. 7 shows, for a water content of 186 g kg-1, air flow as a function of head pressure at the three levels of compaction employed. The curvilinear part of each curve reflects turbulent flow, and a head pressure of 1 kPa was therefore chosen for subsequent measurements. In the test, air permeability ranged from 56 to 266 µm2 and showed the expected decrease in permeability with increasing moisture and compaction level (data not shown).



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 7. Relationships between air flow and head pressure for soil (gravimetric soil moisture, 18.6%) compacted by pressures of 6, 12, and 18 kN m-2 (n = 6). The curvilinearity was caused by turbulent flow.

 
Five field trials with simulated disc injection and harrow tine injection were conducted to evaluate the effect of injection method and soil conditions, as reflected in air permeability measurements. A variety of soil physical methods have shown compaction at the bottom of slits made by disc coulters for direct drilling, while no compaction was found with harrow tine (chisel) coulters (Baker and Mai, 1982; Iqbal et al., 1998; Munkholm et al., 2003). Therefore, different effects of disc injection and harrow tine injection on soil permeability were also expected. However, the results presented in Table 2 were complex and not readily interpreted. In two trials (Trials 3 and 4), the air permeability was significantly greater after disc injection than after harrow tine injection, while the three other trials showed no difference between methods. Both methods showed reduced permeability compared with the undisturbed control soil in Trial 1, as expected, while in Trials 2 and 5 there was no significant difference between control soil and injection slits. No relationship between soil moisture and air permeability could be identified, and a major effect of injection method on injection slit wall permeability could thus not be shown for the soil conditions examined. Bulk density of the soil cores ranged from 1.3 to 1.7 g cm-3 with no treatment effects (data not shown), suggesting that soil structure varied considerably. Permeability is greatly influenced by macropores and in particular by pore continuity (Ball et al., 1997), and such factors may have dominated the levels of air permeability recorded.


View this table:
[in this window]
[in a new window]
  		Table 2. Air permeability across injection slit walls was quantified following simulated disc injection or harrow tine injection. Five different trials were conducted on two soil types as indicated (n ranged from 6 to 20).

 
Distribution of Slurry Liquids
The cattle slurry and pig slurry selected for the field experiment differed both quantitatively and qualitatively (Table 1). The greater dry matter content of cattle slurry is typical for animal production in Denmark, as is the high proportion of total N as NH+4 in pig slurry. With respect to particle size distribution, cattle slurry had a higher proportion of large particles (>1000 µm) than pig slurry, whereas pig slurry had more particles in the 500- to 1000-µm fraction. However, in both slurries most of the dry matter was in the 0- to 25-µm fraction.

About 1 d after simulated injection and manual application of slurry, PCB-TDR probes were installed in and around cross sections of injection slits, as shown in Fig. 4B, for monitoring of {theta} and ECa. There was considerable variation between replicate injection slits in the absolute responses observed, but the temporal dynamics were consistent. Possible reasons for the variability was fluctuating slit width leading to varying amounts of slurry in the 5-cm length intervals covered by the probes, or incomplete contact between probe and soil or slurry.

Figures 8 and 9 show a selected data set for each injection method with pig slurry and cattle slurry, respectively. Actual readings for the outmost sampling position (Position C in Fig. 4B) are shown in the left-hand column, and relative deviations at the other probe positions are shown in the middle ({theta}) and right-hand column (ECa). In the 1- to 9-d period after probe installation, {theta} and ECa fluctuated around a constant level in all treatments, though with some difference in the amplitudes. Rainfall (6.6 mm) during Day 9 temporarily increased {theta}, and also significantly increased ECa.



View larger version (31K):
[in this window]
[in a new window]
  		Fig. 8. Volumetric soil moisture ({Theta}) and apparent electrical conductivity (ECa) were measured in and around injection slits between 1 and 13 d after application of pig slurry with disc injection (upper panels) or harrow tine injection (bottom panels). Actual readings are shown in the left-hand column, while the plots in the second ({Theta}) and third columns (ECa) show relative deviations from Position C, the probe installed at a 5-cm distance from the injection slit (see Fig. 4B). The variables x and z refer to, respectively, horizontal and vertical distance relative to the injection slit.

 


View larger version (32K):
[in this window]
[in a new window]
  		Fig. 9. Volumetric soil moisture ({Theta}) and apparent electrical conductivity (ECa) were measured in and around injection slits between 1 and 13 d after application of cattle slurry with disc injection (upper panels) or harrow tine injection (bottom panels). Actual readings are shown in the left-hand column, while the plots in the second ({Theta}) and third column (ECa) show relative deviations from Position C, the probe installed at 5-cm distance from the injection slit (see Fig. 4B). The variables x and z refer to, respectively, horizontal and vertical distance relative to the injection slit.

 
For pig slurry, moisture was more concentrated around the injection slit (difference between Probes A and C > difference between Probes B and C) with disc injection than with harrow tine injection (Fig. 8, top panel), and the moisture level below the slit was higher. Apparent electrical conductivity was highest in the injection slit, but values converged during the 14-d period, indicating diffusion of charged compounds away from the injection slit. With harrow tine injection of pig slurry there appeared to have been a considerable horizontal distribution of slurry liquids during the first 24 h, or initial mixing of slurry into a large soil volume, as indicated by similar ECa levels in all positions (Fig. 8, bottom panel). Here, the water content registered in Position A was lower than in the reference position (C), possibly because of incomplete contact between probe and soil as suggested above.

For cattle slurry, moisture gradients were less pronounced than for pig slurry (Fig. 9). This was unexpected in view of the higher organic matter content and thus water retention capacity of cattle slurry, and in contrast to the profiles of ECa and chemical parameters (see next section). Cellulosic materials have a complex structure with an internal surface area of 300 m2 g-1 (Cowling and Brown, 1969), and water in these voids could be shielded from the TDR trace. The abrupt increase of soil moisture in the treatment with disc-injected cattle slurry after 6 d could not be explained, but probably represented a change in the shape of the TDR waveforms influencing the travel time determination for probe C, as indicated by the sudden change relative to the other probes (middle plot). As with pig slurry, the horizontal distribution of ECa showed stronger gradients after disc injection than after harrow tine injection of cattle slurry (Fig. 9).

The most striking feature of the data presented in Fig. 8 and 9 is the fact that differences in {theta} between probe positions were maintained across time and reestablished after a rainfall event. It implies that soil conditions around the injection slit were modified on a more permanent basis. This may reflect a collapse of soil structure due to the mass flow of water at the time of slurry injection (Or, 1996), or slurry particles could have been transported into the soil matrix, contributing to an increased water retention capacity (Olesen et al., 1997b).

Distribution of Dissolved Compounds
Concomitantly with the application of untreated slurry, other slit sections were amended with labeled slurry to trace the distribution of dissolved compounds. These treatments were sampled after about 24 h and tracer concentrations in the eight sections of each sample (Fig. 5) quantified.

The inert Br- ion (Table 3) was expected to follow the liquid phase of the slurry and thus to supplement the observations made by TDR. There were strong gradients of Br- with distance from the injection slit, reflecting the predominance of slurry in and around the injection slit. Concentrations of Br- near the injection slit were much higher for cattle slurry than for pig slurry, indicating that water retention in the injection slit was greater with cattle slurry, which also had the highest VS content (Table 1). Furthermore, Br- showed a greater horizontal distribution with harrow tine injection, in accordance with the TDR data, although the difference was only significant for pig slurry, Section 5 (Table 3). Bromide concentrations showed no treatment effects with respect to the downward movement of liquid (Sections 6–8), in contrast to the TDR measurements. This may be because Br- concentrations represented a greater depth interval (4- to 8-cm depth) than TDR measurements (approximately 5- to 6-cm depth).


View this table:
[in this window]
[in a new window]
  		Table 3. Bromide concentrations and concentrations of 13C and 15N in each of the eight sections isolated after application of pig and cattle slurry to slits prepared by simulated disc injection or harrow tine injection.

 
Concentrations of 13C and 15N (atom % excess) also indicated a greater retention of slurry liquids with cattle slurry, and stronger gradients with disc as opposed to harrow tine injection (Table 3). Concentrations of 15N in the injection slits (Section 1) with pig slurry were higher relative to cattle slurry when compared with Br- and 13C, since more 15NH+4 had been added to pig slurry than to cattle slurry. The distribution of tracers thus confirmed the establishment of slurry liquid gradients around the injection slit, as well as treatment effects, but any differentiation in the distribution of slurry C and N was not evident from these data.

To reduce the impact of soil heterogeneity, ratios of atom % 13C to atom % 15N were considered. The 13C to 15N ratios of tracer-amended pig and cattle slurry were calculated to be 3.1 and 2.0 assuming 40% C in slurry organic matter, while that of unamended soil was measured to be 2.7. Figure 10 shows 13C to 15N ratios of soil from all treatments and sections analyzed. Surprisingly, the 13C to 15N ratios of Sections 1 and 2 (all treatments) and Section 3 (pig slurry only) were around 1.0 and thus lower than both slurries and soil. In Fig. 10, 13C to 15N ratios of the sections most distant from the injection slit (Sections 5 and 8), were close to the value for undisturbed soil with all treatments, indicating that little slurry had reached these parts of the soil profile. One exception was pig slurry applied with harrow tine injection, which had a 13C to 15N ratio < 2 in Section 5; this confirmed the other evidence for large horizontal distribution of pig slurry after harrow tine injection. Also, pig slurry gave 13C to 15N ratios around 2.0 in Sections 6 and 7 (below the injection zone), suggesting that the dilute pig slurry was more likely to move vertically through the soil than cattle slurry, as also indicated by the TDR data.



View larger version (46K):
[in this window]
[in a new window]
  		Fig. 10. Ratios of atom % 13N to atom % 15N in each of the eight sections sampled in the field experiment with application of labeled slurry to slits prepared by simulated injection.

 
The observation of 13C to 15N ratios below the level achievable by simple mixing of components implies that the composition of slurry in and around the injection zone had changed with respect to C and N during the first 24 h after application. The following mechanisms could have contributed to a decrease in 13C relative to 15N: (i) the infiltration process itself, (ii) acetate degradation, and (iii) retention of NH+4 by negatively charged soil particles. Infiltration, by removing the same proportion of dissolved C and N, would lower the 13C:12C ratio of the injection zone more than the 15N:14N ratio, resulting in a lower 13C to 15N ratio. The reason is that the fraction of slurry C in soluble form (compare VS and dissolved organic C in Table 1) was lower than the fraction of slurry N present as (soluble) NH+4. In contrast, the labeled components added, 13CH3COOH and 15NH+4, were both 100% soluble. The effect of simple infiltration on the 13C to 15N ratio within the injection zone was assessed by assuming that 80% of pig slurry liquids and 50% of cattle slurry liquids were reallocated to the soil (compare Table 1 and Fig. 6). However, even if no native slurry C was mobile and all slurry N was mobile, this could not explain 13C to 15N ratios below 1.5 (calculations not shown). Therefore, physical redistribution of slurry liquids could not alone explain the low isotopic ratios observed in and around the injection slit. Acetate is readily degraded in soils (Alexander, 1981), and losses of 13C to the atmosphere would lower the 13C to 15N ratio. Previous laboratory studies with a soil–slurry model system have found intense biological activity at the soil–slurry interface within 24 h of slurry application (Petersen et al., 1996; Frostegård et al., 1997), which could have removed part of the 13CH3COOH. Ammonium retention by adsorption to soil particles is related to residence time, or slurry liquid infiltration rate. Olesen et al. (1997b) found the infiltration rate to decrease with increasing slurry dry matter content and proposed that a low slurry dry matter content would increase the travel distance of NH+4 derived from injected slurry. In the present study, this would result in relatively lower 13C to 15N ratios at distance from the injection slit with pig slurry compared with cattle slurry, in accordance with the observations in Fig. 10, and so this mechanism could be significant. Both CH3COOH degradation and NH+4 retention could thus have contributed to lower the 13C to 15N ratios in the injection slit. A mass balance for 13C and 15N could have helped distinguish between transport and transformations, but total fresh weights of each section were not recorded.

Impact on Nitrogen Transformations
Various studies have shown that slurry distribution, as well as slurry and soil properties, can influence the course of N transformations. Both net N mineralization (Sørensen and Jensen, 1995) and denitrification activity (Petersen et al., 1992) was greater with discrete as opposed to homogeneous application of slurry, and laboratory studies have shown an intense microbial turnover and stimulation of nitrifying and denitrifying populations around the soil–slurry interface (Petersen et al., 1992; Frostegård et al., 1997). Both processes are sources of nitrous oxide (N2O), particularly under low-oxygen conditions (Firestone and Davidson, 1989). Removal of degradable C by anaerobic digestion reduced the emission of N2O from band-applied slurry by 20 to 40% during growth of spring barley (Petersen, 1999), indicating that slurry composition can influence the activity of these microbial groups significantly under field conditions.

Soil texture also interacts with slurry components. Microplot studies with simulated injection of 15N-labeled slurry showed that plant uptake of slurry NH+4 increased with soil clay content (Sørensen and Jensen, 1995), which was explained by a greater retention of NH+4 near the zone of injection where there was less potential for microbial immobilization. In contrast, clay content had no effect on the plant availability of organic N in slurry (Sørensen and Jensen, 1998). Indirectly, slurry organic matter may of course affect the interaction between NH+4 and soil particles by the partial retention of slurry liquids, as shown in this paper.

These observations indicate that distribution and properties of both slurry and soil have a significant effect on slurry N transformations and plant availability of slurry N. The quantitative importance of differences between injection techniques must be studied further.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
The experiments reported in this paper describe the spatial distribution of slurry components after field application by direct injection. The injection zone was characterized by significant heterogeneity in the distribution of water and dissolved C and N, but systematic effects of both injection method and slurry properties were identified, albeit not by all the methods applied. Injection slit profiles were characterized by temporally stable moisture gradients, and slurry water retention of different slurries followed the same relationships. This indicates that the heterogeneity of the injection slit environment is predictable and could be taken into account in calculations of nutrient transformations. The relationship between slurry organic matter content and soil water potential defined in Eq. [3] could perhaps help define the boundary conditions of a model incorporating spatial variability. A generalization to accommodate all soil water potentials would be helpful in this respect.


   ACKNOWLEDGMENTS
 
The authors wish to thank Dr. P. Schjønning for helpful inputs to this study. The work was supported by The Danish Ministry for Food, Agriculture, and Fisheries (BÆR98).


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2003 32: 1931-1938. [Full Text]  




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (4)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Petersen, S. O.
Right arrow Articles by Ambus, P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Petersen, S. O.
Right arrow Articles by Ambus, P.
Agricola
Right arrow Articles by Petersen, S. O.
Right arrow Articles by Ambus, P.
Related Collections
Right arrow Animal Waste
Right arrow Nutrients
Right arrow Spatial Variability
Right arrow Infiltration
Right arrow Isotopes

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome