JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 1 May 2008
Published in J Environ Qual 37:925-936 (2008)
DOI: 10.2134/jeq2006.0486
© 2008 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
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 Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lapen, D. R.
Right arrow Articles by McLaughlin, N.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Lapen, D. R.
Right arrow Articles by McLaughlin, N.
Agricola
Right arrow Articles by Lapen, D. R.
Right arrow Articles by McLaughlin, N.
Related Collections
Right arrow Water Quality
Right arrow Preferential Flow
Right arrow Microorganisms
Right arrow Tillage
Right arrow Vadose Zone Processes and Chemical Transport
Right arrow Municipal Waste

TECHNICAL REPORTS

Waste Management

Effect of Liquid Municipal Biosolid Application Method on Tile and Ground Water Quality

D. R. Lapena,*, E. Toppb, M. Edwardsa, L. Sabourinb, W. Curnoec, N. Gottschalla, P. Boltona, S. Rahmand, B. Ball-Coelhob, M. Paynee, S. Kleywegtf and N. McLaughlina

a Agriculture and Agri-Food Canada, Ottawa, ON, Canada, K1A 0C6
b Agriculture and Agri-Food Canada, London, ON
c Univ. of Guelph-Kemptville, Kemptville, ON
d National Swine Research & Information Center, Iowa State Univ., Ames, IA
e Ontario Ministry of Agriculture, Food, and Rural Affairs, Stratford, ON
f Ontario Ministry of Environment, Toronto, ON

* Corresponding author (lapend{at}agr.gc.ca).

Received for publication November 7, 2006.

   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 Materials and Methods
 Results
 Discussion and Conclusions
 REFERENCES
 
This study examined bacteria and nutrient quality in tile drainage and shallow ground water resulting from a fall land application of liquid municipal biosolids (LMB), at field application rates of 93,500 L ha–1, to silt-clay loam agricultural field plots using two different land application approaches. The land application methods were a one-pass AerWay SSD approach (A), and surface spreading plus subsequent incorporation (SS). For both treatments, it took between 3 and 39 min for LMB to reach tile drains after land application. The A treatment significantly (p < 0.1) reduced application-induced LMB contamination of tile drains relative to the SS treatment, as shown by mass loads of total Kjeldahl N (TKN), NH4–N, Total P (TP), PO4–P, E. coli., and Clostridium perfringens. E. coli contamination resulting from application occurred to at least 2.0-m depth in ground water, but was more notable in ground water immediately beneath tile depth (1.2 m). Treatment ground water concentrations of selected nutrients and bacteria for the study period (~46 d) at 1.2-m depth were significantly higher in the treatment plots, relative to control plots. The TKN and TP ground water concentrations at 1.2-m depth were significantly (p < 0.1) higher for the SS treatment, relative to the A treatment, but there were no significant (p > 0.1) treatment differences for the bacteria. For the macroporous field conditions observed, pre-tillage by equipment such as the AerWay SSD, will reduce LMB-induced tile and shallow ground water contamination compared to surface spreading over non-tilled soil, followed by incorporation.

Abbreviations: A, AerWay • LMB, liquid municipal biosolids • SS, surface spreading • TKN, total Kjeldahl N • TP, total phosphorus


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
Materials and Methods
 Results
 Discussion and Conclusions
 REFERENCES
 
LIQUID sewage biosolids, or liquid municipal biosolids (LMB), are often applied to land as a nutrient source for crop growth and for reducing sewage disposal pressures on landfills, incinerators, and composting facilities. However land application of biosolids must be managed to reduce the risk of contaminants such as pathogens, nutrients, and metals from reaching ground water, adjacent surface water, and/or subsurface artificial drainage networks (Tan et al., 1993; Kanwar et al., 1997; Richards et al., 2004; Akhand et al., 2006). Tile drainage networks are often within 1m of the surface and can effectively short circuit soil drainage waters to adjacent surface water sources via flow efficient conduit systems. Moreover, in many of the poorly drained soils in eastern Ontario, Canada ground water is often very near the surface on a year round basis (Wicklund and Richards, 1962). As a result of their shallow nature, these artificial drainage and ground water systems are at heightened risk of exposure to infiltrating contaminants. To complicate matters, for many soils, macropores (e.g., worm burrows, soil cracks, abandoned root channels) can facilitate rapid gravity flow of contaminant-laden material in the vadose zone to tile and shallow ground water depth (Steenhuis et al., 1994; Shipitalo and Gibbs, 2000; Stone and Wilson, 2006). Macropores can cause a majority of pore-weighted flow that effectively bypasses the soil matrix (Watson and Luxmoore, 1986) where contaminants have greater potential to be sequestered, degraded, or in the case of nutrients, available for plant uptake.

Maintaining contaminants derived from organic amendments in the rooting zone away from ground water, artificial drainage systems, and surface water sources, is the cornerstone of responsible land application practices. This can be accomplished by better management practices for land application that include, but are not limited to: (i) judicious rates of application (Pote et al., 2001; Ball Coelho et al., 2007), (ii) pre-application tillage of the soil to foster disruption of continuous macropore networks and improve surface soil sorptivity (Fleming and Bradshaw, 1992), (iii) land application during time periods when soil macroporosity is reduced (e.g., optimal soil physical conditions based on soil water content for instance) and when soil sorptive capacity is higher (van Es et al., 2004; Turpin et al., 2007b), and/or (iv) control product placement and use application equipment that decreases the absolute amount of amendment available for ‘local’ (e.g., furrow) infiltration/surface runoff (e.g., Rausch et al., 2005).

One of the most convenient ways to deliver the product to the land surface is to apply it using one of a variety of available surface spreading methods (Kleinman and Sharpley, 2003). Many jurisdictions require that surface-applied sewage materials must be incorporated into the soil shortly after application, generally within 24 h. (e.g., Government of Ontario, 2002). For some macroporous soils, this approach may actually promote macropore flow (Shipitalo and Gibbs, 2005) of applied liquid amendments to tiles and ground water and therefore may not be the most prudent option. A surface application will have greater potential to interact with a larger number of undisturbed surface macropores per unit area, relative to more spatially-discrete furrow (e.g., Comfort et al., 1988; Ball-Coelho et al., 2007) or banding-type (e.g., Bittman et al., 2004) application approaches, for instance. In the absence of pre-tillage, the post-application incorporation may come too late with respect to minimizing infiltration or runoff potential. In addition, field trafficking during post-application tillage on soils recently wetted by liquid amendments can cause soil smearing and surface soil compaction to the detriment of soil tilth (Hakansson et al., 1988).

Given these considerations, one-pass application approaches that till the soil immediately before LMB deposition and place the product in a manner that augments soil sorption in the surface soil would seem desirable. However, the province of Ontario for instance, does not recognize surface application over pre-tilled soil as proper incorporation of LMB (Government of Ontario, 2002), and by regulation such a practice would have to be followed by a post-application cultivation. The AerWay SSD is one such one-pass approach that uses an aerator-type rolling tine to vertically till the soil immediately prior to surface deposition (Bittman et al., 2004). Nevertheless, this approach has been shown to augment infiltration in the tillage layers; and moreover, the discrete soil pockets formed during the tillage action can serve as slurry reservoirs (Turpin et al., 2007a). However, most slurry deposition research associated with these types of systems have focused attention on reducing surface runoff potential (Harrigan et al., 2006; van Vliet et al., 2006), and little work has been carried out on evaluation of such a system for reducing amendment-derived contamination to tile drain systems (Akhand et al., 2006).

The purpose of the study undertaken here was to obtain information on the potential environmental benefits of an AerWay-type one-pass LMB application method relative to a surface spreading plus subsequent incorporation approach. Specific focus was on nutrient and enteric microorganism tile and ground water quality resulting from said practices.


   Materials and Methods
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
 Discussion and Conclusions
 REFERENCES
 
Study Site and Liquid Municipal Biosolid
The study was conducted in Winchester, Ontario, Canada (latitude 45°03' N, longitude 75°21' W) in fall 2005 (October to December 2005). The soil is primarily a poorly-drained North Gower clay loam classified as a Orthic Humic Gleysol (Canadian soil classification system) or a Mollic Gleysol (FAO soil classification system). Average soil particle size (hydrometer method) over the study field for 0 to 15 cm (N = 46) and 30 to 50 cm (N = 46) depth, were 0.21 (SD = 0.03) Mg Mg–1 sand, 0.38 (SD = 0.04) Mg Mg–1 clay and 0.41 (SD = 0.04) Mg Mg–1 silt, and 0.10 (SD = 0.02) Mg Mg–1 sand, 0.42 (SD = 0.03) Mg Mg–1clay, and 0.48 (SD = 0.03) Mg Mg–1and silt, respectively. The field is relatively flat with a general N-S slope of about 0.0008 m m–1, and the field has been tile drained for several decades. The site had been under primarily conventional tillage practices since 2002, consisting of spring cultivation and fall moldboard plowing (maximum plow depth of approximately 0.2 m). Crops on this field consisted of corn (Zea mays L.), soybean (Glycine max L.), and wheat (Triticum spp.). For 2003, the field was planted in wheat and conventional tillage was employed (spring and fall cultivation). In 2004, soybeans were planted on lightly cultivated soils in spring but no fall tillage was performed. The field was left fallow and untilled in 2005 and glyphosate-based herbicides were used to control the weed population. The LMB was obtained from a municipal wastewater treatment plant that handles the waste from about 20,000 people. The sludge from the primary treatment facility is anaerobically digested.

Experimental Plots, Liquid Municipal Biosolid Land Application Methods
A portion of the Winchester field site was subdivided into eight plots (six treatment plots and two control plots); each 100 m long (length of tile drain contributing area) and centered over tile lines (100 mm diam. plastic tiles for treatment plots and 100 mm diam. clay tiles for control plots) (Fig. 1 ). The control plots were located on an adjacent field that was hydrologically (tile and surface water flow) isolated from the treatment plots. Tile depth is approximately 0.8 m at its deepest point along the plots.


Figure 1
View larger version (31K):
[in this window]
[in a new window]

  		Fig. 1. Layout of treatment and control plots (not to scale). Note: control plots were located approximately 500 m south of the treatment plots to minimize potential for cross-plot contamination. T1-6 and C1-2 = number for treatment tiles (T) and number for control tiles (C); A and SS = AerWay and surface spreading plus incorporation, respectively.

 
This study evaluated the impact on water quality resulting from a common rate of LMB application using two distinctly different application methods. There were three plot replications for each of the two application treatments; with one control plot per treatment. The application approaches consisted of an AerWay SSD (Holland Equipment Limited, Norwich, Ontario, Canada) system (designated treatment A), and surface spreading followed by incorporation using a Kongskilde vibro-flex cultivator (Kongskilde, Exeter, Ontario, Canada) (designated treatment SS). The AerWay SSD slurry deposition system used in this study is described in Akhand et al. (2006). The AerWay SSD system surface-applies slurry close to the ground immediately following the pass of rolling tines that affect aerator-type tillage of the soil (Bittman et al., 2004). The AerWay SSD unit was mounted on the back of a Husky 9000 Elite tanker (Husky Farm Equip. Ltd., Alma, Ontario, Canada) that was pulled by a tractor. The nominal tillage depth for the AerWay SSD was 0.13 m (determinations of need for ballast were made immediately before application (McLaughlin et al., 2006)), and the tines were set to a 10° angle to augment tillage action and soil pocket formation (Turpin et al., 2007a). It should be noted that the use of trade names in this manuscript is for description of research methods only, and does not imply an endorsement of a particular brand or product. The surface spreading treatment was conducted using the AerWay SSD by depositing the LMB within 0.5 m above the surface, but with the tines lifted above the ground to avoid tillage action. Subsequent tillage of the LMB over the SS area was conducted around 20 h after application at a nominal incorporation depth of 0.1 m.

Both treatments used an LMB application rate of 93,500 L ha–1. Rates were calibrated before land application using a Greyline Doppler flow meter (Greyline Inst. Inc., Long Sault, Ontario, Canada) fitted to SSD drop tube hoses, and bucket calibration methods for full, medium, and low tanker fill conditions. After each plot application, which consisted of two passes either side of the tile line (Fig. 1), the tanker was reloaded with LMB from a 35,000 L nurse tanker to maintain tanker-fill uniformity among plots. The nurse tankers were fitted, for this study, with a liquid circulation system consisting of a suite of pumps to ensure the material was mixed before field application activities. The control plots were treated in a manner similar to the treatment plots but without receiving LMB. Trafficking in the control plots was normalized with respect to the treated plots by driving the tanker filled with water over them (no water was applied however). All significance test results reported in the following sections were derived from t tests using SYSTAT v. 10 (SPSS Inc., Chicago, IL). Log-transformations were conducted if data distributions were not normal.

Tile Drain, Ground Water, and Meteorological Data Collection
At the outflow end of each treatment plot, tile drain access pits were excavated to facilitate the measurement of tile drain discharge and monitoring of tile water quality (Fig. 2 ). The excavated pit walls were supported by 1.14 m diam. steel culverts set 2.4 m deep into the ground with tile input and output access holes. The input tile line was made impermeable for approximately 1.5 m up flow of the input port to facilitate more accurate discharge measurements. Sump pumps set at the bottom of the access pits were used to remove water draining into the pits off site. Calibrated tipping bucket systems were used to measure tile discharge (tips were totaled every 15 minutes). Tips were counted by means of magnetic switches recorded with a Campbell CR23X data logger (Campbell Scientific Inc., Logan, UT). Before the tile drain effluent filled the tipping buckets, incoming tile effluent first interacted with water collection/monitoring systems installed immediately above the tipping buckets (Fig. 2). Essentially, tile effluent leaving the tile was first allowed to fill up a small water collection cup which overflowed into a larger funnel (fitted with splash plates to avoid water loss) that directed the water to the tipping bucket system. Calibrated ISCO 6712 (ISCO Inc., Lincoln, NE) automatic water samplers fitted with Teflon line were used to collect tile water from the water collection cup. The 1 L ISCO sample bottles were lined with Teflon bags. The bags and bottles were cleaned after each collection event. During no tile flow conditions, ISCO liquid level actuators (Model 1640) were used to trigger the ISCO 6712 sampling programs. The actuators were set in the tile drain water collection cup to trigger once drainage water half filled the collection cup. However, during conditions when the tiles were flowing, the actuators were set in small rainfall collectors that were placed outside the pits to actuate once rainfall reached 5-mm depth. For the control plots, tile discharge and tile effluent sampling systems were set up according to methods described in Akhand et al. (2006).


Figure 2
View larger version (18K):
[in this window]
[in a new window]

  		Fig. 2. Tile effluent sampling/monitoring system used for treatment plots. Tile effluent splash plates installed around the collection cup/funnel system to ensure all tile flow was directed to the tipping bucket are not illustrated for clarity purposes. Control plot tile effluent/discharge monitoring followed the approach given in Akhand et al. (2006).

 
For flow trigger situations, the ISCO water sampling programs were set to collect water when triggering occurred, every 15 min for 45 min post trigger, then every 30 min for 120 min, then every hour for 5 h, and the rest of the remaining samples were collected at 2-hr intervals. However, at LMB application time, samples were collected manually every 15 min starting 30 min before application (pass 1) to 2 h post application.

Three piezometers were installed in each control and each treatment plot. These were nested at depths of 0.6, 1.2, and 2.0 m (approximate relative plot locations given in Fig. 1) to quantify piezometric head, estimate saturated hydraulic conductivity (pump tests as per Freeze and Cherry, 1979), and collect ground water for chemical and microbiological analyses. Each piezometer was made from 35 mm diameter PVC pipe with 150 mm intake (slotted) lengths. The piezometers were packed with sand around the intakes and sealed with bentonite over the remaining length of pipe. To protect the piezometers from damage during cultivation/application, and from direct contamination by the LMB application, they were cut into two sections at a depth 0.3 m below the soil surface. Before LMB application, the top sections of all piezometers were removed and the open piezometer sealed tightly with plastic bags and fitting caps with sealant. After sealing the piezometers, soil was backfilled into the soil-disturbed area immediately surrounding the piezometers. One day following LMB application, the piezometer tops were carefully reattached to the bottom section by means of a PVC coupler.

Before ground water sampling, piezometer water levels were recorded using a Solinst model 101 electronic water level recorder (Solinst Canada Ltd., Georgetown, Ontario, Canada). Then the ground water was pumped out of the piezometers into containers using a Masterflex E/S portable peristaltic pump (Model 7518-02, Cole-Parmer Instrument Company, IL) (this water was discarded off site). The pump was attached to a Teflon tube water collection apparatus that was set inside each piezometer. The piezometers were left to recover for roughly 6 h, after which time, secondary pumping for water quality analyses commenced. All ground water samples were collected in sterile 1 L containers. A ground water well (slotted along its entire 3-m extent) was installed between the treatment and control plots to monitor the water table continuously at an hourly interval using a Global WL15X Water Level Logger (Global Water Instrument Inc., Gold River, CA). Hourly meteorological information was collected at the site. The weather data were logged with a Campbell Scientific Inc. CR10 datalogger.

Tile/Ground Water Quality, Liquid Municipal Biosolids, and Enumeration of Bacteria
For each treatment plot, YSI 6600 multi-parameter water quality sondes (YSI Inc., Yellow Springs, OH) were installed in the treatment tile water collection systems (Fig. 2). All sondes were set with sensors measuring: turbidity (with wiper) (range = 0– ~1000 NTU), electrical conductivity (0–100 dS m–1), oxidation–reduction potential (ORP) (–999–999 mV), pH (0–14 units), dissolved oxygen (0–50 mg L–1), temperature (–5– 45°C), and NO3–N and NH4–N (0–200 mg L–1-N). The sondes were set to record water quality parameters every 15 min. All sondes were calibrated every 2 wk. Sondes were also installed before biosolid application to obtain background water quality data. For all ground water and control/treatment tile samples collected, turbidity was measured using a WQ770 Global water turbidity meter (Global Water Instrument Inc., Gold River, CA). These measurements were compared statistically to concurrently monitored turbidity values derived from the YSI sondes to ensure comparability between the different measurement systems. All water samples were brought back to the laboratory within 24 h after water sample collection. There, samples were immediately divided and shipped overnight express to several laboratories for analysis. During transport and shipping, samples were kept in coolers with ice packs.

The Ontario Ministry of Environment's (MOE) Laboratory Services Branch in Etobicoke, Ontario, Canada conducted the water nutrient (NO3–N, NH4–N, NO2–N, TKN, PO4–P, and TP) analyses. All TKN and TP concentrations were determined by colorimetry, preceded by acid digestion, according to MOE method E3368. The other N and P species were determined colorimetrically according to MOE method E3366. Samples were prepared for enumeration of E. coli and C. perfringens as follows: All samples were decanted or collected in sterile bottles. E. coli were enumerated, using the most probable numbers (MPN) method with the USEPA approved IDEXX Colilert system (IDEXX Laboratories, Inc., Westbrook, MA). Dilutions were employed to quantify MPN for all samples. C. perfringens was determined by MPN using a modified m-CP medium (Armon and Payment, 1988). C. perfringens is often an indicator of human-based contamination (Payment and Franco, 1993).

Soil Physical and Hydraulic Properties
Soil bulk density (0- to 0.1-m depth) was measured at six locations for each plot using a 0.1 m length by 0.06 m diam. soil corer immediately before application. Larger bucket volume based determinations of bulk density were made over sites tilled by the implements used in the study in areas where no LMB was applied (to reflect post tillage bulk density). Volumetric water content of the soil was logged every 30 min to a maximum depth of 0.9 m using ESI Environmental Sensors Inc. (Environmental Sensors Inc., Victoria, British Columbia, Canada) time domain reflectometry (TDR) moisture profiling system as described in Akhand et al. (2006). Measurements of field saturated hydraulic conductivity from surface to 0.4-m depth were made on site before application using pressure and tension infiltrometers (Reynolds, 1993ab; Akhand et al., 2006; Turpin et al., 2007a). Guelph permeameter measurements of field saturated hydraulic conductivity of AerWay tilled soil (aeration pockets) were made according to the methods described in Turpin et al. (2007a).


   Results
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
Discussion and Conclusions
 REFERENCES
 
General Background Information
The application of LMB took place on Julian Day (JD) 294. LMB application times during this day differed between the plots due to nurse tanker refilling. The monitoring period post-application extended to JD 340, after which freezing temperatures hampered data acquisition. Average hourly air temperature and total hourly precipitation during the study period are given in Fig. 3 . Total rain precipitation between JD 294 and JD 340 was 124 mm. Hourly air temperature averaged 2.6°C over the study period with hourly minimum and maximum values being –14.7 and 18.8°C, respectively. The bulk density of surface soils (0–0.10 m depth) before application field activities was approximately 1.31 g cm–3 (SD = 0.08 g cm–3). For AerWay tilled soils, bulk density was found to be roughly 0.8 to 1.2 g cm–3 (total porosity = 0.70 to 0.55 m3 m–3, and air-filled porosity = 0.43 to 0.28 m3 m–3). The profile of soil water content taken immediately before LMB application, from the surface to 0.9 m, is given in Table 1 . The soil water content for the 0.6- to 0.9-m depth was near saturation as total porosity averaged ~0.44 m3 m–3 for this soil depth (SD = 0.05 m3 m–3). The soil above 0.6-m depth was not saturated at time of application. Pre-application field saturated hydraulic conductivity at surface, 0.1-m, 0.2-m, 0.25-m, and 0.4-m depth for SS and A plot soils ranged between 10–2 and 10–4 cm s–1. Saturated hydraulic conductivity derived from pump tests on the 1.2 and 2.0 m piezometers ranged between 10–3 and 10–6, and 10–4 and 10–6 cm s–1, respectively. Field saturated hydraulic conductivity for AerWay tine influenced surface soils (primarily soil pockets formed by the aeration tine made at field locations adjacent to treatment plots) ranged between 3 x 10–2 and 3 x 10–3 cm s–1. Table 2 provides key parameters of the LMB used in this study. The total solids in the LMB was relatively small, at approximately 11,933 mg L–1.


Figure 3
View larger version (24K):
[in this window]
[in a new window]

  		Fig. 3. Hourly total precipitation and average hourly temperature recorded at the site for the study period.

 

View this table:
[in this window]
[in a new window]

  		Table 1. Range in soil water contents measured using time domain reflectometry (TDR) immediately before application.

 

View this table:
[in this window]
[in a new window]

  		Table 2. Key parameters for raw liquid municipal biosolids (LMB) used in the study.

 
Tile Drain Discharge and Water Quality
Tile contamination following LMB application occurred within 15 min for most treatment plots (Table 3 ; Fig. 4 ). Plot T4(SS) did display a somewhat longer response time to application-induced tile flow (39 min from application to apparent breakthrough at tile monitoring sites) and T2(SS) was the most rapid responding tile system as LMB tile breakthrough occurred approximately 3 min after the first LMB application pass. There were no significant (p > 0.1) differences between treatment breakthrough times. t tests indicated that the total volume of soil water+biosolids (minus tile baseflow contributions) that discharged through the tile during the LMB application-induced tile hydrograph was significantly (p < 0.1) higher for the SS treatments (average volume = 1.06 m3, SD = 0.38 m3) relative to the AerWay treatments (average volume = 0.24 m3, SD = 0.17 m3) (Table 4 ). After LMB application, the tile drainage responded markedly to several key precipitation events: these notable events occurred on JD 296–299, 310, 313, 319–320, and 332–333.


View this table:
[in this window]
[in a new window]

  		Table 3. Minutes after first liquid municipal biosolid application pass that tile contamination was observed.

 

Figure 4
View larger version (24K):
[in this window]
[in a new window]

  		Fig. 4. Tile drain water quality parameters, in (mg) or (cfu) per 15 min for samples collected during the LMB application-induced tile hydrograph event (JD 294–295). Note: LMB applications on the plots did not occur simultaneously.

 

View this table:
[in this window]
[in a new window]

  		Table 4. Liquid municipal biosolid (LMB) application-induced tile hydrograph information.

 
Figure 4 shows tile water quality with respect to nutrients and bacteria for ~24 h post-application. Table 5 gives the total mass loads of all observed contaminants for the entire length of the LMB-induced tile hydrograph (Fig. 4). The mass loads of all these contaminants for this application-induced event were significantly (p < 0.1) higher for the SS treatment relative to the A treatment. Moreover, the control plot mass load data were significantly (p < 0.1) smaller than those for both A and SS treatments. The application-induced contamination events were episodic and brief, and for most treatments, nutrient mass loads returned to pre-application conditions within about 24 h after application. However, nearly 24 h post-application, mass loads of both E. coli and C. perfringens were still high; between 1e+6 and 2e+6 cfu 15min–1 for E. coli, and 4e+4 to 2e+5 cfu 15min–1 for C. perfringens. Most treatment tiles had close to zero output for these enteric microorganisms immediately before application on application day.


View this table:
[in this window]
[in a new window]

  		Table 5. Treatment average (standard error) of total mass of contaminants exported out of tile for the liquid municipal biosolid induced hydrograph period (Table 4). Pre-application background values were subtracted from data used to calculate tile mass loads.{dagger}

 
For post-application flow conditions (JD 295 and on), there was more randomness in the peak observed contaminant load values for specific tile hydrograph events. Maximum observed post-application hydrograph mass loads (per 15 min) were approximately: 1622 (mg) for TKN, 611 (mg) for TP, 383 (mg) for NH4–N, 30 (mg) for PO4–P, 7739 (mg) for NO3–N, 1.5e+7 (cfu) for E. coli, and 8.0e+6 (cfu) for C. perfringens. Except for NO3–N (SS higher than A), t tests indicated that there were no significant differences (p > 0.1) between mass export of contaminants among the treatments for data collected from the end of the application-induced tile hydrograph events to the end of the study period. Control plot contaminant mass loads, except TP, were significantly lower (p < 0.1) than respective treatment loads for this time period. Over a study season basis, E. coli exhibited significant declines in mass loads post LMB application hydrograph (Fig. 5 ). Such a trend for C. perfringens was not as strongly characterized, albeit the two highest observed (cfu 15 min–1) values, for T2(SS) and T3(A), skewed what would have been a more strongly expressed exponential decay trend. For TKN, TP, PO4–P, and NO3–N, mass loads tended to generally increase with time post application hydrograph to end of study period.


Figure 5
View larger version (14K):
[in this window]
[in a new window]

  		Fig. 5. Tile drain E. coli and C. perfringens, in cfu per 15 min, over the study period (JD > 295). Data were averaged by treatment and by tile hydrograph event (or baseflow sampling time). First-order exponential decay models (Y = predicted mass load, x = day) were fit to the E. coli data (model significant at p < 0.1); but models for C. perfringens data were not significant (p > 0.1). Error bars represent standard error.

 
Multi-parameter Sonde Data for Treatment Tiles
The multi-parameter sonde data collected for each treatment tile provided water quality information at 15-min intervals over the course of the study period and helped to confirm that the SS treatment did not reduce contaminant transport to tile as effectively as the A treatment during application. Figure 6 illustrates dramatic tile breakthrough/water quality changes for turbidity, oxidation–reduction potentials, specific conductivity, and NH4–N for all treatments. Clearly the most marked sonde parameter response to LMB application occurred for T2(SS), followed by the other SS treatment plots. On a study season (JD > 296) basis turbidity responded to precipitation rapidly and transiently (many peak values were > 100 NTU during these events), indicating rapid pore flushing at the front end of the flow events and subsequent weeping of less turbid ground waters after such events occurred. However, it should be noted that water bubble interference may have occurred to an unknown degree at high flow rates (Teti, 1996). The ORP in tile water for all treatment plots plummeted from 200 to 300 mV immediately before LMB application, to values between 50 and –25 mV during peak LMB-induced discharge; however, it took nearly 10 d post-application for all the treatment tile ORP values to return to pre-application ranges. The A plot ORP values showed the strongest delay in recovery, averaging 199 mV (SD = 43 mV) for JD 296–305, relative to SS ORP averages for that period of 267 mV (SD = 18 mV). All post LMB application hydrograph NH4–N values were generally below 0.5 mg L–1 for JD > 296; however, for JD 296 there were some preferential flow peaks around 2 mg L–1 for both treatments (T5(A) and T2(SS)). While spiking to over 2 dS m–1 at application for T2(SS), specific conductivity returned within a few days to pre-application values (0.6 to 0.8 dS m–1) for all plots. Transient drops in conductivity over the remainder of the study period were linked presumably to flow-induced dilution.


Figure 6
View larger version (33K):
[in this window]
[in a new window]

  		Fig. 6. Selected multi-parameter sonde data for treatment tiles for liquid municipal biosolid application-induced hydrograph event. ORP = oxidation reduction potential, Sp. Cond. = specific conductivity.

 
Ground Water Quality
Only the concentrations of TKN, TP, C. perfringens, and E. coli for 1.2- (Fig. 7 ) and 2.0-m (Fig. 8 ) depth piezometers for each treatment and control plot are discussed here. Pre-application ground water samples were collected on JD 265, 276, and 293. Data acquisition for the 0.6-m depth was discontinuous as the water table fluctuated in and out of this depth. The soils were saturated at 1.2-m depth over the course of the study period, however. For TKN, pre-application ground water concentrations for all treatment and control plots were below 1.2 and 0.48 mg L–1 for the 1.2-m and the 2.0-m depth piezometers respectively (this included samples taken during preferential flow events). For TP, the maximum observed pre-application concentration (1.2 m) was 0.57 mg L–1. C. perfringens was not detected in any 1.2 or 2.0 m ground water samples collected pre-application. For E. coli, the largest pre-application background concentration for all 1.2 m piezometers was observed for T5(A) at >25000 cfu L–1. Thus, interpretation of T5(A) ground water E. coli concentrations was somewhat problematic. The largest E. coli (1.2 m) pre-application reading, outside of T5(A) was only 300 cfu L–1. The greatest observed pre-application E. coli concentration for all 2.0-m depth piezometers was 40 cfu L–1 (for both T5(A) and C2). Overall, E. coli (1.2 m and 2.0 m) concentrations showed apparent exponential decreases with time (inferred die-off) (Fig. 7) post-application. For C. perfringens, such apparent die-off trends were not as strongly expressed. Comparing pre- versus post-application concentrations, C. perfringens was significantly (p < 0.1) higher post-application for both treatments at 1.2 and 2.0 m, while at 1.2 m only the SS treatments showed significantly higher post-application concentrations of TKN and E. coli. There were no significant differences (p > 0.1) in E. coli concentrations pre- versus post-application for both treatments at 2.0 m, and nutrient concentrations at this depth could not be compared due to insufficient data.


Figure 7
View larger version (22K):
[in this window]
[in a new window]

  		Fig. 7. Selected nutrient and bacteria concentrations in ground water collected from 1.2 m piezometers. Data were averaged by treatment for each sampling event. Error bars represent standard error. First-order exponential decay models (Y = predicted concentration, x = day) were fit to the E. coli and C. perfringens data. Both models were significant (p < 0.1).

 

Figure 8
View larger version (20K):
[in this window]
[in a new window]

  		Fig. 8. Selected nutrient and bacteria concentrations in ground water collected from 2.0 m piezometers. Data were averaged by treatment for each sampling event. Error bars represent standard error. First-order exponential decay models (Y = predicted concentration, x = day) were fit to the E. coli and C. perfringens data. Only the E. coli model was significant (p < 0.1).

 
For 1.2-m depth, post-application TKN, TP, E. coli, and C. perfringens concentrations were significantly (p < 0.1) higher for both treatment plots relative to controls (Table 6 ). Moreover, t tests showed post-application concentrations of TKN and TP (1.2 m) to be significantly (p < 0.1) higher for SS relative to A treatments, but there were no significant differences in study period E. coli or C. perfringens concentrations (1.2 m) among treatments. At 2.0-m depth, control plots had significantly (p < 0.1) higher TKN concentrations than both treatment plots, and TP concentrations were significantly higher for control and SS treatment plots than they were for the A plots. There were significantly higher (p < 0.1) concentrations of E. coli, but not C. perfringens, for both treatments relative to controls, and no significant differences were found for bacteria among treatments (as for 1.2-m depth). Thus, there was bacterial contamination to at least 2.0-m depth.


View this table:
[in this window]
[in a new window]

  		Table 6. Summary of t test results for selected post-application study period ground water nutrients and bacteria concentrations. TKN, total Kjeldahl N; TP, total P.{dagger}{ddagger}

 

   Discussion and Conclusions
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
 Discussion and Conclusions
REFERENCES
 
Overall, this study found that the AerWay SSD land application approach significantly reduced application-induced contaminant transport of LMB to tile drains and shallow ground water, relative to surface spreading LMB and subsequent (within 24 h) incorporation of the material into surface soils. On average, the greatest contaminant mass loads for all singular tile hydrograph events occurred as a result of LMB application. Average mass load ratios (SS:A) for this application-induced tile hydrograph event was 19 for TKN, 21 for NH4–N, 3 for NO3–N, 12 for TP, 15 for PO4–P, 6 for E. coli, and 12 for C. perfringens. Overall, the SS treatment was a less desirable LMB land application approach with respect to tile contamination that can occur as a direct result of application, at least for the given low-relief silty-clay-loam soils observed. It was also found for these soil conditions that tile contamination can potentially occur within minutes following application, irrespective of land application treatment. Based on LMB breakthrough at tile monitoring sites, contributing pore networks directly interacting with the tile drain (Shipitalo and Gibbs, 2000) facilitated LMB flow at rates between 0.03 to 0.44 cm s–1 and 0.12 to 0.15 cm s–1 for SS and A treatments, respectively. The fact that these flow rate values were similar between treatments (p > 0.1) suggests that there were similar sub-tillage layer pore networks contributing flow to the tiles, irrespective of application approach. Tile drain LMB breakthrough times were clearly dominated by preferential LMB flow through macropore networks interconnected to tile lines where gravity flow dominated, and where soil matrix filtering and LMB viscosity effects were less strongly expressed. The total amount of flow that discharged during the LMB application event averaged 1.06 and 0.24 m3 for SS and A treatment plots, respectively. The aggressively set (i.e., increased rolling tine angle to augment tillage action and soil pocket formation in the soil) aerator-based tillage modification of the soil was directly responsible for such differences in tile hydrograph volumes since the same SSD applicator was used to apply LMB to the surface for both treatments (SS sans tillage). Increased porosity, lateral infiltration, and sorptivity in the tillage layers above the resident plow pans occurred as a result of this tillage action (Turpin et al., 2007a, 2007b). These factors combined to augment LMB storage in the tillage layers. Without pre-tillage of the soil, the surface-spread LMB was allowed to interact with: (i) undisrupted macropores at the surface that were interconnected to tile or deep in vertical extent and (ii) surface soils with lower porosities.

The silty-clay loam soils studied here have been determined to be highly macroporous with respect to earthworm activity. In fact, Ouellet et al. (2008) found that for approximately 250 agricultural field sites studied, mostly throughout eastern Ontario, the Winchester soils had some of the greatest worm biomasses. The largest vertical burrows contributing flow to depth were formed by L. terrestris. Burrow depths as observed at this site went to at least 1.2-m depth (D.R. Lapen, unpublished data, 2004), which helps to explain the nature of some of the flow-contributing pore networks (Shipitalo and Gibbs, 2000) to shallow ground water. Disruption of these large pores beneath the tillage layer is difficult, thus, partly explaining common contaminant breakthrough times during application among the different treatments. So in this case, increasing surface soil LMB storage potential via AerWay tillage action reduced the amount of contaminated material that reached sub-tillage layer contributing pore networks. But effective pre-tillage approaches under the observed conditions did not completely attenuate contaminant transport to tile, nor decrease the speed at which contaminants reached tile post-application.

Clearly, LMB contamination occurred in ground water to at least 2.0-m depth, as observed immediately after application. On average, the SS treatment appeared to promote more generalized contamination of shallow ground water than did the A treatment by virtue of the lack of ‘pre’-tillage action effects on soil hydraulic properties discussed previously. In addition, ground water E. coli concentrations at 1.2- and 2.0-m depth after application exhibited first-order die-off trends with rate coefficients (around 0.11 d–1) consistent with those of E. coli in water (0.15 d–1) at 5°C (Reddy et al., 1981).

For most contaminants examined, differences between the two treatments with respect to water quality were not significant after application-induced tile drain hydrographs ceased. This was due in part to: (i) all treatment plots receiving a form of tillage action before the major rain event post-application, and (ii) homogenized treatment differences due to large post-application rain events (that occurred soon after application) potentiating translocation of LMB contaminants to sub-tillage layer pore domains (i.e., taking LMB out of tillage layer storage). These rain events (e.g., JD 296) saturated the surface soils and therefore maximized (i) water-LMB interactions in the surface soils and (ii) gravity-based transport associated with the larger intrinsic pores that operate at low soil water tensions.

Overall, this study suggests that under field conditions characteristic of many fine-textured soils in eastern Ontario that are tile-drained, application of LMB by equipment such as the AerWay SSD will reduce tile water contamination compared to surface spreading LMB followed by incorporation.


   ACKNOWLEDGMENTS
 
Funding for this project was provided by: The Agriculture Policy Framework's GAPs program and Agriculture and Agri-Food Canada Matching Investment Initiative in cooperation with Ontario Federation of Agriculture, Ontario Ministry of Agriculture, Food and Rural Affairs, and Ontario Ministry of Environment. We wish to thank Mr. D. Irving and Mr. A. Smith (Univ. of Guelph-Kemptville), and Mr. M. Mayer (Holland Equipment Ltd.) for invaluable project support.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
 Discussion and Conclusions
 REFERENCES
 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
 Discussion and Conclusions
 REFERENCES
 




This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
N. Gottschall, M. Edwards, E. Topp, P. Bolton, M. Payne, W. E. Curnoe, B. B. Coelho, and D. R. Lapen
Nitrogen, Phosphorus, and Bacteria Tile and Groundwater Quality Following Direct Injection of Dewatered Municipal Biosolids into Soil
J. Environ. Qual., May 1, 2009; 38(3): 1066 - 1075.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
M. Larsbo, D. R. Lapen, E. Topp, C. Metcalfe, K. C. Abbaspour, and K. Fenner
Simulation of Pharmaceutical and Personal Care Product Transport to Tile Drains after Biosolids Application
J. Environ. Qual., April 27, 2009; 38(3): 1274 - 1285.
[Abstract] [Full Text] [PDF]

Home page
 Agron. J.Home page
B. R. B. Coelho, R. C. Roy, A. J. Bruin, A. More, and P. White
Zonejection: Conservation Tillage Manure Nutrient Delivery System
Agron. J., January 8, 2009; 101(1): 215 - 225.
[Abstract] [Full Text] [PDF]

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 Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lapen, D. R.
Right arrow Articles by McLaughlin, N.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Lapen, D. R.
Right arrow Articles by McLaughlin, N.
Agricola
Right arrow Articles by Lapen, D. R.
Right arrow Articles by McLaughlin, N.
Related Collections
Right arrow Water Quality
Right arrow Preferential Flow
Right arrow Microorganisms
Right arrow Tillage
Right arrow Vadose Zone Processes and Chemical Transport
Right arrow Municipal Waste

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