| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Plant and Earth Science, Univ. of Wisconsin-River Falls, 410 S. 3rd St., River Falls, WI 54022
b Dep. of Soil, Water, and Climate, Univ. of Minnesota, 1991 Upper Buford Circle, St. Paul, MN 55108
* Corresponding author (sgupta{at}umn.edu).
Received for publication July 24, 2007.
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
Abbreviations: ELISA, enzyme-linked immunosorbent assay • GS, growing season, NGS, non-growing season
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
Antibiotics are extensively used in human and veterinary medicine. Although human use of antibiotics is restricted to therapeutic purposes, veterinary use includes therapeutic and subtherapeutic (growth promotion and prophylactic) purposes. In the USA, it is estimated that between 40 and 84% of antibiotics are used for agricultural purposes, with the majority being used subtherapeutically in livestock production (Shea, 2003). Although dosages are small (<200 g ton–1 feed), as much as 70 to 90% of the antibiotics fed are excreted in manure (Levy, 1992). The most commonly used subtherapeutic veterinary antibiotics include tetracyclines (chlortetracycline and oxytetraycline), ionophores (monensin and lasalocid), peptides (bacitracin), macrolides (tylosin and erythromycin), sulfonamides (sulfathiazole and sulfamethazine), and β-lactams (penicillin) (Mellon et al., 2001). With the exception of the ionophores, the majority of veterinary antibiotics are identical or similar to those used to treat human diseases (Shea, 2003).
Because antibiotics are only partially metabolized by animals, manure often contains high levels of antibiotics (Kumar et al., 2005). Antibiotic concentrations in manure range from trace levels to >200 mg kg–1 or L–1, with typical concentrations in the 1 to 10 mg kg–1 or L–1 range (Kumar et al., 2005). In the USA, animal production generates an estimated 132 million metric tons (dry weight) of manure annually (USDA-ERS, 2005). The majority of manure is applied to agricultural land as a nutrient source for crop production. Approximately 9.2 million hectares of land receive manure applications annually in the USA (USDA-NASS, 2005). There is concern that land-applied antibiotics in manure are reaching surface and ground waters (Yang and Carlson, 2003; Kim and Carlson, 2006; Lissemore et al., 2006) and ultimately contributing to the development and spread of antibiotic resistance in the environment.
As concerns over increased prevalence of antibiotic resistance grow, widespread subtherapeutic use of antibiotics for livestock production has been heavily scrutinized and continues to be controversial (Mellon et al., 2001). However, there are a limited number of field studies that have quantified the extent of antibiotic transport from manure-applied fields. A study by Hamscher et al. (2002) investigated leaching losses of tetracycline and chlortetracycline from a manure-amended sandy soil. Although manure contained 4.0 and 0.1 mg kg–1 of tetracycline and chlortetracycline, respectively, these antibiotics were not detected in 80-cm-deep soil water or 200-cm-deep ground water. Kay et al. (2004 and 2005b) evaluated leaching and surface runoff of sulfachloropyridazine, oxytetracyline, and tylosin. Peak concentrations of sulfachloropyridazine and oxytetraycline in leachate were 613 and 36 µg L–1, respectively, whereas peak concentrations in runoff were 703 and 72 µg L–1, respectively. Tylosin was not detected in leachate or runoff waters. Using simulated rainfall, Burkhardt et al. (2005) reported peak concentrations of sulfadiazine, sulfadimidine, and sulfathiazole in runoff from manure-amended grassland ranged from 490 to 1030 µg L–1, whereas mass losses ranged from 0.6 to 2.1% of the antibiotics applied with manure. A similar simulated rainfall study by Kreuzig et al. (2005) reported mass losses of sulfonamides varying from 0.1 to 2.5% of the antibiotics applied with manure for arable land and 13.3 to 27.4% for grassland. From a cultivated bare soil under simulated rainfall, Davis et al. (2006) reported concentrations of a variety of tetracycline, macrolide, sulfonamide, and ionophore-class antibiotics in surface runoff were generally <5 µg L–1, with mass losses <0.1% of the applied antibiotics.
The influence of soil management practices such as tillage on antibiotic losses is poorly understood. Isensee et al. (1990) showed that leaching losses of organic contaminants are often greater under no-tillage compared with other conservation or conventional tillage practices because of macropore persistence. A few studies have shown that tillage can reduce antibiotic losses in leachate and runoff (Kreuzig et al., 2005; Kay et al., 2004; Kay et al., 2005a); however, antibiotic losses from different tillage practices have not been systematically evaluated.
The karst region of the Upper Midwest (southeastern Minnesota, southwestern Wisconsin, northeastern Iowa, and northwestern Illinois) is a major animal production area, and land application of manure is a common practice in this region. The soils in the area are relatively shallow and contain macropores and interpedal fractures, which increases the interconnectivity between surface and ground water systems. A Minnesota Department of Agriculture survey (MDA, 2006) showed that approximately 34% of private wells in Southeastern Minnesota exceed the 10 mg L–1 nitrate-nitrogen drinking water standard, compared with only 12% of private wells outside of the Minnesota karst and sand plain aquifers. A link between high nitrate-nitrogen concentrations and pesticide detections was also reported for this region (MDA, 2006).
The objective of this study was to quantify chlortetracycline, monensin, and tylosin losses in leaching and runoff under two conservation tillage practices (chisel plowing and no-tillage) in the karst region of the Upper Midwest. Physicochemical characteristics of chlortetracycline, monensin, and tylosin are presented in Table 1 . Collectively, these three antibiotics account for approximately 40% of all subtherapeutic antibiotic usage in livestock production (Mellon et al., 2001). Contrary to previous studies that have been conducted for short periods under simulated rainfall, this study was designed to quantify year-round antibiotic losses under natural weather conditions, including when soils are frozen.
|
|
Materials and Methods |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
The dominant soil at the site is Rozetta silt loam (fine-silty, mixed, superactive, mesic Typic Hapludalfs). The A horizon was approximately 30 cm thick. Below the A horizon was a series of argillic horizons from a depth of approximately 30 to 150 cm. Chert fragments and bedrock were present at a depth of approximately 2.5 m (De Bonis, 1996). Particle size distribution of the surface soil at the site is 19% sand, 69% silt, and 12% clay, whereas the organic matter content is 0.036 g g–1. Surface soil pH is approximately 7.0. Ponded infiltration rates at the site vary depending on the time of the year and the presence of a surface seal. Munyankusi (1999) measured double-ring infiltration rates of 12.0 ± 5.3, 28.4 ± 11.0, and 9.1 ± 5.7 m d–1 at the soil surface, 45 cm, and 90 cm, respectively. Hanewall (1996) found that the earthworm population was as high as 153 worms m–2 at the site.
Tillage and Manure Treatments
The experimental setup at the site consisted of a randomized complete block design in a split-plot arrangement, with tillage as the main treatment and manure source as the subtreatment. Tillage treatments were chisel plowing and no-tillage, whereas manure sources were beef manure, hog manure, and no manure (control). Different manure sources were selected because antibiotic usage varies with animal species. No antibiotics were present in the control treatment because there was no manure application. Tillage treatments at the site have been in operation since 1993, and manure had not been applied since 1997. The site has been managed under continuous corn (Zea mays L.) since 1992. Each tillage and manure source treatment combination was replicated twice for a total of 12 plots, each measuring 18.3 m long by 4.9 m wide. Plots were separated by a 1.8-m cropped buffer. Slope across the study site was approximately 12% facing northeast.
Manure was applied in the fall shortly after harvest, a common practice in this region. Manure was obtained from local farms in the area that mixed subtherapeutic antibiotics into their livestock feed. The same farms were used during the entire study period. Hog manure was obtained from a liquid lagoon, whereas beef manure was a solid daily scrape and haul system. Hog manure contained chlortetracycline and tylosin, whereas beef manure contained chlortetracycline, monensin, and tylosin. During the second and third years of the study, manure was supplemented with a small quantity of feed-grade antibiotics (<0.5 kg feed additive per ton beef manure and <1 kg feed additive per 1000 gallon of liquid hog manure); tylosin was applied as Tylan 40 (Elanco Animal Health, Greenfield, IN), chlortetracycline as Aureomycin 50 (Alpharma, Inc., Bridgewater, NJ), and monensin as Rumensin 80 (Elanco Animal Health). However, the addition of feed antibiotics did not have a measurable effect on antibiotic concentrations in manure, possibly due to adsorption onto manure.
Twenty-four hours before manure application, each feed additive was mixed with approximately 10 L of water and shaken several times to make a uniform suspension. The suspension was added to the liquid hog manure tank at the time of filling. Uniform mixing was achieved during transportation (approximately 2 mi). In 2005, the manure spreader had an auger that continuously mixed the manure during filling and land application. A pump circulated the manure through the tank when the applicator was stationary. For beef manure, the feed additive suspension was spread over the beef manure and mixed many times with the bucket of a front-end loader. To improve uniformity, beef manure mixing was done by dividing the pile into quarters and then intermixing the quarters together.
Manure application rates were based on University of Wisconsin nitrogen recommendations for continuous corn in the area (180 kg N ha–1). In the chisel plow treatment, manures were surface applied and incorporated with tillage, which took place along landscape contours. For the no-tillage treatment, manures were surface applied and were not incorporated. Hog manure was surface applied using a liquid side spreader in 2003 and 2004 and an injector operated above ground in 2005. Visual observations showed more uniformity of manure application with the above-ground injector than with the side spreader. In 2003 and 2004, manure samples were collected in four aluminum trays placed in each plot before manure application. These samples were mixed together to obtain one composite sample per plot. In 2005, four manure samples were taken directly from the manure tank before land application.
Beef manure was applied from an auger-driven back-end manure spreader. Beef manure application rates varied from 54 to 90 Mg ha–1, whereas liquid hog manure rates varied from 65,500 to 131,000 L ha–1 over the 3-yr period. Solid content averaged 29.3% for beef manure and 4.3% for hog manure. After beef manure application, a small amount of manure from several locations within each plot was combined to obtain one composite sample per plot. These small samples were combined into four subsamples before antibiotic analysis. The no-manure treatment was supplemented with an equivalent nitrogen application of urea in the spring for crop production. The average residue cover after planting in the spring was 36% for the chisel plow and 81% for the no-tillage treatment.
Water Collection
Each plot was equipped with a pan and a wick lysimeter (Barbee and Brown, 1986) that were installed 60 cm below the soil surface. Pan lysimeters collected water from macropores or when the soil was saturated, whereas wick lysimeters collected water from macropores, saturated soil, and unsaturated soil (matrix flow at 50 cm suction). The lysimeters were constructed from PVC sheets and measured 79 cm long and 36 cm wide (Munyankusi, 1999; Gupta et al., 2004). The lysimeters continuously collected percolating water, which was directed into a 19-L collection vessel. Lysimeter collection efficiency was not measured. Between April and November, water in the collection vessel was measured twice monthly and then subsampled for antibiotic analysis. Samples were frozen at –20°C immediately after collection. During December through March, leachate collected in the vessel; however, lysimeters were not sampled due to the possibility of water freezing in the small-diameter (6.4 mm OD and 4.3 mm ID) sampling tubes, thus preventing subsequent sampling.
For runoff measurements and collection, each plot was isolated from the surrounding area using galvanized corrugated steel (16 gage, 20.3 cm by 81.3 cm sheets), which were pounded into the ground approximately 15 cm deep. Runoff was directed to a 1.7-m-wide collector at the bottom of each plot, which was connected to a PVC pipe that routed the runoff to a tipping bucket device that was connected to a datalogger (Campbell Scientific, Inc., Logan, UT). The datalogger recorded number of tips, which was converted to runoff volume by multiplying the number of tips by the volume of each tip. A small portion of the runoff water from alternating tips was collected in a 19-L PVC container. Runoff samples were retrieved immediately after each snowmelt or rainfall event and frozen at –20°C.
Analytical Procedures
Water and manure samples were prepared for analysis by thawing the frozen samples at room temperature followed by mechanical homogenization (liquid samples) or manual homogenization (solid samples). Leachate and runoff water samples were filtered though a nonsorptive 0.45-µm filter (dissolved phase) before analysis. Although leachate did not contain significant sediment, sediment-bound antibiotics in runoff were not analyzed.
Antibiotic analysis of manure samples included water-extractable and total-extractable antibiotics. For water-extractable antibiotics, 5 g of manure was mixed with 10 mL of nano-pure water, vortexed for 1 min, shaken on an end-over-end shaker for 15 min at 4°C, and centrifuged at 2000 g for 20 min, and the supernatant was collected. This step was repeated with an additional 10 mL of nano-pure water. Supernatants were combined, centrifuged at 2000 g for 20 min, and filtered through a non-sorptive 0.45-µm filter. For total-extractable antibiotics, the remaining solids were mixed with 10 mL of an 80:20 (v/v) methanol:water solution, vortexed, shaken, and centrifuged, and the supernatant collected. This step was repeated with an additional 5 mL of the methanol:water solution. Both supernatants were combined, centrifuged, and filtered as described for the water extraction.
Antibiotic analysis was conducted using enzyme-linked immunosorbent assay (ELISA) methodology. The underlying basis of ELISA analysis is a highly specific and sensitive antibody–antigen reaction. The ELISA method was chosen primarily for its high sensitivity, along with its ability to rapidly and cost-effectively analyze large numbers of samples. Although generally considered semi-quantitative, ELISA analysis has been effective as a quantitative analytical technique when sufficient selectivity and sensitivity have been established (Aga et al., 2003; Kumar et al., 2004; Aga et al., 2005; Dolliver, 2007).
Commercially available tetracycline (r-Biopharm, South Marshall, MI), monensin (Immuno-Diagnostic Reagents, Vista, CA), and tylosin (International Diagnostic Systems Corp, St. Joseph, MI) ELISA kits for antibiotic residue analysis in food products were adapted for analysis of water and manure samples (Kumar et al., 2004; Dolliver, 2007). Analysis was performed as outlined by the kit manufacturer. Quantification was performed at the specified wavelength using a spectrophotometer (Molecular Devices, Sunnyvale, CA). The tetracycline kit had 100% cross-reactivity toward chlortetracycline (Kumar et al., 2004). There was no cross-reactivity for the antibiotics used in this study across kits. Water samples were diluted 1:1 with kit-specific dilution buffer, whereas manure samples were diluted equal to or greater than 1:10 to eliminate the sample matrix effects (Aga et al., 2003; Dolliver, 2007).
Limits of quantification were 0.25 µg L–1 for chlortetracycline and tylosin and 3.0 µg L–1 for monensin. Standard curves (r2 > 0.95) were constructed for each kit from standards run in triplicate. Initially, all water samples were run in singlicate. Based on the initial analysis, near-positive (one SD below the limit of quantification) and positive samples were re-run for confirmation. Due to variability, manure samples were run in triplicate. The coefficient of variation for intra- and inter-plate variability was typically <20%.
Soluble phosphorus was determined using the ascorbic acid method of Murphy and Riley (1962) after filtration of water samples through a 0.45-µm membrane filter. Dissolved organic carbon was analyzed by high-temperature combustion of filtered (0.45 µm) HCl-acidified water samples using a Shimadzu TOC-V CSH/CSN analyzer (Shimadzu Scientific Instruments, Columbia, MD).
Statistical Analysis
Data were divided into non-growing season (NGS) and growing season (GS) to examine seasonal effects. Non-growing season was the period from harvest/manure application to planting (approximately November–April), whereas the growing season started from the time of planting and lasted until harvest (approximately May–October).
Statistical analyses of water percolation and runoff were performed using mixed model procedures for repeated measures with SAS 9.1 (Littell et al., 1998; SAS Institute, 2004). Year and season were modeled as repeated measures. Data were log-transformed to meet assumptions of normality and equality of variances. Tillage treatment, manure source, sampler, season, and year, along with their interactions, were considered fixed effects, whereas replicates and interactions were assumed to be random effects. Significance of fixed effects and interactions were determined based on the Kenward-Roger adjusted denominator degrees of freedom approximation (Kenward and Roger, 1997). Covariance structures (compound symmetry, unstructured, and first-order autoregressive) were assessed using the restricted highest likelihood method (Littell et al., 1998). Akaike's Information Criterion was used to select the best-fitting covariance structure for each response variable. Variance components with zero variance were removed from the model. Mean separation was conducted from pairwise differences of least-squares means for fixed effects. Effects were considered significant at P 
0.10. Data are reported as back-transformed means.
Due to the prevalence of zero values in the antibiotic transport data, non-normality could not be remedied with transformations. Therefore, the nonparametric Kruskal-Wallis test was used to evaluate differences in rank sums for main effects. Effects were considered significant at P 0.10.
|
Results |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
Water Percolation
As a percent of total precipitation, water percolation to 60 cm was 9% in 2004, 8% in 2005, and 11% in 2006. Due to matrix flow, water percolation was slightly higher, but not significantly different, in the wick versus pan lysimeters; therefore, percolation values reported in this paper are averaged over both lysimeters. The lack of significant differences in water percolation between pan and wick lysimeters indicates that water percolation was primarily through macropore flow or when the soil was saturated.
In general, water percolation was higher for no-tillage compared with the chisel plow treatment (Fig. 1 ), which likely reflects a combination of greater residue cover and abundance of earthworm macropores. Higher surface residue cover helps prevent the development of surface seal in silty soils, which in turn results in greater infiltration (Ela et al., 1992). The lack of annual soil disturbance in the no-tillage treatment helps maintain the surface continuity of macropores, thus allowing greater water entry and therefore higher water percolation. Hanewall (1996) showed that no-tillage plots at the site had significantly higher total earthworm numbers compared with chisel plow plots. Lumbricus terrestris L., a deep-burrowing, surface-feeding earthworm, accounted for approximately 15% of the total earthworm population. These earthworms typically form vertical burrows that are open at the surface; however, tillage practices such as chisel plowing disrupt the continuity of these macropores, thus reducing water entry (Zachmann et al., 1987).
|
Water Runoff
There were large variations in surface runoff as a result of varying site conditions and precipitation patterns. In general, runoff water losses were much smaller than percolation water losses (Fig. 1 and 2
). Annual runoff was highest in 2005 as a result of several large snowmelt and precipitation events, whereas runoff in 2006 was substantially lower than in 2004 and 2005 (Fig. 2). As a percent of annual precipitation, runoff was 1% in 2004, 3% in 2005, and <1% in 2006. There were no seasonal trends in runoff (Fig. 2).
|
During our study, the damming effect from chisel plowing was confounded in some years due to localized conditions. In a few chisel plow plots, water broke through the tillage dam and led to channelization. In a few cases, these channels reached the runoff collector, resulting in higher runoff. When the incidence of channelization was excluded, runoff was at least 50% greater from the no-tillage compared with the chisel plow treatment during the NGS across all 3 yr. During the GS, the tillage trend reversed with slightly higher runoff from the chisel plow compared with no-tillage treatment. This reflects the diminished surface roughness and damming in chisel plow as well as the dominance of the residue cover in minimizing soil detachment by raindrop impact in the no-tillage plots. Reduced roughness in the chisel plow plots were due to soil settling and smoothing from planting practices and successive rainfall and runoff events. Runoff was highest from the no manure, followed by the hog manure and beef manure treatment (Fig. 2); however, this was not statistically significant.
Antibiotic Application from Manure
Antibiotic application varied between manures and among years (Table 2
). Total and water-extractable chlortetracycline application rates were similar between beef and hog manure treatments (Table 2). Total-extractable chlortetracycline application varied from 12 to 36 g ha–1 for beef manure, compared with 16 to 38 g ha–1 for hog manure, over the study period. Total-extractable tylosin application rates were significantly higher for hog manure (47–775 g ha–1) than beef manure (7–15 g ha–1). Total-extractable monensin application rates in beef manure ranged from 242 to 922 g ha–1. Monensin was not detected in hog manure because it is not approved for use in swine production.
|
Antibiotic Detections and Concentrations
Leachate
Due to a lack of significant difference in percolation volume between pan and wick samplers, antibiotic data were averaged across samplers. Chlortetracycline was not detected in leachate from any of the manure treatments during the study period (Table 3
). Monensin and tylosin were detected relatively infrequently (in approximately 8% of samples). For tylosin, which was present in beef and hog manures, 76% of leachate detections occurred from hog manure plots (Table 3). Tylosin concentrations in hog manure were approximately 85% higher than concentrations in beef manure (Table 2), which likely contributed to the higher frequency of detection. The highest monensin and tylosin concentrations in leachate from the beef manure plots were 40.9 and 1.0 µg L–1, respectively. Comparatively, the highest tylosin concentration in leachate from the hog manure plots was 1.2 µg L–1.
|
|
Runoff
Chlortetracycline in runoff was detected in approximately 2% of samples, whereas monensin and tylosin were detected in approximately 20% of samples (Table 4
). Chlortetracycline in runoff was only detected from the hog manure treatment (Table 4). The highest monensin and tylosin concentrations in runoff from the beef manure plots were 57.5 and 1.9 µg L–1, respectively. The highest chlortetracycline and tylosin concentrations in runoff from the hog manure plots were 0.5 and 6.0 µg L–1, respectively.
|
With the exception of 2005, there was no consistent effect of tillage on antibiotic detections in runoff. During 2005, 80% of monensin detections and 94% of tylosin detections were from the no-tillage treatment, which had approximately 78% greater water runoff than the chisel plow treatment (Fig. 2).
Antibiotic Losses
Leaching
There was no leaching of chlortetracycline during the study period. Annual monensin losses from the beef manure treatment were generally <20 mg ha–1, with a highest loss of 74 mg ha–1, which occurred in 2006 (Fig. 4
). Annual tylosin losses from the beef and hog manure treatments were generally <0.5 mg ha–1; the highest loss was 1 mg ha–1, which occurred in 2006. Despite significantly lower application rates and detections of tylosin in the beef manure compared with the hog manure treatment, tylosin losses from both manure treatments were similar as a result of higher water percolation in the beef manure compared with the hog manure treatment (Fig. 5
). On a relative mass basis, leaching losses for all three antibiotics were <0.005% of the total amount applied in manure.
|
|
Runoff
Runoff losses of chlortetracycline, monensin, and tylosin were variable during the study period (Fig. 5). Annual runoff losses of chlortetracycline from the hog manure treatment were generally <0.5 mg ha–1; the highest loss of 3 mg ha–1 occurred in 2006. There were no chlortetracycline losses in runoff from the beef manure treatment. Annual monensin losses from the beef manure treatment and tylosin losses from the beef and hog manure treatments were substantially higher in 2005 compared with 2004 and 2006 (Fig. 6
), mainly because of extremely high NGS runoff from the no-tillage treatment (Fig. 2). The highest monensin loss from the beef manure treatment in 2005 (14,210 mg ha–1) was more than 400 times higher than the highest loss observed in 2004 or 2006 (15 mg ha–1) (Fig. 5). Comparatively, the highest tylosin loss from the hog manure treatment in 2005 (419 mg ha–1) was approximately 20 times higher than the highest loss in 2004 or 2006 (20 mg ha–1). Approximately 97% of the annual monensin loss and 98% of the annual tylosin loss in 2005 occurred during two large consecutive 4-d snowmelt runoff events on 4 through 7 February and 12 through 15 February.
|
Monensin, which was only applied in beef manure, was detected in runoff from the hog manure and no-manure treatments (Fig. 5). This is most likely due to cross-contamination between plots, possibly by manure application crossover during spreading, carry-over of manure mixed soil by implements during tillage and planting, and/or breaks in the sheet metal edging during high-flow runoff events. In all cases, cross-contamination occurred in plots adjacent to beef manure plots.
Non-growing season runoff losses from the beef manure treatment accounted for >99% of monensin and 100% of tylosin losses in runoff. For the hog manure treatment, NGS losses accounted for 100% of chlortetracycline and >99% of tylosin losses in runoff. With the exception of 2005, there was no significant effect of tillage on annual losses of chlortetracycline, monensin, or tylosin in runoff due to high temporal and spatial variability. In 2005, monensin losses from the beef manure treatment (8314 vs. 1254 mg ha–1) and tylosin losses from the beef manure (209 vs. 0 mg ha–1) and hog manure (256 vs. 0.01 mg ha–1) treatments were significantly higher from the no-tillage compared with the chisel plow treatment.
Total Losses
Consistent with earlier findings, >99% of total (leaching and runoff) losses for each of the three antibiotics occurred during the NGS. With the exception of chlortetracycline losses, which were too infrequent to characterize, total losses for monensin and tylosin followed NGS water loss trends (Fig. 1, 2, and 6). Leaching losses accounted for approximately 60% of the total losses for monensin and tylosin from beef manure treatments in 2004 and 2006 and tylosin losses from the hog manure treatment in 2004 (Fig. 6). Non-growing season percolation during these years was nearly 100 times greater than the NGS runoff. Conversely, runoff accounted for nearly 100% of monensin and tylosin losses in 2005 as a result of several very large snowmelt runoff events. Annual runoff was more than 1000% greater, whereas annual percolation was approximately 150% lower in 2005 compared with 2004 and 2006. Although not significant due to high variability between plots and between lysimeters, total losses were generally higher from the no-tillage compared with the chisel tillage treatment (data not shown).
|
Discussion |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
Chlortetracycline, monensin, and tylosin have been shown to rapidly degrade in soil (half-lives <25 d) (Table 1). However, Galvalchin and Katz (1994) showed that degradation for a variety of antibiotics, including chlortetracycline and tylosin, was slower at colder temperatures. In their study, degradation at 4°C was <30% of initial levels on average, whereas degradation at 30°C was approximately 85% of initial levels. In this study, manure application and antibiotic losses occurred during the winter months (November–April). Average temperature during the NGS was <1°C, which likely slowed antibiotic degradation. Higher degradation during the warmer GS may have contributed to lower antibiotic availability for transport during the GS. This suggests that fall manure applications in cold climates are particularly susceptible to antibiotic losses during snowmelt.
Although there were no overall significant tillage effects, there is evidence that antibiotic losses are linked to water percolation and runoff. Therefore, tillage and other management practices that reduce water losses, especially during the NGS, are the most effective at limiting antibiotic losses from fall manure applications. In this landscape, where macropores are prevalent, no-tillage is not as effective in controlling water percolation and runoff during the NGS as compared with chisel plowing along the contour.
Concentrations and annual losses were in agreement with adsorption characteristics (Kd values) for these antibiotics (Table 1). Concentrations and losses were greatest for monensin, which had the lowest adsorption to soil. The infrequent loss of dissolved chlortetracycline in this study mainly reflects its high affinity to soil. These results indicate that adsorption to soil is the primary mechanism limiting antibiotic losses from manure-amended agricultural land. On a relative mass basis, dissolved antibiotic losses in this study were within the range (<5%) that has been reported for short-term plot studies (Kay et al., 2004; Kay et al., 2005b; Burkhardt et al., 2005; Davis et al., 2006).
|
Conclusions |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResultsDiscussionConclusionsREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. L. Kuchta, A. J. Cessna, J. A. Elliott, K. M. Peru, and J. V. Headley Transport of Lincomycin to Surface and Ground Water from Manure-amended Cropland J. Environ. Qual., June 23, 2009; 38(4): 1719 - 1727. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. A.S. Dolliver and S. C. Gupta Antibiotic Losses from Unprotected Manure Stockpiles J. Environ. Qual., May 1, 2008; 37(3): 1238 - 1244. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
