• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Antimicrobial Contaminant Removal By Multistage Slow Sand Filtration

Posted on: Thursday, 5 January 2006, 06:00 CST

By Rooklidge, Stephen J; Miner, J Ronald; Kassim, Tarek A; Nelson, Peter O; Howard, Ren S

In rural areas, where passive filtration of surface water is a viable option, aquatic systems can become contaminated by antimicrobials from wastewater treatment plants or diffuse pollution. This study examined slow sand filter biomass (schmutzdecke) sorption behavior and removal efficiencies in a pilot roughing and slow sand filter that was fed 0.2 mg/L of five compounds from four classes of antimicrobials. High-performance liquid chromatography/tandem mass spectrometry (HPLC MS/MS) was used to analyze aqueous antimicrobial concentrations. Schmutzdecke sorption coefficients were comparable to those previously found for soils but did not correlate well with estimates derived from octanol- water partition coefficients. Roughing filtration exhibited low removal efficiencies, but antimicrobials with high sorption coefficients can accumulate in roughing filter waste and schmutzdecke. At the end of a 14-day slow sand filtration period, tylosin was not detected in filter effluent and trimethoprim was >99% removed. Slow sand filtration, however, exhibited <25% removal of lincomycin and <4% removal of the sulfonamide class of antimicrobials. Multistage filtration is regarded as an ineffective treatment method for antimicrobials with low filter-media sorption coefficients.

Recent investigations of pharmaceutical contaminants in surface waters of the United States and Europe revealed that antimicrobials, such as synthetic and naturally derived antibiotics, were present at concentrations in the range of nanograms to micrograms per litre (Kolpin et al, 2004; Golet et al, 2002; Kolpin et al, 2002; Lindsey et al, 2001). Human and veterinary antimicrobials in the aquatic environment and the rising trend of antibacterial resistance found in aquatic and soil microorganisms have unknown consequences to the effectiveness of current drug therapies (Rooklidge, 2004). Consumption of antimicrobials below therapeutic levels has the potential to reduce the efficacy of drug regimens prescribed to treat infectious diseases (Levy, 1998), and the presence of antimicrobials in drinking water sources may enhance resistance transfer among bacteria in water distribution systems (Schwartz et al, 2003). Although pharmaceutical contaminants in drinking water are not currently regulated, the effects of chronic human exposure to trace concentrations of antimicrobials, along with the fate of these drugs in water treatment systems, remain poorly understood.

Surface water in rural areas can become contaminated by antimicrobials in effluent from wastewater treatment plants or by diffuse pollution introduced from land application of animal manure and processed biosolids that contain antimicrobial residues. Water- soluble antimicrobials are used in livestock and poultry production to promote growth and avoid bacterial infections that spread easily, and waste streams from these operations are potential routes of environmental contamination. A study of water samples from a western US river showed that concentrations of the sulfonamide class of antimicrobials eventually diminished downstream of urban centers that discharge wastewater effluent, but tetracycline antimicrobials increased to greater than 0.9 g/L as the river flowed through agricultural areas (Yang & Carlson, 2003). Tetracycline concentrations of 1 g/L were found in springs and streams near US poultry farms (Campagnolo et al, 2002), and agricultural practices were also assumed to be the source of 0.25 g/L concentrations of lincomycin found in river water downstream of rural communities in Italy (Calamari et al, 2003). Small water treatment systems that rely on contaminated sources may not use the oxidation or adsorption processes employed in larger municipal water treatment facilities. These processes have been shown to be effective treatment methods for many pharmaceutical contaminants (Huber et al, 2003).

Antimicrobials sorbed to schmutzdecke waste may be partially biodegraded during a filtration period, but land application of drinking water waste products may introduce a source for diffuse pollution in previously unexposed regions.

Rural areas of North America and developing nations experienced a resurgence of slow sand filtration during the past century. Logsdon et al (2002) have reviewed operational and design criteria, extent of filtration use, and recent research intended to expand the applicability of the treatment method. Roughing filter pretreatment of source water (multistage filtration) removes suspended solids before slow sand filtration and prolongs the filtration cycle during seasonal periods of poor water quality (Wegelin, 1996). Small communities with available land are the primary benefactors of multistage filtration because of the passive nature of the treatment process, which relies on sedimentation, sorption, biodegradation, and predation within the filter media. Although multistage filtration is an effective treatment method for removal of pathogenic microorganisms and suspended paniculate matter, the unit processes involved were not specifically designed to remove anthropogenic organic contaminants.

FIGURE 1 Pilot roughing filter section diagram

TABLE 1 Roughing filter media size ranges

An investigation of herbicide contaminant behavior in pilot slow sand filters indicated that extensive microbial degradation occurred within the first eight days of filtration, but sorption within the filter column was not significant (Woudneh et al, 1997). Removal of several pharmaceuticals by municipal slow sand filtration was regarded as inconclusive (Ternes et al, 2002), but antimicrobial removal has yet to be investigated in many types of water treatment systems. Laboratory studies of conventional drinking water treatment processes showed that sulfonamides were effectively removed by carbon adsorption and oxidation (Adams et al, 2002); however, these treatment techniques may have limited feasibility in developing nations and rural areas because of cost and maintenance considerations.

This study examined roughing and slow sand filter removal efficiencies for five compounds from four classes of antimicrobials identified as contaminants in a recent US nationwide surface water investigation (Kolpin et al, 2002). The antimicrobial challenge experiment was conducted on a pilot roughing and slow sand filter system built at the municipal water treatment facility of the city of Salem, Ore., which uses slow sand filtration to treat source water from the North Santiam River. The unit processes of multistage filtration were subjected to 0.2 mg/L influent concentrations of each antimicrobial for more than 20 detention periods. The roughing filter was dosed with antimicrobials for four days, and aqueous samples were collected at influent and effluent sample ports. The mature pilot slow sand filter system was dosed with antimicrobials for 14 days, and water samples were collected from the influent, effluent, and sample ports at 10 cm (0.3 ft) intervals down the sand column every day for seven days and on the final day. Aqueous antimicrobial concentrations were analyzed using a direct- injection, highperformance liquid chromatography/tandem mass spectrometry (HPLC MS/MS) method developed for this research.

FIGURE 2 Schematic of North Santiam River pilot slow sand filter

TABLE 2 Summary of pilot-study analytical methods

Laboratory sorption studies were conducted on slow sand filter biomass (schmutzdecke) formed at the sand-water interface of the municipal system to determine antimicrobial sorption coefficients for schmutzdecke (K^sub d^) and its organic carbon (K^sub oc^) and organic matter (K^sub om^) fractions. Organic carbon and organic matter sorption coefficients have been used for many years to develop simulation models that predict partitioning behavior of organic contaminants in natural sediments (Brown & Flagg, 1981). Environmental partitioning behavior of hydrophobic chemicals can be evaluated without extensive field investigation by estimating K^sub oc^ and K^sub om^ from laboratory observations of the chemical distribution between octanol and water. Development of predictive models for antimicrobial sorption to slow sand filter media and schmutzdecke would be greatly aided by the ability to estimate sorption coefficients from the octanol-water partition coefficient (K^sub ow^), but linear relationships between antimicrobial K^sub oc^ and K^sub ow^ have not been consistently observed in other complex matrixes such as soils and manure (Loke et al, 2002; Tolls, 2001). The schmutzdecke sorption coefficients calculated in this experiment were compared with values estimated by classical empirical equations to determine the appropriateness of considering antimicrobial K^sub ow^ as a suitable predictor of schmutzdecke sorption behavior.

TABLE 3 Antimicrobials used in the pilot roughing and slow sand filter experiment

FIGURE 3 Percent of influent antimicrobial concentrations at 100- mm (4-in.) depth intervals down the sand column(day 14)

This research was designed to investigate the removal of antimicrobials in multistage filtration, provide information for rural communities affected by antimicrobial surface water pollution, identify topics for future research, and aid in environmental and human health risk assessments.

METHODS

Pilot filter construction. The pilot horizontal roughing filter was manufactured from exterior-grade plywood supported by a beam and pier-block foundation, and was sealed with epoxy at wood connections and fastener points to form a watertight box. The box containing the media was 0.6-m (2-ft) high, 1.22-m (4-ft) wide, and 2.44-m (8-ft) long. The filter had an initial layer of calcite limestone1 in the first filter cell (Figure 1), with crushed and sieved basaltic river rock2 of decreasing size ranges (Table 1) making up the remaining filter media. The final cell length ratio was 0.45:2.55:2:1. The calcite amendment was found to enhance influent clay turbidity removal and increase pH and alkalinity of multistage filter effluent (Rooklidge & Ketchum, 2002).

Before the antimicrobial challenge study began, the pilot filters were run to waste for three weeks to mature the schmutzdecke. The roughing filter influent flow rate during the study, provided by a constant-head reservoir filled from the municipal facility's source water intake structure, was 0.5 m/h (294 gal/sq ft/d). The flow rate and filter proportions were within the range suggested by Wegelin (1996), and the roughing filter was not backwashed before the challenge experiment.

The pilot slow sand filter (Figure 2) was a 3.05-m (10-ft) tall polyethylene tank with an internal diameter of 1.22 m (4 ft). The tank was fed source water at a rate of 0.15 m/h (88 gal/sq ft/d). The 1-m (3.3-ft) deep basaltic sand media^sub 2^ had a uniformity coefficient of 2.26 and an effective size of 0.30 mm. The sand column was supported by crushed river rock (0.6-m [2-ft] deep), installed in layers of decreasing average size from 20 mm (0.8 in.) at the bottom to 1.5 mm (0.06 in.) at the sand-gravel interface. The pilot filter was a vertical fullscale filter representing the same depth and media characteristics of the city of Salem's municipal slow sand filters, and the flow rate was within the range suggested by Galvisetal (1998).

The slow sand filter was regarded as mature when effluent turbidity fell below 1 ntu, the supernatant level reached 25 cm (0.8 ft) above the schmutzdecke, effluent total coliform was <2 cfu/100 mL, and Escherichia coli was not detectable. Water quality characteristics were determined in grab samples during the filtration period using the methods listed in Table 2.

Antimicrobial sampling and analysis. The representative antimicrobials used as analytes for this study (Table 3) included analytical-grade3 veterinary medications: sulfamethazine (SMZ) and tylosin (TYL); drugs prescribed for human use: sulfamethoxazole (SMX) and trimethoprim (TRI); and lincomycin (LIN), a dual-use antimicrobial. The daily injected mass of analytes was dissolved in 200 mL methanol and diluted in 4 L (1.06 gal) of distilled water brought to neutral pH with sodium hydroxide. Analytes were injected into the filter influent pipe by peristaltic pump from clean amber glass bottles to provide a 0.2 mg/L final influent concentration of each antimicrobial. The total antimicrobial concentration fed to the pilot system (1 mg/L) was not considered detrimental to slow sand filter biological activity and treatment performance because concentrations did not exceed inhibitory limits of the aerobic, primarily gram-negative bacteria of schmutzdecke microbial populations (Ingerslev et al, 2001).

TABLE 4 HPLC MS/MS parameters for analytes and internal standard

TABLE 5 Calculated schmutzdecke sorption coefficients and estimated coefficients using K^sub ow^ for five antimicrobials

Aqueous samples for chromatographic analysis were collected during the challenge study in 40-mL clean amber glass vials, amended with 3 mM sodium azide (WeberShirk & Dick, 1997) and stored at 4C until analyzed. Duplicate samples of 1.5 mL from each vial were centrifuged at 10^sup 5^ g for 20 min, and 1 mL of the supernatant was transferred to HPLC autosampler vials. Sulfadimethoxine (SDM) was added to each HPLC vial as a 20 pg/L internal standard.

A 3-pm, 4.6 150 mm column4 was used with an HPLC system and mass spectrometer.5 The sample injection volume was 350 pL at a mobile- phase flow rate of 0.35 mL/min. The mobile phase consisted of two components: 2-propanol (A) and distilled water with 0.6% formic acid (B) added for a pH of 2.5. The mobile phase had an initial v/v concentration of 10% A and 90% B. The mobile-phase gradient was increased to 100% A over 13 min, kept at 100% for 2 min, decreased to 10% in 3 min, with an additional 7 min of equilibration time at the initial concentration. Quantitation was performed using integrating software,6 which multiplied the peak area of analyte by the concentration/area ratio of the internal standard.

Limits of detection (LOD) and quantitation (LOQ) for all analytes were calculated by the signal-to-noise ratio (S/N) from triplicate, six-point calibration curves made with centrifuged river water. Table 4 lists the LOD for all compounds at an S/N of 3:1, LOQ at an S/N of 10:1, and multiple reaction-monitoring MS parameters. Blanks and water for the calibration curves were made by spiking centrifuged river water collected on each day of the study. Antimicrobial concentrations were calculated from six-point linear calibration curves ranging from 0.1 to 500 pg/L. Recovery from standards made in river water ranged from 93 to 110%, and the coefficients of determination (R^sup 2^) for the daily calibration curves were >0.985.

Sorption study. Schmutzdecke was collected from the municipal filter by pressing clean glass petri dishes into the drained sand column and removing the top 1 cm (0.4 in.) of material. After maturation in the municipal slow sand filter, the mass fraction of organic carbon in the schmutzdecke (f^sub oc^) was 0.4% and the fraction of organic matter (f^sub om^) was 1.54%.

Schmutzdecke sorption studies of each antimicrobial were performed by placing 5 g of air-dried solid in 50-mL centrifuge tubes with 30 mL of river water containing 3 mM of sodium azide. Samples were shaken at 200 rpm for 24 h and centrifuged. The aqueous antimicrobial concentrations were then immediately analyzed with HPLC MS/MS. Calibration curves were generated as before, using centrifuged river water from a control vial containing only schmutzdecke and sodium azide.

The sorption study river water collected in March 2004 had a pH of 7.2 and was not adjusted during the sorption experiments. The final pH of shaken schmutzdecke samples ranged from 7.1 to 7.6. Sorption coefficients derived for this pH range were regarded as representative of the studied source water because the North Santiam River fluctuates annually between pH 6.8 and pH 7.6 (Wujcik, 2004). Analysis of the blank river water and filter effluent before the challenge experiment indicated that none of the antimicrobials were present in detectable concentrations.

TABLE 6 Mean and 95% confidence intervals of Santiam River water quality parameters from grab samples collected during the antimicrobial challenge study

TABLE 7 Roughing filter antimicrobial removal efficiencies calculated for each day during the four-day challenge experiment

The schmutzdecke sorption coefficient K^sub d^ was calculated for all anaIytes using Eq 1, in which C^sub s^ is the milligram of sorbed antimicrobial per kilogram of solid and C^sub e^ is the aqueous antimicrobial concentration (milligrams per litre) after 24 h of equilibration. Isotherms were created by analyzing duplicate equilibrated samples at five analyte concentrations (100-500 g/L). Each K^sub d^ was derived from the slope of a linear regression of data forced through the origin, and linearity was evaluated by R^sup 2^ calculated from the regressed data.7

RESULTS AND DISCUSSION

Sorption comparison. Sorption of all five antimicrobials reached equilibrium within 24 h (data not presented). The regression of sorption data (Table 5) exhibited significant linearity within the studied range of concentrations, and schmutzdecke K^sub s^ values were well represented by Eq 1. Schmutzdecke antimicrobial sorption followed the order TYL > TRI > LIN > SMX > SMZ, and the results for TYL were comparable to coefficients found for soil (Rabolle & Spliid, 2000) and manure (Loke et al, 2002). The results for the sulfonamides SMX and SMZ indicated that sorption to schmutzdecke was the same low order of magnitude as that found for sandy and clay loam soils (Boxall et al, 2002). Therefore, a potential for high sulfonamide mobility in sand filtration systems could be expected.

A comparison of the calculated and estimated sorption parameters listed in Table 5 revealed that octanol-water partition coefficients and empirical equations of hydrophobic compounds are not suitable descriptors of antimicrobial sorption to schmutzdecke. The values of log K^sub oc^ and log K^sub om^ estimated from empirical Eqs 2 and 3 did not significantly approximate the calculated values. The order of K^sub oc^ and K^sub om^ estimated from K^sub ow^ did not correlate with the calculated sorption order, and it was evident that the sorption behavior could not be explained by purely hydrophobic partitioning. As a result, schmutzdecke sorption coefficients estimated from K^sub ow^ are unlikely to be suitable predictors for models designed to simulate antimicrobial sorption behavior in slow sand filtration.

TABLE 8 Slow sand filter antimicrobial removal efficiencies calculated during the 14-day challenge experiment

Roughing filtration. Table 6 shows the mean and 95% confidence intervals of influent water quality parameters from grab samples collected during the pilot study filtration period. A significant tr\end of increasing removal efficiency was not observed for any of the antimicrobials during the roughing filter study period (Table 7). The removal efficiencies of roughing filtration exhibited the same order as schmutzdecke sorption. This behavior was not surprising, because the schmutzdecke was primarily composed of sand filter media processed from the same material as the basaltic gravel media of the roughing filter.

Roughing filtration is regarded as a physical removal system for particulate matter that has an effect on slow sand filtration, and biological growth within the gravel media is not generally considered a relevant mechanism for contaminant removal because of high filtration rates and low detention periods. However, removal of antimicrobials with higher sorption coefficients, such as TYL, indicated that contaminants might sorb to particulate matter deposited within the interstitial voids of the filter media. Consequently, further research is necessary to investigate whether sorbed contaminants accumulate in roughing filters and could be reintroduced to the environment when the filter media is washed and the waste spread on agricultural soils.

Slow sand filtration. No significant trends of increasing removal were observed for any of the antimicrobials during the slow sand filter challenge study. Increased antimicrobial removal efficiency would be expected if the antimicrobials were readily degraded by microbial populations within the filter media (Woudneh et al, 1997). Schmutzdecke maturation appeared not to be affected during the study because filter head loss increased from 25 to 50 cm (0.8 to 1.6 ft) within 14 days.

Slow sand filter removal efficiencies of the studied antimicrobials (Table 8) demonstrated that average removal efficiencies decreased in a manner comparable to the batch sorption order of schmutzdecke (TYL > TRI > LIN > SMX > SMZ). TYL was regarded as completely removed during slow sand filtration because the antimicrobial was not detected in the filter effluent. TRI exhibited >99% removal, LIN showed <25% removal, and slow sand filtration removed <4% of the influent sulfonamide concentrations by the last day of the study period.

The low removal efficiencies of SMX and SMZ in unit processes of multistage filtration suggested that additional treatment processes are necessary to eliminate sulfonamide contaminants from drinking water. Sulfonamides are used in veterinary and human therapeutic applications and residues in the range of nanograms to milligrams per litre have been detected in land-applied manure (Haller et al, 2002), wastewater treatment plant effluent (Hartig et al, 1999), groundwater (Lindsey et al, 2001; Sacher et al, 2001), and surface water (Hirsch et al, 1998). As a result, sulfonamides should be considered as indicators for source waters suspected of veterinary and human pharmaceutical contamination because of their observed persistence in aquatic systems and lack of significant removal by conventional and multistage drinking water treatment processes.

Figure 3 shows the percentages of influent aqueous concentrations at the end of the filtration period for 10cm (0.3-ft) intervals within the sand column. TYL concentrations fell below the LOQ within 40 cm (1.3 ft) of the sand column, and these results are in agreement with the Rab011e and Spliid (2000) laboratory studies that found >98% removal when aqueous TYL concentrations fell below the analytical detection limit in a 30-cm (12-in.) column of sandy soil.

Antimicrobials with the highest sorption response attach to media in the region of greatest biological activity of a slow sand filter (Ellis & Aydin, 1995) and the effects of continued contaminant exposure to slow sand filter microbial communities have yet to be investigated. Slow sand filter treatment facilities can generate >500 metric tons/yr of waste biomass for every hectare of filter area (550 tons/acre/yr), and the removed schmutzdecke can be applied to agricultural soils or washed to reclaim the sand for further use. Although antimicrobials sorbed to schmutzdecke and roughing filter waste may be partially biodegraded over a 60-90-day filtration period (Liguoro et al, 2003; Teeter & Meyerhoff, 2003), land application of drinking water waste products may introduce a source for antimicrobial terraccumulation and diffuse pollution in regions previously unexposed to this pressure (Rooklidge, 2004).

The effect and significance of antimicrobial residues on antibacterial resistance formation among soil microorganisms are difficult to assess (Onan & LaPara, 2003; Sengelov et al, 2003). By introducing antimicrobials mobilized through diffuse pollution pathways, land application of agricultural waste, wastewater biosolids, and drinking water waste products could contribute to aquatic organism resistance. Further research is needed to investigate resistance transfer in heterogeneous microbial communities of biological drinking water filtration as well as antimicrobial residue mobilization from water treatment waste products into environmental systems that may adversely affect humans and aquatic organisms.

CONCLUSIONS

Antimicrobial sorption coefficients for schmutzdecke biomass grown on basaltic sand filter media were comparable to those previously found for soils. Sorption to slow sand filter schmutzdecke did not correlate well with octanol-water partition coefficients and empirical equations derived to model hydrophobic organic contaminant behavior in natural sediments. Therefore, K^sub ow^ is unlikely to be a suitable predictor for simulation models of slow sand filter antimicrobial removal.

Roughing filtration exhibited low removal efficiencies for antimicrobials with low schmutzdecke sorption coefficients, but antimicrobials with high sorption coefficients, such as TYL, may accumulate in paniculate matter deposited within the roughing filter. Roughing filter waste may eventually be applied to agricultural soils and should be investigated further as a potential source for diffuse pollution of the aquatic environment.

Slow sand filtration exhibited antimicrobial removal efficiencies in the order of TYL > TRI > LIN > SMX > SMZ. Tylosin was not detected in the effluent during the filtration period, and the antimicrobial was primarily sorbed within the top 40 cm (1.3 ft) of the sand filter column. This result supported earlier findings from laboratory-scale soil column studies.

At the end of the 14-day study period, slow sand filtration exhibited >99% removal of TRI, <25% removal of LIN, and <4% removal of the sulfonamide class of antimicrobials from contaminated river water. Removal results for LIN and the sulfonamides were consistent with schmutzdecke sorption behavior. Multistage filtration is regarded as an ineffective treatment method for antimicrobials with low filter-media-sorption coefficients.

Sulfonamides may be suitable indicators for suspected pharmaceutical contamination of source water used for drinking water treatment because of filtration behavior and a greater relative environmental persistence among classes of antimicrobial contaminants.

The antimicrobial contaminant behavior observed within a biologically active slow sand filter calls for continued research to investigate antibacterial resistance transfer in biological filtration systems. Further research into antimicrobial contamination and residue mobility in land-applied schmutzdecke is also warranted for adequate environmental risk assessment.

ACKNOWLEDGMENT

The authors acknowledge editorial suggestions from Diana M. Fantov and the anonymous reviewers of this article. The water treatment operators of the city of Salem, Ore., provided pilot filter assistance. Antimicrobial sample analysis was conducted in the Oregon State University (OSU, Corvallis) laboratory of Jennifer A. Field, with assistance from Melissa M. Schultz. Partial funding of this research was furnished by M. Robin Collins of the University of New Hampshire Water Treatment and Technology Assistance Center and the US Environmental Protection Agency under grant number X827736-01-0. Conclusions expressed in this article do not necessarily reflect those of the funding agencies.

If you have a comment about this article, please contact us at journal@awwa.org.

Aquatic systems, particularly in rural areas, may be contaminated by antimicrobials from wastewater treatment effluent or diffuse pollution. Recent investigations of pharmaceutical contaminants in surface waters in the United States and Europe revealed the presence of antimicrobials such as synthetic and naturally derived antibiotics used in human and veterinary medicine. The fate of these antimicrobials in the aquatic environment, along with the rising trend of antibacterial resistance in aquatic and soil microorganisms, is poorly understood.

This research was designed to investigate the removal of antimicrobials in multistage filtration, provide information for rural communities affected by antimicrobial surface water pollution, identify topics for future research, and aid environmental and human health risk assessments. The study examined schmutzdecke sorption behavior and removal efficiencies in a pilot roughing and slow sand filter fed 0.2 mg/L of five compounds from four classes of antimicrobials. Aqueous antimicrobial concentrations were analyzed by high-performance liquid chromatography/tandem mass spectrometry.- RSH

FOOTNOTES

1 Ash Grove Cement Co., Portland, Ore.

2 Morse Brothers, Stayton, Ore.

3 MP Biomedicals, Inc., Aurora, Ohio

4 Luna C8(2), Phenomenex, Torrance, Calif.

5 52690 HPLC System, Quattro Micro mass spectrometer, Waters Corp., Milford, Mass.

6 MassLynx 4.0, Waters Corp., Milford, Mass.

7 Excel 2002, Microsoft, Redmond, Wash.

8 Syracuse Research Corp., (1999-2004).

REFERENCES

Adams, C. et al, 2002. Removal of Antibiotics From Surface and Distilled Water in Conventional Water Treatment Processes. Jour. Envir. Engrg., 128:3:\153.

Boxall, A. et al, 2002. The Sorption and Transport of a Sulphonamide Antibiotic in Soil Systems. Toxicol. Letters, 131:19.

Briggs, G. 1981. Theoretical and Experimental Relationships Between Soil Adsorption, Octanol-Water Partition Coefficients, Water Solubilities, Bioconcentration Factors, and the Parachor. Jour. Agric. & Food Chem., 29:1050.

Brown, D. & Flagg, E., 1981. Empirical Prediction of Organic Pollutant Sorption in Natural Sediments. Jour. Envir. Qual, 10:3:382.

Calamari, D. et al, 2003. Strategic Survey of Therapeutic Drugs in the Rivers Po and Lambro in Northern Italy. Envir. Sci. & Technol., 37:1241.

Campagnolo, E. et al, 2002. Antimicrobial Residues in Animal Waste and Water Resources Proximal to Large-scale Swine and Poultry Feeding Operations. Sci. Total Envir., 299:89.

Ellis, K. & Aydin, M., 1995. Penetration of Solids and Biological Activity Into Slow Sand Filters. Water Res., 29:5:1333.

Galvis, G. et al, 1998. Multi-Stage Filtration: An Innovative Water Treatment Technology. Technical Paper 34-E, IRC International Water and Sanitation Centre, The Hague, Netherlands.

Golet, E. et al, 2002. Environmental Exposure and Risk Assessment of Fluoroquinolone Antibacterial Agents in Wastewater and River Water of the Glatt Valley Watershed, Switzerland. Envir. Sci. & Technol., 36:17:3645.

Haller, M. et al, 2002. Quantification of Veterinary Antibiotics (Sulfonamides and Trimethoprim) in Animal Manure by Liquid Chromatography-Mass Spectrometry. Jour. Chromatogr. A., 952:111.

Hartig, C. et al, 2001. Membrane Filtration of Two Sulphonamides in Tertiary Effluents and Subsequent Adsorption on Activated Carbon. Water Res., 35:16:3998.

Hirsch, R. et al, 1998. Determination of Antibiotics in Different Water Compartments Via Liquid Chromatography-Electrospray Tandem Mass Spectrometry. Jour. Chromatogr. A., 815:213.

Huber, M. et al, 2003. Oxidation of Pharmaceuticals During Ozonation and Advanced Oxidation Processes. Envir. Sci. & Technol., 37:1016.

Ingerslev, F. et al, 2001. Primary Biodegradation of Veterinary Antibiotics in Aerobic and Anaerobic Surface Water Simulation Systems. Chemosphere, 44:865.

Karickhoff, S. et al, 1979. Sorption of Hydrophobic Pollutants on Natural Sediments. Water Res., 13:241.

Kolpin, D. et al, 2004. Urban Contribution of Pharmaceuticals and Other Organic Wastewater Contaminants to Streams During Differing Flow Conditions. Sci. total Envir:, 328:119.

Kolpin, D. et al, 2002. Pharmaceuticals, Hormones, and Other Organic Wastewater Contaminants in U.S. Streams, 1999-2000: A National Reconnaissance. Envir. Sci. & Technol., 36:1202.

Levy, S., 1998. The Challenge of Antibiotic Resistance. Sci. Amer., 278:46.

Liguoro, M. et al, 2003. Use of Oxytetracycline and Tylosin in Intensive Calf Farming: Evaluation of Transfer to Manure and Soil. Chemosphere, 52:203.

Lindsey, M. et al, 2001. Analysis of Trace Levels of Sulfonamide and Tetracycline Antimicrobials in Groundwater and Surface Water Using Solid-phase Extraction and Liquid Chromatography/Mass Spectrometry. Anal. Chem., 73:4640.

Logsdon, G. et al, 2002. Slow Sand Filtration for Small Water Systems. Jour. Envir. Engrg. & Sci., 1:339.

Loke, M. et al, 2002. Determination of the Distribution Coefficient (log K^sub d^) of Oxytetracycline, Tylosin A, Olaquindox and Metronidazole in Manure. Chemosphere, 48:351.

Onan, L. & LaPara, T, 2003. Tylosin-resistant Bacteria Cultivated from Agricultural Soil. FEMS Toxicol. Letters, 220:15.

Rablle, M. & Spliid, N., 2000. Sorption and Mobility of Metronidazole, Olaquindox, Oxytetracycline, and Tylosin in Soil. Chemosphere, 40:715.

Rooklidge, S., 2004. Environmental Antimicrobial Contamination FromTerraccumulation and Diffuse Pollution Pathways. Sci. Total Envir., 325:1:1.

Rooklidge, S. & Ketchum, L, 2002. Calciteamended Horizontal Roughing Filtration for clay Turbidity Removal. Jour. Water Supply Res. & Technol.-Aqua, 51:6:333.

Sacher, F. et al, 2001. Pharmaceuticals in Groundwaters: Analytical Methods and Results of a Monitoring Program in Baden- Wrttemberg, Germany. Jour. Cnromafogr.A938:199.

Schwarz, T. et al, 2003. Detection of Antibioticresistant Bacteria and Their Resistance Genes in Wastewater, Surface Water, and Drinking Water Biofilms. FEMS Microbiol. Ecol., 43:325.

Sengelov, G. et al, 2003. Bacterial Antibiotic Resistance Levels in Danish Farmland as a Result of Treatment With Pig Manure Slurry. Envir. Intl., 28:587.

Standard Methods for the Examination of Water and Wastewater, 1998. (20th ed.). APHA, AWWA, and WEF, Washington.

Syracuse Research Corp., 1999-2004. Environmental Science Database. http://www.syrres.com/esc/kowwin.htm, accessed July 7, 2004.

Teeter, J. & Meyerhoff, R., 2003. Aerobic Degradation of Tylosin in Cattle, Chicken, and Swine Excreta. Envir. Res., 93:1:45.

Ternes, T. et al, 2002. Removal of Pharmaceuticals During Drinking Water Treatment. Envir. Sci. & Technol., 36:3855.

Tolls, J. 2001. Sorption of Veterinary Pharmaceuticals in Soils: A Review. Envir. Sci. & Technol., 35:17:3397.

Weber-Shirk, M. & Dick, R., 1997. Biological Mechanisms in Slow Sand Filters. Jour. AWWA, 89:2:72.

Wegelin, M., 1996. Surface Water Treatment by Roughing Filters. Swiss Centre for Development Cooperation in Technology and Management, Gallen, Switzerland.

Woudneh, M. et al, 1997. The Behavior of 2,4D as it Filters Through Slow Sand Filters. Jour. Water Supply Res. & Technol.-Aqua, 46:3:144.

Wujcik, H., 2004. Personal communication of city of Salem water operator with the authors.

Yang, S. & Carlson, K., 2003. Evolution of Antibiotic Occurrence in a River Through Pristine, Urban, and Agricultural Landscapes. Water Res., 37:4645.

ABOUT THE AUTHORS

Stephen J. Rooklidge (to whom correspondence should be addressed) is an engineer with ECO.-LOGIC Engineering, 777 N. Pershing Ave., Stockton, CA 95203; e-mail rooklidge@alumni.nd.edu. He has a bachelor's degree in chemical engineering from Oregon State University, Corvalis (OSU); a master's, degree in environmental engineering from the University of Notre Dame in Notre Dame, Ind., and a PhD in bioresource engineering from OSU. An AWWA member, he received a 1999 AWWA LARS Scholarship and second place in the 2002 AWWA Master's Thesis Academic Achievement Award competition. J. Ronald Miner was a professor of bioengineering at OSU for 31 years but passed away during the preparation of this article. Tarek A. Kassim is in the OSU Department of Bioengineering, and Peter O. Nelson is with the OSU Department of Civil, Construction and Environmental Engineering.

Copyright American Water Works Association Dec 2005

Source: American Water Works Association. Journal

More News in this Category