APPLID AND ENVIRONMENTAL MICROBIOLOGY, May 1976, p. 705-710
Copyright © 1976 American Society for Microbiology
Vol. 31, No. 5 Printed in U.SA.
Airborne Coliphages from Wastewater Treatment Facilities K. F. FANNIN,* J. C. SPENDLOVE, K. W. COCHRAN, AND J. J. GANNON Departments of Environmental and Industrial Health and of Epidemiology,* School of Public Health, The University of Michigan, Ann Arbor, Michigan 48109, and Dugway Proving Ground, Dugway, Utah 84022 Received for publication 12 January 1976
The emission (from wastewater treatment plants) of airborne coliphages that form plaques on two strains ofEscherichia coli was investigated. Two activatedsludge and two trickling-filter plants were studied. Field sampling procedures used large-volume air samplers with recirculation devices. Coliphages were enumerated by a most-probable-number (MPN) procedure. Temperature, relative humidity, windspeed, and presence of sunlight were monitored. Concurrent samples of sewage were taken during each air-sampling run. Average coliphage levels in the airborne emissions of trickling-filter beds and activated-sludge units were 2.84 x 10-l and 3.02 x 10-' MPN/m3, respectively, for all positive observations, and sewage liquor concentrations from the sources were 4.48 x 105 and 2.94 x 106 plaque-forming units/liter, respectively, depending upon the E. coli host used for assay. This work establishes minimal airborne-coliphage concentrations from the plants studied. The procedures employed will be useful in evaluating the animal virus levels in these emissions. Initial work on the airbome transmission of viruses was concerned with influenza and other respiratory pathogens (11, 20, 29). However, increasing attention has recently been directed to the possible aerosol spread of nonrespiratory viruses that resulted in infections. Rabies virus, for example, is transmissible by air (8). This route of spread has also been implicated in an epidemic of infectious hepatitis among children who had ridden together on a schoolbus (1). Long lists of laboratory-acquired infections, many caused by aerosolized viruses, are available. Recent studies show a number of known oncogenic viruses to be transmissible via aerosols. Yaba tumor virus can -be carried by aerosols to monkeys (30). Aerosol-transmitted infections with other oncogenic viruses have also been reported (17). The occurrence and survival of viruses in water and wastewater have been the subject of an increasing number of recent investigations (5) focusing upon the possible waterborne transmission of the enteric viruses (those which are excreted through human feces). These viruses include the polio; coxsackie; echo; reo; adeno; infectious hepatitis, and other recently identified viruses. Although many outbreaks of infectious hepatitis spread via water have been reported (21), the inability of investigators to isolate and confirm the causative agent has prevented direct water-related studies (19). Since initial isolations of poliomyelitis virus from sewage (22), viruses of fecal origin have
been repeatedly isolated from wastewater treatment plants. The treatment of wastewater by activated sludge, trickling filters, and spray irrigation of land releases into the air certain bacteria that can be recovered at various distances downwind from the site of treatment (12, 23, 24). These bacteria, many of which are pathogenic, can be airborne for at least 0.8 miles (about 0.13 km) (2). Although it has been shown that bacteria are transmissible by aerosols from sewage plants and that certain viruses can travel very long distances under varied weather conditions and may initiate infections at points far from their sources (26), a literature review indicates that the potential health hazards to populations exposed to airborne viruses from wastewater treatment facilities have not been successfully evaluated. The purpose of this study is to develop and apply procedures for evaluating emissions of airborne viruses from wastewater treatment facilities. Since high concentrations of coliphages occur in sewage (relative to animal viruses) (16), these procedures were applied for their airborne isolation. The methods used in this investigation may, with slight modification, be applied to evaluating airborne animal virus emissions in similar environments. (This paper was taken in part from a dissertation submitted by K.F.F. in partial fulfillment of the requirements for the Ph.D degree at The University of Michigan, Ann Arbor.) 705
706
FANNIN ET AL.
APPL. ENVIRON. MICROBIOL.
Sampling fluid was kept on ice during sampling and until initial processing. Wind direction was obField-sampling locations. Four wastewater treat- served, and velocity was measured with a hand-held ment plants were selected for sampling. Plant no. 1, anemometer manufactured by Davis Instrument a trickling-filter plant, was located in a metropoli- Manufacturing Co., Baltimore, Md. Relative hutan region of northern Utah. This plant operated at midity (RH) was monitored with a battery-operated approximately 13 million gallons per day (MGD). psychrometer, model no. 566, manufactured by The Plant no. 2 was an activated-sludge plant located in Bendix Corp., Friez Instrument Division, Baltisoutheastern Michigan. Operation exceeded its orig- more. Air temperature, sky condition, precipitation, inal design capacity of 15 MGD. Here, during the and daylight or darkness were also noted. Liquid sewage samples from plants 1 and 4 were study period, ferric chloride was added early in the treatment process and unsettled portions of the sew- taken from the boom of the trickling-filter bed. Liqage were recirculated. Plant no. 3 was also an acti- uid samples from plants 2 and 3 were taken from the vated-sludge plant located in southeastern Michi- activated-sludge unit. All liquid samples were taken gan. This plant had a sewage flow rate of approxi- from the source immediately upwind from the point mately 8.5 MGD, with 4.0 and 4.5 MGD, respec- of air sampling. Liquid samples were pooled tively, distributed through each of two activated- throughout each air-sampling run. Sample processing. Airborne samples were mixed sludge subplants operated in parallel. Samples were taken from the 4.5-MGD subplant, in which sewage thoroughly with a Vortex mixer and then centriwas subjected to primary treatment before aeration. fuged for 30 min at 940 x g at 4 C to remove gross Plant no. 4 was a trickling-filter plant located in particulate matter. The fluid was then decanted into southeastern Michigan, with a sewage flow rate of sterile tubes and the pH was adjusted to approximately 9.0, using 0.1 N NaOH. Using a 20- to 50-ml 1.5 MGD. Field-sampling procedures. Airborne samples syringe with an attached long 15-gauge pipetting from the Utah plant (plant no. 1) were taken with a cannula, the sample was aspirated. A Millipore Multi Slit Impinger (MSI) manufactured by Envi- Swinnex filter holder, containing a 25-mm-diameter ronmental Research Corporation, St. Paul, Minn. filter with a 0.45-,m pore size, that had been preSamples were taken from the Michigan plants treated with 1 ml of heat-inactivated (30 min at (plants 2, 3, and 4) with large-volume air samplers 56 C) fetal calf serum was attached to this syringe, (LVAS), manufactured by Litton Systems, Inc., and the sample was filtered into sterile tubes. The Minneapolis, Minn. The efficiency of these samplers pH of each tube was adjusted (7.0 to 7.2) with 0.1 N is discussed by Decker et al. (9). The sampling fluid HCI. Liquid sewage samples were blended at high used with the MSI was 35 ml of Hanks balanced salt solution plus 10% heat-inactivated fetal calf serum, speed in a Waring blender for 4 min and then centriplus 0.002 M tris(hydroxymethyl)aminomethane fuged at 940 x g for 30 min at 4 C to remove gross buffer. The sampling fluid was modified for use with particulate matter. Ten milliliters of the supernathe LVAS. This fluid, referred to as phosphate sam- tant was filtered through the Millipore filter arpling fluid, consisted of 30 ml of phosphate-buffered rangement described above. To avoid dilution with saline, plus 2% of a 1% phenol red solution, plus any serum retained in the filter, the 1st ml of filtrate 0.03% GE Antifoam 10, plus 2% heat-inactivated was discarded. fetal calf serum, and had a pH of approximately 7.2. Sample assay. Airborne coliphages were isolated Fluid flow rates were 5 ml/min with the MSI and 7 and enumerated by a most-probable-number (MPN) ml/min with the LVAS. Collected material was con- procedure as described by Kott (15), with slight centrated by passing the fluid through a recircu- modification. Ten tubes, each containing 10 ml of lation apparatus. Double-distilled water was added phage assay broth, were used for each sample with to replace water lost by evaporation. The air sam- each bacterial host. To each tube, 0.5 ml of processed pling rate was 1,000 liters/min, and sampling pe- sample was added. Bacterial hosts used were Escheriods were 35 min with the MSI and 30 to 120 min richia coli C3000 (obtained from Geoffrey Orr, Dugwith the LVAS. way Proving Ground, Utah) and E. coli K-12 HfrD Air was sampled from the edge of trickling-filter (obtained from Rex Spendlove, Utah State Univerand activated-sludge units 2 to 15 m downwind from sity). A phage-resistant mutant ofE. coli 11303 (obthe source, with most samples taken within 5 m of tained from G. Orr) was used in early samples as a the source. All air samples were taken from mobile control against certain bacteria, such as Bdellovibcarts at a height of approximately 1.5 m. Control air rio, that may be filterable and produce lysis on E. samples were taken upwind from the source. Precise coli cells. To each of the inoculated tubes was added sampling locations depended upon wind direction 0.1 ml of a 4-h culture of one of the above E. coli and physical accessibility. Between sampling runs, strains. The racks of tubes were vigorously shaken the samplers and attached recirculation appara- and incubated for 18 to 24 h at 37 C. For each set of tuses were washed in turn with 500 ml of an Alconox inoculated sample tubes, a lawn of one of the host detergent solution, 1,000 ml of distilled water, and strains was prepared on a phage assay base agar 500 ml of double-distilled water. Between individual plate, which had been incubated overnight at 37 C. air samples during each sampling run, the MSI, The bacterial lawn was prepared by adding 0.5 ml of LVAS, and attached recirculation devices were a 4-h host culture to 2.5 ml of melted phage assay washed with 500 ml of sterile distilled water fol- base agar containing 0.7% agar. This suspension lowed by 500 ml of sterile double-distilled water. was mixed and poured onto a phage assay base plate MATERIALS AND METHODS
VOL. 31, 1976
AIRBORNE COLIPHAGES
and evenly distributed over it. Excess surface moisture of the seeded lawn was permitted to dry, and a drop from each of the 10 tubes per sample per host strain was placed on the plate, using sterile wooden applicator sticks. After the spots dried, the plates were incubated at 37 C for 5 to 12 h and then examined for bacterial lysis at each of the inoculated areas. Clear zones in the spotted areas were interpreted as coliphage lysis and reported as positive for isolation of phage to which the particular host strain is susceptible. Results were reported as MPN per
cubic meter. Liquid sewage samples were assayed for coliphage by the soft-agar overlay, a method described by Adams (3). Tenfold dilutions of sewage samples were made in phage assay base, and 1 ml of each dilution was plated in duplicate on phage assay base
707
agar plates. Plates were incubated for 5 to 18 h, and plaques were counted by using a Quebec colony counter. Plaque counts were recorded as plaqueforming units per liter of sewage for the particular
host strain used.
RESULTS The MSI and LVAS samplers, with fluid recirculation, were capable of the level of sensitivity required for isolating airbome coliphages emitted from wastewater treatment plants. Coliphage levels in liquid sewage varied widely among samples taken during different sampling runs and at different plants. This variation was usually reflected in the air samples taken at these times (Tables 1 and 2). Mean
TABLE 1. Airborne E. coli C3000 phage isolates related to liquid-sewage concentration no. Plant no.
1 1 1 1 2 2 2 2 3 3 3 4
4
Liquid-sewage concn
No. of air
4.44 x 105 1.60 x 105 0 NDb 9.00 X 105 1.00 X 103 0 0 5.90 x 104 3.90 x 104 7.45 x 105 8.10 x 105 8.75 x 105
3 3 3 4 5 2 2 5 3 1 4 4
(PFU/liter")
samples
4
Mean 3.36 x 105 Mean (exclud4.48 x 105 ing zero) a PFU, Plaque-forming units. b ND, Not done.
Mean airborne concn
(MPN/m3) 3.33 x 10-1 4.67 x 10-1
6.00 x 10-' 1.00 x 10-1 4.00 x 10-2 0 0 0 2.60 x 10-1
1.10 x 10B5.13 x 10-' B2.88 x 10-1 1.88 x 10-1 2.24 x 10-' 2.84 x 10-'
Mean airborne concn/
sewage concn
7.50 2.96
x x
10-7 10-';
4.44 x 10-" 0 4.41 x 10-'; 2.82 x 10-'; 6.89 x 10-7
3.56 x 10-7 2.15 x 10-7 6.67 x 10-7 6.34 x 10-7
TABLE 2. Airborne E. coli K-12 HfrD phage isolates related to liquid-sewage concentration Plant no.
1 1 1 1 2 2 2 2
3 3 3 4 4
Mean Mean (exciuding zero) a.b See Table 1.
concn Liquid-sewage (PFU/litera)
No. of air
5.46 x 105 9.00 x 104 2.40 x 107 NDb ND 1.50 x 103 0 0 2.10 x 105 1.50 x 105 4.25 x 105
3 3 3 4
5.75 x 105 4.55 x 105
2.40 x 10'i 2.94 x 106
samples
2 2 5 3 1 4 4
4
Mean airborne
concn
(MPN/m3) 3.33 x 10-' 6.67 x 10-1 2.00 x 10-1
3.00 x 10ND 0 0 0 1.70 x 10-2 0 3.88 x 10-1 ¢3.75 x 10-1 1.40 x 10-1
2.23 x 10-' 3.02 x 10-'
Mean airborne concn/ sewage concn
6.10 7.41 8.33
x x x
10-7
10-'"
10-9
0
8.10
x
10-"
0 9.13 x 10-7 6.52 x 10-7 3.08 x 10-7 9.29 1.03
x
10-"
x
10-7
708
APPL. ENVIRON. MICROBIOL.
FANNIN ET AL.
concentrations of airborne coliphages for all nonzero samples were 2.84 x 10-1 MPN/m3 on E. coli C3000 and 3.0 x 10-1 MPN/m3 on E. coli K-12 HfrD. Average ratios of mean concentrations (airborne to sewage) for these samples were 6.34 x 10-7 and 1.03 x 10-7, respectively. Coliphage levels from plant 2 were substantially lower than those observed at the other three plants (Tables 1 and 2). The conditions at plant 2 (addition of ferric chloride and recirculation) were not typical of its general method of operation. Consequently, data from this plant were not included in general comparisons of the effects of meteorological conditions on the recovery of airborne coliphages from the plants studied. Table 3 compares wind velocity with airborne-coliphage isolations. No trend is evident from examination of the table although, excluding the one sample at 6.0 to 8.9 mph, the number of isolations is higher at wind velocities of 1 mph or above than below this level. Application of the Spearman rank correlation coefficient (r,), with correction for ties, as described by Siegel (25), shows no significant correlation between windspeed and bacteriophage isolations (95% confidence levels are used for determination of significance of all statistical tests in this investigation). A comparison of airbome-coliphage isolations with temperature (Table 4) also shows little correlation. Although the number of isolations appeared to be lower at very low temperatures than at higher ones, no correlation could be observed. A stronger correlation was, however, observed with RH and airborne-coliphage isolations (Table 5). E. coli K-12 HfrD phage isolations increased significantly as RH increased. Although a significant correlation was not observed with E. coli C3000 phages, the r, value of 0.30 showed a much stronger correlation with RH than with the other meteorological parameters examined. The effects of sunlight on the recovery of TABLE 3. Airborne emission of coliphages compared to windspeed at plants 1, 3, and 4
airborne coliphages were observed by noting whether samples were taken in daylight (light), at night (dark), or at dusk or dawn (light/dark) (Table 6). Samples taken under light and dark conditions were compared by applying the Mann-Whitney U test (25). Although the differences were not significant, a larger number of coliphage isolates appeared to have been made during night or dusk/dawn than during daylight. DISCUSSION The isolation of low levels of coliphages from the airborne emissions of wastewater treatTABLE 4. Airborne emission of coliphages compared to ambient air temperature at plants 1, 3, and 4 Temp (C)
No. of samples
1.00 x 10-1 2 1.00 x 10-1 10.0-14.9 3.21 x 10-1 12 3.01 x 10-1 15.0-19.9 2.24 x 10-1 11 3.85 x 10-1 20.0-24.9 4 4.88 x 10-1 3.50 x 10-1 25.0-29.9 a r, values were 0.19 and 0.13 for C3000 and HfrD phages, respectively. t,n-2) values were, respectively, 1.01 (not significant) and 0.68 (not significant). TABLE 5. Airborne emission of coliphages compared to relative humidity at plants 1, 3, and 4 Relative Relativ
humidity
ity (mph)
No. of
samples
E. coli C3000
E. coli K-12
phage
HfrD phage
2.3 x 10-1 3.5 x 10-1 3.0 x 10-1 2.0 x 10-1 a r, values were 0.21 and 0.11 for C3000 and HfrD phages, respectively. t(,-2) values were, respectively, 1.12 (not significant) and 0.58 (not significant).
Copyright © 1976 American Society for Microbiology
Vol. 31, No. 5 Printed in U.SA.
Airborne Coliphages from Wastewater Treatment Facilities K. F. FANNIN,* J. C. SPENDLOVE, K. W. COCHRAN, AND J. J. GANNON Departments of Environmental and Industrial Health and of Epidemiology,* School of Public Health, The University of Michigan, Ann Arbor, Michigan 48109, and Dugway Proving Ground, Dugway, Utah 84022 Received for publication 12 January 1976
The emission (from wastewater treatment plants) of airborne coliphages that form plaques on two strains ofEscherichia coli was investigated. Two activatedsludge and two trickling-filter plants were studied. Field sampling procedures used large-volume air samplers with recirculation devices. Coliphages were enumerated by a most-probable-number (MPN) procedure. Temperature, relative humidity, windspeed, and presence of sunlight were monitored. Concurrent samples of sewage were taken during each air-sampling run. Average coliphage levels in the airborne emissions of trickling-filter beds and activated-sludge units were 2.84 x 10-l and 3.02 x 10-' MPN/m3, respectively, for all positive observations, and sewage liquor concentrations from the sources were 4.48 x 105 and 2.94 x 106 plaque-forming units/liter, respectively, depending upon the E. coli host used for assay. This work establishes minimal airborne-coliphage concentrations from the plants studied. The procedures employed will be useful in evaluating the animal virus levels in these emissions. Initial work on the airbome transmission of viruses was concerned with influenza and other respiratory pathogens (11, 20, 29). However, increasing attention has recently been directed to the possible aerosol spread of nonrespiratory viruses that resulted in infections. Rabies virus, for example, is transmissible by air (8). This route of spread has also been implicated in an epidemic of infectious hepatitis among children who had ridden together on a schoolbus (1). Long lists of laboratory-acquired infections, many caused by aerosolized viruses, are available. Recent studies show a number of known oncogenic viruses to be transmissible via aerosols. Yaba tumor virus can -be carried by aerosols to monkeys (30). Aerosol-transmitted infections with other oncogenic viruses have also been reported (17). The occurrence and survival of viruses in water and wastewater have been the subject of an increasing number of recent investigations (5) focusing upon the possible waterborne transmission of the enteric viruses (those which are excreted through human feces). These viruses include the polio; coxsackie; echo; reo; adeno; infectious hepatitis, and other recently identified viruses. Although many outbreaks of infectious hepatitis spread via water have been reported (21), the inability of investigators to isolate and confirm the causative agent has prevented direct water-related studies (19). Since initial isolations of poliomyelitis virus from sewage (22), viruses of fecal origin have
been repeatedly isolated from wastewater treatment plants. The treatment of wastewater by activated sludge, trickling filters, and spray irrigation of land releases into the air certain bacteria that can be recovered at various distances downwind from the site of treatment (12, 23, 24). These bacteria, many of which are pathogenic, can be airborne for at least 0.8 miles (about 0.13 km) (2). Although it has been shown that bacteria are transmissible by aerosols from sewage plants and that certain viruses can travel very long distances under varied weather conditions and may initiate infections at points far from their sources (26), a literature review indicates that the potential health hazards to populations exposed to airborne viruses from wastewater treatment facilities have not been successfully evaluated. The purpose of this study is to develop and apply procedures for evaluating emissions of airborne viruses from wastewater treatment facilities. Since high concentrations of coliphages occur in sewage (relative to animal viruses) (16), these procedures were applied for their airborne isolation. The methods used in this investigation may, with slight modification, be applied to evaluating airborne animal virus emissions in similar environments. (This paper was taken in part from a dissertation submitted by K.F.F. in partial fulfillment of the requirements for the Ph.D degree at The University of Michigan, Ann Arbor.) 705
706
FANNIN ET AL.
APPL. ENVIRON. MICROBIOL.
Sampling fluid was kept on ice during sampling and until initial processing. Wind direction was obField-sampling locations. Four wastewater treat- served, and velocity was measured with a hand-held ment plants were selected for sampling. Plant no. 1, anemometer manufactured by Davis Instrument a trickling-filter plant, was located in a metropoli- Manufacturing Co., Baltimore, Md. Relative hutan region of northern Utah. This plant operated at midity (RH) was monitored with a battery-operated approximately 13 million gallons per day (MGD). psychrometer, model no. 566, manufactured by The Plant no. 2 was an activated-sludge plant located in Bendix Corp., Friez Instrument Division, Baltisoutheastern Michigan. Operation exceeded its orig- more. Air temperature, sky condition, precipitation, inal design capacity of 15 MGD. Here, during the and daylight or darkness were also noted. Liquid sewage samples from plants 1 and 4 were study period, ferric chloride was added early in the treatment process and unsettled portions of the sew- taken from the boom of the trickling-filter bed. Liqage were recirculated. Plant no. 3 was also an acti- uid samples from plants 2 and 3 were taken from the vated-sludge plant located in southeastern Michi- activated-sludge unit. All liquid samples were taken gan. This plant had a sewage flow rate of approxi- from the source immediately upwind from the point mately 8.5 MGD, with 4.0 and 4.5 MGD, respec- of air sampling. Liquid samples were pooled tively, distributed through each of two activated- throughout each air-sampling run. Sample processing. Airborne samples were mixed sludge subplants operated in parallel. Samples were taken from the 4.5-MGD subplant, in which sewage thoroughly with a Vortex mixer and then centriwas subjected to primary treatment before aeration. fuged for 30 min at 940 x g at 4 C to remove gross Plant no. 4 was a trickling-filter plant located in particulate matter. The fluid was then decanted into southeastern Michigan, with a sewage flow rate of sterile tubes and the pH was adjusted to approximately 9.0, using 0.1 N NaOH. Using a 20- to 50-ml 1.5 MGD. Field-sampling procedures. Airborne samples syringe with an attached long 15-gauge pipetting from the Utah plant (plant no. 1) were taken with a cannula, the sample was aspirated. A Millipore Multi Slit Impinger (MSI) manufactured by Envi- Swinnex filter holder, containing a 25-mm-diameter ronmental Research Corporation, St. Paul, Minn. filter with a 0.45-,m pore size, that had been preSamples were taken from the Michigan plants treated with 1 ml of heat-inactivated (30 min at (plants 2, 3, and 4) with large-volume air samplers 56 C) fetal calf serum was attached to this syringe, (LVAS), manufactured by Litton Systems, Inc., and the sample was filtered into sterile tubes. The Minneapolis, Minn. The efficiency of these samplers pH of each tube was adjusted (7.0 to 7.2) with 0.1 N is discussed by Decker et al. (9). The sampling fluid HCI. Liquid sewage samples were blended at high used with the MSI was 35 ml of Hanks balanced salt solution plus 10% heat-inactivated fetal calf serum, speed in a Waring blender for 4 min and then centriplus 0.002 M tris(hydroxymethyl)aminomethane fuged at 940 x g for 30 min at 4 C to remove gross buffer. The sampling fluid was modified for use with particulate matter. Ten milliliters of the supernathe LVAS. This fluid, referred to as phosphate sam- tant was filtered through the Millipore filter arpling fluid, consisted of 30 ml of phosphate-buffered rangement described above. To avoid dilution with saline, plus 2% of a 1% phenol red solution, plus any serum retained in the filter, the 1st ml of filtrate 0.03% GE Antifoam 10, plus 2% heat-inactivated was discarded. fetal calf serum, and had a pH of approximately 7.2. Sample assay. Airborne coliphages were isolated Fluid flow rates were 5 ml/min with the MSI and 7 and enumerated by a most-probable-number (MPN) ml/min with the LVAS. Collected material was con- procedure as described by Kott (15), with slight centrated by passing the fluid through a recircu- modification. Ten tubes, each containing 10 ml of lation apparatus. Double-distilled water was added phage assay broth, were used for each sample with to replace water lost by evaporation. The air sam- each bacterial host. To each tube, 0.5 ml of processed pling rate was 1,000 liters/min, and sampling pe- sample was added. Bacterial hosts used were Escheriods were 35 min with the MSI and 30 to 120 min richia coli C3000 (obtained from Geoffrey Orr, Dugwith the LVAS. way Proving Ground, Utah) and E. coli K-12 HfrD Air was sampled from the edge of trickling-filter (obtained from Rex Spendlove, Utah State Univerand activated-sludge units 2 to 15 m downwind from sity). A phage-resistant mutant ofE. coli 11303 (obthe source, with most samples taken within 5 m of tained from G. Orr) was used in early samples as a the source. All air samples were taken from mobile control against certain bacteria, such as Bdellovibcarts at a height of approximately 1.5 m. Control air rio, that may be filterable and produce lysis on E. samples were taken upwind from the source. Precise coli cells. To each of the inoculated tubes was added sampling locations depended upon wind direction 0.1 ml of a 4-h culture of one of the above E. coli and physical accessibility. Between sampling runs, strains. The racks of tubes were vigorously shaken the samplers and attached recirculation appara- and incubated for 18 to 24 h at 37 C. For each set of tuses were washed in turn with 500 ml of an Alconox inoculated sample tubes, a lawn of one of the host detergent solution, 1,000 ml of distilled water, and strains was prepared on a phage assay base agar 500 ml of double-distilled water. Between individual plate, which had been incubated overnight at 37 C. air samples during each sampling run, the MSI, The bacterial lawn was prepared by adding 0.5 ml of LVAS, and attached recirculation devices were a 4-h host culture to 2.5 ml of melted phage assay washed with 500 ml of sterile distilled water fol- base agar containing 0.7% agar. This suspension lowed by 500 ml of sterile double-distilled water. was mixed and poured onto a phage assay base plate MATERIALS AND METHODS
VOL. 31, 1976
AIRBORNE COLIPHAGES
and evenly distributed over it. Excess surface moisture of the seeded lawn was permitted to dry, and a drop from each of the 10 tubes per sample per host strain was placed on the plate, using sterile wooden applicator sticks. After the spots dried, the plates were incubated at 37 C for 5 to 12 h and then examined for bacterial lysis at each of the inoculated areas. Clear zones in the spotted areas were interpreted as coliphage lysis and reported as positive for isolation of phage to which the particular host strain is susceptible. Results were reported as MPN per
cubic meter. Liquid sewage samples were assayed for coliphage by the soft-agar overlay, a method described by Adams (3). Tenfold dilutions of sewage samples were made in phage assay base, and 1 ml of each dilution was plated in duplicate on phage assay base
707
agar plates. Plates were incubated for 5 to 18 h, and plaques were counted by using a Quebec colony counter. Plaque counts were recorded as plaqueforming units per liter of sewage for the particular
host strain used.
RESULTS The MSI and LVAS samplers, with fluid recirculation, were capable of the level of sensitivity required for isolating airbome coliphages emitted from wastewater treatment plants. Coliphage levels in liquid sewage varied widely among samples taken during different sampling runs and at different plants. This variation was usually reflected in the air samples taken at these times (Tables 1 and 2). Mean
TABLE 1. Airborne E. coli C3000 phage isolates related to liquid-sewage concentration no. Plant no.
1 1 1 1 2 2 2 2 3 3 3 4
4
Liquid-sewage concn
No. of air
4.44 x 105 1.60 x 105 0 NDb 9.00 X 105 1.00 X 103 0 0 5.90 x 104 3.90 x 104 7.45 x 105 8.10 x 105 8.75 x 105
3 3 3 4 5 2 2 5 3 1 4 4
(PFU/liter")
samples
4
Mean 3.36 x 105 Mean (exclud4.48 x 105 ing zero) a PFU, Plaque-forming units. b ND, Not done.
Mean airborne concn
(MPN/m3) 3.33 x 10-1 4.67 x 10-1
6.00 x 10-' 1.00 x 10-1 4.00 x 10-2 0 0 0 2.60 x 10-1
1.10 x 10B5.13 x 10-' B2.88 x 10-1 1.88 x 10-1 2.24 x 10-' 2.84 x 10-'
Mean airborne concn/
sewage concn
7.50 2.96
x x
10-7 10-';
4.44 x 10-" 0 4.41 x 10-'; 2.82 x 10-'; 6.89 x 10-7
3.56 x 10-7 2.15 x 10-7 6.67 x 10-7 6.34 x 10-7
TABLE 2. Airborne E. coli K-12 HfrD phage isolates related to liquid-sewage concentration Plant no.
1 1 1 1 2 2 2 2
3 3 3 4 4
Mean Mean (exciuding zero) a.b See Table 1.
concn Liquid-sewage (PFU/litera)
No. of air
5.46 x 105 9.00 x 104 2.40 x 107 NDb ND 1.50 x 103 0 0 2.10 x 105 1.50 x 105 4.25 x 105
3 3 3 4
5.75 x 105 4.55 x 105
2.40 x 10'i 2.94 x 106
samples
2 2 5 3 1 4 4
4
Mean airborne
concn
(MPN/m3) 3.33 x 10-' 6.67 x 10-1 2.00 x 10-1
3.00 x 10ND 0 0 0 1.70 x 10-2 0 3.88 x 10-1 ¢3.75 x 10-1 1.40 x 10-1
2.23 x 10-' 3.02 x 10-'
Mean airborne concn/ sewage concn
6.10 7.41 8.33
x x x
10-7
10-'"
10-9
0
8.10
x
10-"
0 9.13 x 10-7 6.52 x 10-7 3.08 x 10-7 9.29 1.03
x
10-"
x
10-7
708
APPL. ENVIRON. MICROBIOL.
FANNIN ET AL.
concentrations of airborne coliphages for all nonzero samples were 2.84 x 10-1 MPN/m3 on E. coli C3000 and 3.0 x 10-1 MPN/m3 on E. coli K-12 HfrD. Average ratios of mean concentrations (airborne to sewage) for these samples were 6.34 x 10-7 and 1.03 x 10-7, respectively. Coliphage levels from plant 2 were substantially lower than those observed at the other three plants (Tables 1 and 2). The conditions at plant 2 (addition of ferric chloride and recirculation) were not typical of its general method of operation. Consequently, data from this plant were not included in general comparisons of the effects of meteorological conditions on the recovery of airborne coliphages from the plants studied. Table 3 compares wind velocity with airborne-coliphage isolations. No trend is evident from examination of the table although, excluding the one sample at 6.0 to 8.9 mph, the number of isolations is higher at wind velocities of 1 mph or above than below this level. Application of the Spearman rank correlation coefficient (r,), with correction for ties, as described by Siegel (25), shows no significant correlation between windspeed and bacteriophage isolations (95% confidence levels are used for determination of significance of all statistical tests in this investigation). A comparison of airbome-coliphage isolations with temperature (Table 4) also shows little correlation. Although the number of isolations appeared to be lower at very low temperatures than at higher ones, no correlation could be observed. A stronger correlation was, however, observed with RH and airborne-coliphage isolations (Table 5). E. coli K-12 HfrD phage isolations increased significantly as RH increased. Although a significant correlation was not observed with E. coli C3000 phages, the r, value of 0.30 showed a much stronger correlation with RH than with the other meteorological parameters examined. The effects of sunlight on the recovery of TABLE 3. Airborne emission of coliphages compared to windspeed at plants 1, 3, and 4
airborne coliphages were observed by noting whether samples were taken in daylight (light), at night (dark), or at dusk or dawn (light/dark) (Table 6). Samples taken under light and dark conditions were compared by applying the Mann-Whitney U test (25). Although the differences were not significant, a larger number of coliphage isolates appeared to have been made during night or dusk/dawn than during daylight. DISCUSSION The isolation of low levels of coliphages from the airborne emissions of wastewater treatTABLE 4. Airborne emission of coliphages compared to ambient air temperature at plants 1, 3, and 4 Temp (C)
No. of samples
1.00 x 10-1 2 1.00 x 10-1 10.0-14.9 3.21 x 10-1 12 3.01 x 10-1 15.0-19.9 2.24 x 10-1 11 3.85 x 10-1 20.0-24.9 4 4.88 x 10-1 3.50 x 10-1 25.0-29.9 a r, values were 0.19 and 0.13 for C3000 and HfrD phages, respectively. t,n-2) values were, respectively, 1.01 (not significant) and 0.68 (not significant). TABLE 5. Airborne emission of coliphages compared to relative humidity at plants 1, 3, and 4 Relative Relativ
humidity
ity (mph)
No. of
samples
E. coli C3000
E. coli K-12
phage
HfrD phage
2.3 x 10-1 3.5 x 10-1 3.0 x 10-1 2.0 x 10-1 a r, values were 0.21 and 0.11 for C3000 and HfrD phages, respectively. t(,-2) values were, respectively, 1.12 (not significant) and 0.58 (not significant).
