Loss of infectivity of poliovirus 1 in river water under simulated field conditions

Wattr Resturth Vot It pp 1091 to 1099 {3 Pergamon Prc~s Lid 1979 Prmted m Great Brttam

004"LI3~;4 79 II01 109150200,0

LOSS OF INFECTIVITY OF POLIOVIRUS 1 IN RIVER WATER UNDER SIMULATED FIELD CONDITIONS C P CUBBAGE*,J J GANNON*, K W COCHRAN?and G W WILLIAMS~ Departments of Enwronmentai and Industrial Health*. Epldemmlogyt and Bmsta0sttcs~/, School of Pubhc Health, University of Michigan. Ann Arbor, Michtgan 48109, U S A

(Recewed 13 March 1979) Abstract--The effects of hght, vtrus coneantratmn, and turb,dtty on the rate of loss of mfcctmty (LOI) of pohowrus I were mvesUgated m two test systems, which utahzed flowing river water Two levels of each variable were used in a 23 confounded factorml design The seeded systems were sampled at regular intervals to estabhsh LOI rates Virus mfecttvtty was measured by plaque assay Loss of infectivity followed a two-component curve, an mmal, rapid phase followed by a second, slower component The slopes of the two components were examined by the analym of varmnce to determine the potentml mfluenea of each vanable Both hght and turbadRy exerted a s~gmficant influence on the LOI rate m the second component of the LOI curve and also an the transmon period between the two components, however, during the mmal rapid phase none of the varmbles influenced the LOI rate (at the 005 sagmficance level) This research demonstrates the slgmficanc¢ of hght as a wruc~dal component in the aquatic enwronment

INTRODUCTION Imtml work on the waterborne transmissmn of entenc viruses has been directed at the detection and enumeration of viral contaminants m sewage effluents and recelwng waters (Akin et al, 1971) More recently, attention has been directed toward the behavior of enteric viruses in the aquatic environment (Berg, 1976, Berg, 1977, Carlson et al, 1968, Floyd & Sharp, 1977, Gerba & Schalberger, 1975a, Schalberger, 1975b) Current wastewater treatment techmques do not completely remove viral hazards from the effluents of treatment faohtles (Berg, 1976, Berg, 1977) The contmu|ng assooaUon of enteric viruses with a number of human diseases points up the lack of epldemmloglcal evidence to estabhsh the extent of the viral hazard m our potable water supphes (DeMichele, 1974) Increas|ng reliance on recycled water will accelerate the need to understand the factors wMch influence the surwval of these viruses in water The purpose of this study is to evaluate the mfluenee of two variables known to affect virus behavmr in the aquaUc enwronment, turbidity (Bmon, 1975) and virus concentration (Alan et al 1971), and a third variable, hght (ambient solar radlaUon), which has not been previously examined for its effect on v|ruses m the aquatm environment MATERIALS AND METHODS

E ~penmental facdmes Two reorculatmg flow systems were used (Figs 1 and 2) The tray system consisted of four plastic trays (46 u 27 u 4 5 cm) fitted wRh pumps whmh reclrculated the 5 51 capaoty approxtmately 11 times m

10mm The trays were placed m a water bath for temperature control The experimental channel system was a reorculatmg arttficaal stream constructed from four-foot alummmm sections hned with 15 mll vinyl to form two parallel channels 2146m long and 15cm deep Each channel was approximately 18 crn wide Including the reorculatmg hardware the system capacity was 30001 A more complete description of the channel facflRy has been pubhshed (Gannon et al, 1966)

Laght The tray experiments were conducted m June 1975 Solar mtensRy was recorded on a pyrhehometer using an arbRrary scale to provide relative values for cornparing experimental blocks of tray data Except for a 5 h period of heavy ram, the dayhght hours were typ,fied by full sun and occasional scattered cloud~ The temperature of the water m both dark and ambmnt trays ranged from 16 to 23 5°C w,th a mean of 18 7°C There did not appear to be a slgmficant dlfferenee between the hght (ambient) and dark tray temperatures

TurbMaty Turbidity was measured as nephelometnc turbtdity units (NTU) Natural turbidity was less than'2 5 NTU and was used as the low level High level turbldRy was prowded by ra, smg the turbidity to 36 NTU with a stock solutmn of organically loaded bentonite The bentonite stock solution was prepared m 3% beef broth, then centrifuged and resuspended three Umes m autoclaved river water to remove excess dissolved beef broth Prehmlnary work comparing virus, recovery from river water, wRh r,ver water plus bentonite showed a shght recovery reductmn from the river water plus benton|te

1091

1092

C P (UBBAOI I J (JANNO\ I'~ W ( ( X H R A N and (~ ~

IF////////~,J~'."~////////AI

I'

I1~'///~~],.~,'//////,///~1

J~~,~~~l

I~~~f~l

Constont temp water bott~ -142x Co6cm

~"

WILLIAMS

~iF//~~Ef//YA

~

I

Inle# " --"------~-- f - ~ - ~

~

I

Outlet

/

"

/

Coolant outlet flow

/ / / / Recfrculatmg reservoir

Tray and reservoir capacity = 5 5 t Pump flow flow rote = 225 I h-' Fig I Experimental hay system reclrculatmg flow pattern

0

©

Virus seeding location OunCe channel t"ns0de chonnel t26,500 l

Reserver ) 26, 500 I

Huron nver

\.\

A B C

4t6 I m=n-~ pump 0 Flow meter I~ Outlet I~ox [ ] O Inlet box [] E San'~ling StOtlon [~ Well Stom~ ~ r~

Fig 2 Experimental channel system

Loss of infectivity of poliovlrus 1

Virus concentration levels The high and low starting wrus levels m the tray study were approximately 105 and 104 P F U m i - ~, re~ spectlvely In the channel experiments, the starting levels ranged from 700 to 4200 P F U m l -

Experimental deslon and analyszs The basic statistic used to evaluate the influence of each variable and potentml variable mteracuons was the loss of infectlwty (LOI) rate The virus concentratlon of each sample was plotted against time and the LOI curves were determined by employing the least squares method of curve fitting The LOI curve was assumed to be of the form y = a e b~ This equatmn may be hnearized into

where y = the number of P F U mi-m of inoculated river water at t~me = ~ h b = the rate of LOI, and = the time elapsed in hours The experimental tray design utd~zed each variable at two levels All possible combinations of three variables at two levels were used twice during the 16 experimental units The umts were orgamzed into four blocks which were conducted on alternate weeks to allow prompt assay of the virus-seeded river water The specafic block design (of high and low variable levels) presented m Table 1 strengthened the analys~s 1

Samphno intervals The LOI rates were calculated from virus assays at zero, 025, 1, 6, 12h and each 12h interval thereafter All values were based on the sample assay of the seeded water except the zero hour data which were calculated from the viral assay of the stock wrus inoculum and the ddutmn values of the tray volumes The first tray samples were taken after the mixing period (15 mln)

For the channel experiments, water was pumped directly from the adjacent Huron River into 26,5001 reservoirs for m~xmg to provade a homogeneous supply For the tray experiments, river water was collected and stored in a 7501 carboy for a common source Conductlwty, pH and nutrient data were obtained prior to each set of experiments to estabhsh the uniformity between experiments

Vtrus preparation Culture of cells and virus Pohowrus l (vaccine strata LSc2ab) was propagated m Vero cells (ATCC CCL81) m 05 gallon roller bottles Vero cells were also used to assay v~rus m the channel study, as de-

LOI slope statistics for the first and second components at selected mtervals

Experimental tray* Bltlf Bit2 Bit3 Bit4 BIItl BIlt2 BIIt3 BIIt4 BIIItl BIIIt2 BIIIt3 BIIIt4 BIVtl BIVt2 BIVt3 BIVt4

of the potential effects of each variable and two-vanable interactions at the expense of mformat~on on three-factor interaction, the latter being confounded with the block effects Three channel expertments were conducted w~th slmdar varmble levels to examine the potential of the facdlty for LOI studies

Water source

l n y = Ina + by

Table

High level:[: variables L V TLV T No high levels TV TL LV T V

TLV L

No high levels TV TL LV

0-1 h

Interval 0-6 h 1-72 h

6--72h

-149 -137 - 1 45 - l 54 - l 60 - 1 51 - I 56 - I 61 - l 59 - I 53 - l 53 - 1 48 - l 31 - 1 26 - l 23 - l 24

-073 -076 - 0 73 -072 - 0 82 - 0 82 - 0 82 -082 -079 - 0 77 - 0 79 -083 - 0 73 -072 -068 -076

-012 -008 - 0 15 -006 -007 -007 - 0 15 -025 -006 -008 - 0 14 -023 - 0 l0 -007 - 0 14 -025

B--block, t--tray :~T = turbidity, L ffi hght, V = virus High turbidity > 30 NTU Low turbidity < 2 5 NTU High 118ht--amb~ent solar radiation Low hght--total exclusion High vlrus---110,000-130,000 PFU ml- 1 starting at time 0 Low v~rus--I 1,000-13.000 PFU ml -I starting at time 0 ?

1093

-017 -010 - 0 18 -009 - 0 11 -009 - 0 24 -032 -009 - 0 10 - 0 19 -031 - 0 14 -009 -023 -032

1094

( P Ct aaAOt J J GA"q~ON,K W (_O('HRANand G W WILLIAMS

scribed below The Vero cells were grown MEM m Hanks' balanced salt solution (HBSS) with 10°o fetal calf serum and an antibiotic supplement consisting of I00 units penicillin G 100/.tg streptomycin 100#g kanamycm and 4/~g amphoter]cm B per ml For virus preparation roller bottle monolayers were rinsed three times with 10ml of HBSS and each bottle then inoculated with 5 ml of undiluted dye-free stock virus and 25ml of serum-free, dye-free MEM When cultures showed advanced cytopathology, bottles were frozen and thawed three times Cellular debris was removed by centnfugatlon at 300 g for 20 mm Pooled supernatant fluid was stored at - 2 0 ' C until used Virus assay

BGM cells were used to assay wrus m the tray study BGM cells were grown m MEM with 10~o fetal calf serum, and 5 pg ml - ~ of gentamlcm and amphotenon B, subcultured as described by Dahhng et al (1974) For maintenance, serum concentration was reduced to 5~'o and penlcdhn-streptomycm (100 umts and 100/Jgm1-1) substituted for gentamlcm Levels of infectious virus in water were determined by plaque assay (Melmck & Wenner, 1969) using two-ounce prescription bottles The overlay medium consisted of 1 5% Bacto-Agar (Dlfco, lnc) m MEM 2% heat reactivated fetal calf serum, five-fold concentrate of the antibiotic supplement described above 0005°,0 neutral red and 25 mM MgCI2 Cultures were exammed dmly after 48 h (for a total of 7 days) and all plaques marked and recorded Data were recorded as plaque forming umts (PFU) per ml

sample The 9 ml of eluate was transferred into a sterde screw-capped test tube containing 1 ml of a 10-fold concentrate of HBSS and 50-fold concentrate ol standard antibiotics The sample volume had thus been reduced from 250 ml to 10 ml and was rendered isotonic for tissue culture application Any bacterial contammatlon which may have occurred at the field station was suppressed by the inclusion of antibiotics m the concentrated sample Prehmmary tests for recovery efficiency utlhzed 250 ml of river water seeded with 2066 PFUs per 1 0 ml and resulted in a 42 6°o recovery rate No concentration or clarification techniques were necessary at the v~rus levels used for the tray study Samples were assayed immedmtelv or stored at ¢ C for no longer than 12 h R ESU LTS

Tray expertments

The plotted data points appeared to fit a two stage LOI curve pattern which has been previously observed (Akin et al, 1971, Hlatt, 1964, Mahna et al, 1975) Figures 3-10 contain the curves of rephcared exper|mental umts, averaged for more convement presentat|on The curves Illustrate the virus mocula levels used, duration of LOI, and the pattern of two component slopes Several Ume Intervals were examined to determ|ne the most representat|ve two component curves Table I presents the slopes of the varmus LOI component intervals The transmon or breakpoint between the more rap~d, early LOI slope and the slower final LOI slope occurred during the

Vwus ~oncentratmq techmque

For the channel exper|ments the membrane filter process described by Walhs et al. (1972) was modified to allow rapid handling of the more turbid samples The ease of manipulation, minimal time requirements and sterile concentrate provided by the technique were the basic factors considered m selecting the process To the 250ml sample. 2ml of 01% methyl orange was added as a pH indicator The sample was then acidified to approximately pH 4 with IN HCI Following acidification 2 ml of 005 M AICI3 and 10ml of 1 25% cehte (Hill et al. 1974b) were added to the sample and thoroughly mixed The sample was then vacuum filtered through a combmauon of two 47ram d|ameter filters as a single unit, an AP25 fiberglass prefilter followed by an 0 45/am membrane filter The fiberglass prefilters were pretreated with a 1~,, Tween 80 solution followed by a thorough rinse with distilled water to minimize virus adsorptmn to the fiberglass The river water was vacuum filtered to adsorb the v~rus content of the sample onto the membrane filter A sterile test tube was aseptically placed beneath the filter base and the sample was eluted through the filters with 4 5 ml of 005 M gly~.me-NaOH buffer at pH II 5 An addmonal 4 5 ml of 005 M glycme-HCl buffer at pH 2 0 was passed through the filter to neutrahze the concentrated

i0 5

io4

"r

Sk~peI~--016 O ~.10 2 0.

I 6

12

24

36 T.~rle,

40

i 6O

I "7'2

h

FI& 3 Average tray loss of infecttvlty Bltl and Blllt4 (low turbmd2ty, hagh light, low virus)

Loss of mfectwity of pohowrus 1

1095

Lo~

104

104

s~

o~

b2=-O oe

Temp rot'Re 17-19"C

"r"

"I-

18-22%

t-

,o~ Slope b 2 = - O 0 6

15 0

O

0..

h O-

[ ~o2

I

I

6

[

12

I

24

L

I

36

I

48

60

I

I

72

6

I

12

I

24

I

I

~

T,me,

Tt n',A~, h

Fig 4 Averagetray loss of infect]wry Bit2 and Blllt2 (low turbidity, low hght, high vtrus)

L

48

60

I

72

h

Ftg 6 Average tray loss of mfoctwlty

Blt4 and B l l l t l

(h~gh turbtdlty, low hght, low virus)

104

IO4

o

O Slope b2 = - 0 1 6 Tamp ronge 16-20"(:: 18-23"C

103

=-

's

S k ~ be--O Oe Temp ronQe 16-19"C 20-22~

0

O

b2

IO2

~10 2

IO

I

6

I

12

I

24

I

36 Time,

~

48

I

60

i

72

h

F~g 5 Average tray loss of mfccUvlty

Blt3 and Blllt3

(high turbidity, high hght, high virus)

t

J

6

I

12

I

24.

I

~ Time,

I ~

48

i

60

I

72

h

Fig. 7 Average tray loss of mfcctlvlty

Blltl and BIVtl

(low turbidity, low hght, low wrY).

1096

C P CUBBAGt J J GANNON K W COCHRANand G W WILLIAM*, l - 6 h interval No data were available from within this interval so slope statistics were developed using both the l h point and the 6 h point as the breakpoint The amount of vanablhty m the slope stahstzcs was less for the second component when the breakpoint was arbitrarily placed at the 6h point ~o ~t was used throughout as a matter of convemence

104

Slo~ b~ -o 07 Temp r0n~ 17-~C 20-22~

c~ I

Analysis oJ variance

"6 0 102

I0

I

6

I

12

1,

24

t

I

36

48

l

60

J

Both hght and turbld,ty were indicated as sources of variation which affected the LOI rate during the second component (6-72 h) and the interval expanded to |nclude the translhon period ('1-72 h) The F values for turbidity derived from the analys,s of vanance, exceeded the crJt,cal F value at the 005 s~gmficance level for both of the noted intervals The F values for hght were even more slgmficant during the same intervals and exceeded the crmcal F values at the 001 level Table 2 contains the calculated /~ values None of the variables nor variable interactions were lmphcated as statistically slgmficant sources of LOI rate vanat~on during e~ther of the intervals used as potentml early LOI components (0-1 h and 0-6h)

72

Time, h

Channel e~per:ments

Fig 8 Average tray loss of mfect,vlty Bllt2 and BIVt2 (high turbidity, low light, high virus)

The LOI curves developed for the channel experiments also exhibited the two component curve pattern observed m the tray study Table 3 presents the slope statistics, mclud|ng run 3 which utdlzed both channel and tray fac|ht|es conducted with parallel variable levels 105

105

104

i0 4q

c~ I

~

_ •I03 ~6

103

0

0 ~. 102

Tsmp ron~ 16-19"C 20-22"C

H-

I0

1

6

1

12

I~1

24

36 Time t

I

48

{

I

~!0

h

Fig 9 Average tray loss of mfectwlty Bllt3 and BIVt3 (high turb, dlty. high hght. low virus)

s•. ~--o

6

12

24

,i

1

36

Time,

48

t

60

I 3'2

h

Fig 10 Average tray loss of mfectwlty Bllt4 and BIVt4 (low turbldlty, high hght. high virus)

Loss of Infectivity of pohovtrus 1 Table 2 F values for variables and variable interactions for tray experiments Source of Variation

F Value (6-72 h)

F Value (I-72 h)

8 18" 49 92* 2 83 1 81 I 02 139 1 02 1 37

6 86* 66 23* I 81 001 2 21 061

TurbtdRy Light TurbtdRy and hght Vwus TurbtdRy and virus Light and virus Repilcatlons:~ Blocks within rephcatlons

1 45

2 83

* Exceeds the crmcal F value at the 005 stgmficanee level i" Exceeds the crmcal F value at the 001 stgmficance level The F value for "rephcatmns" mdtcates the extent of varmtton between experimental umts whmchduphcate each other (BI and Bill, BII and BIV) See variable levels, Table I (Remington & Shork, 1970)

Figure 11 demonstrates the poor curve fit for the channel experiments when the data for the zero to six hour interval are used for the initial curve component. DISCUSSION The statistical analysts of the tray data did not identify the starting wrus concentration as a stgmficant factor m the deternunation of the rate of loss of infectivity (LOI) However, both turbidity and hght were implicated as significant sources of variability While the initial virus concentrataon was not slgmficant (at the 005 level) for the rate of LOI, an

1097

mterestmg difference in the duration of the early LOI components was noted between the channel tray curves Differences between the two facthtles rule out a definmve statement, but the lower starting virus levels used m the channel experiments (7~3-4200 P F U m l - t of river water) may have played a role m the reduction of duration of the early LOI curve component (Akin et al. 1971) Any such reference from the current study must be regarded as speculative, however, the low virus levels used m the tray study and the levels of the channel mocula were of the same order of magnitude as those reported by others when a stmllar reduction m the duration of the early LOI component was noted, and suggested that the virus level affected the duration of the mmal phase (Akin et al. 1971) It has been pointed out that two phenomena account for the reduction m the levels of mfectiv~ty observed m enteric viruses m the aquatic environment, actual loss of infectivity brought about by impairment of the mfeetive capability of the virus particle, and/or the apparent loss of infectivity which results from conditions that cause infectious particles to join together into clumps (which act as a single infectious focus) (Akin et al, 1971, Floyd et al, 1977, Gerba & Schaiberger, 1975a, Schaiberger, 1975b, Hill et al, 1974a, Moore et al, 1975, Schaub et al, 1974, Young & Sharp, 1977) The apparent LOI has been divided into s~tuatmns that involve (1) the adsorption of virus partmles onto other particulate matter m intimate contact with the water, and (2) the aggregation of wrus particles into virus clumps (Akin et al, 1971, Bitton. 1975, Carlson et al, 1968, Fowlks, 1959, Floyd & Sharp, 1977. Gerba & Shmberger, 1975a, Hill et al, 1974b) lomc

Table 3 Channel experiments LOI slope statlsUcs

Experimental umt Run I channel I channel 0 Run 2 channel I channel 0 Run 3 channel I tray I channel 0 tray 0

Experimental condmons Low turbidity, low hght and low wrus High turbidity, high hght and low virus

(0-1 h) Slope

Interval (t-last h) Slope r 2.

- 1 53

-018

096

- I 57

- 0 19

098

High turbidity, low hght and low virus Low turbtdRy, high light and low virus

-243

-009

100

-242

- 0 10

099

Low turbidity, high hght and high virus Low turbldRy, high hght and high virus High turbtdRy, low hght and high virus High turbidity, low hght and high virus

- 0 72

-009

097

- 0 74

- 0 09

I 00

-094

-007 -099

-092

-008

1 00

* r 2 ts a measure of how well the data points fit the calculated slope hne The O-I h slopes are all 1 00 since only two data points were available for calculating the slope

1098

C P Ct~Br~A¢.~ ! J (~tANNON K W (.OCHRANand (, V~ ~¢~llLIAblh

~o3

- - - - - Curve fit for O-6h and 6 - 5 7 h ' Curve fit for 0 - I hond 1-57h o Doto points

k

io2

\ o ~

o £

! ::S ~o

io

I I0

I 20

I, 30 Time,

[ 40

[ 50

t 60

B

Fig I1 Channel study comparison of two LOI curve intervals (high turbldfly, low light and low wrus level)

strength and substances which compete for the available bmdmgs sites |nfluence both adsorption and aggregation (Akm et a l , 1971) The appearance of the early component of the LOI curves suggests that the aquatic enwronment provided by the river water was sufficiently dtfferent from the stock v~rus preparatmn to mask the potentml influence of the variables on the rate of LOI durmg the early portion of the experiments Thus, the mfluence of turbidity and hght on the LOI rate became apparent only after the factor(s) responsible for the early LOI component subsided Turb~dity was tmphcated at the 005 significance level as a source of LOI rate vananon during the second curve component and dunng the interval which included the second component and the transition per|od, but not durmg the early component The increased LOI rate assocmted with the higher level of turbl&ty suggests the possibility of sohds associated infectivity and reiterates the necessity for using caution m assuming that the potentmi infectivity level is no higher than the assayed value (Walhs & Melrock, 1967) The number of reported recoveries over 100~o, no doubt result from sohds separation m the sampling process and illustrate the point and reveal the incomplete state of knowledge of the phenomena governing the behavaor of enteric viruses in the aquatic environment (Akin et al, 1971) The effect of solar radmnon has not been previously examined with regard to natural waters and enteric viruses U V hght has been Investigated as a

vlrucldal treatment lor contaminated waters ~n th~ shellfish industry IHIlt et al 1969) and th~ met.h.mlsms responsible for the vtruc~dal effects ha,'e been examined (Kalter & Mlllstem 1974, Schaub ¢,1 .1 1974 Welhngs et al 1974) In addmon th, cxtc.nt and penetration el solar radmnon rote tla~ aquanc environment has been mvesngated {Ruttner ]97q~ and suggests the potentml for ambient solar r~tdlatltm to exert a wrucidal effect As ~as the case wflh turbidny the rapid LOi rates associated vvtth the early component masked the detecnon of any potential influence of hght during that mterval Data flora the trays which rece~ed ambient solar radiation was also examined tor evidence of a diurnal LOI pattern, with a diminished mght-nme LOI rate Unfortunately the rates ~ere so rapid that only three of the eight ambient light trays had more than one dayhght slope interval belore the mfecnvlty dropped below the detection level The dmrnal slope analysis was thus inconclusive l--lt,~ever the staUsncal evaluanon of slope data produt, ed highly significant varmtlon associated with the effects of solar radlanon tollowmg the early LOI component Additional research is needed to determine if the effect ~s direct or ,is has been suggested ~s mdJrect such as enhanced b~ological actlvlt 5 A similar a~,celeraUon in the loss of mlecnwt) has been reported for wruses seeded onto vegetation ~la spray irrigation (Konowalchuck & Splers, Iq71) The results of the current research indicate that solar radiation may be added to the list of variables which determine the mfectlwty of pohovirus m the aquaUc environment The mechamsms by which turb~&ty affects the rate of LOI needs to be further defined In natural waters, where turbidity levels may be higher and depths greater than those used m the current research, the penetration of hght would be diminished and henct. turbidity would shade and protect virus particles from the ~lrucldal effects of light Further research is needed to determine the impact of hght m natural settings such as impoundments with relan~ely clean water Acknowledgements- This paper was taken m part from a

dissertation submitted b) C P Cubbage m partial fulfilment of the requirements for the Ph D Degree at The Umverslty of Michigan Ann Arbor

REFERENCES

AkmE W,Benton W J & Hill W F, Jr (1971)Enteric viruses ground and surlace waters a revlev~ of their occurrence and survwal and a revww of methods and appheatlons Proc 13th Wat Qual Conf Umv llhnols Bull 69, 59-74 Berg G (1976) Microbiology--detection, occurrence and removal of viruses J War Poltut Control Fed 48, 14t0-1416 Berg G (1977) Microbiology--detection, occurrence and removal of viruses J Wat Pollut Control Fed 49, 1290-1299

Loss of mfechvlty of pohowrus I Bitton G (1975) Adsorptmn of wruses onto surfaces m soil and water Water Res 9, 473-484 Carlson G F , Jr. Woodward F C, Wentworth D F & Sproul O J (1968)Virus reactivation on clay particles m natural waters J War PoUut Control Fed 40, R89-R 106 Dahhng, D R. Berg G & Berman D (1974) BGM a continuous cell line more sensitive than primary rhesus and african green kidney cells for the recovery of viruses from water Hlth Lab Scl !1, 275-282 DeMicbele E (1974) Water reuse, virus removal and pubhc health V~rus Surwval :n Water and Wastewater Systems Edited by Mahna J F . Jr &. Saglk B P Center for Research m Water Resources, Umv Texas. Austin. Texas, pp 45-56 Fowlks W L (1959) The mechanism of the photodynamlc effect J invest Derm 32, 233-247 Floyd R & Sharp D G (1977) Aggregatmn of pohowrus and reovirus by ddutlon m water Appl Microbwl 33, 159-167 Gannon J J. Diraslan J A & Phaup J D (1966) A versatde outdoor channel for water pollution mvesUgauons Proc 21st lnd Waste Conf Purdue Umv Eng Bull 121, 234-247 Geldreich E E (1969)Water pollution microbiology J War Pollut Control Fed 41, 1053-1069 Gcrba C P & Schaibergcr G E (1975a) Aggregation as a factor m loss of wral t~ter m seawater Water Res 9, 567-571 Gerba C P & Scha~berger G E (1975b) Effect of particulates on virus survival m seawater J Wat Pollut Control Fed 47, 92-103 Hmtt C W (1964) Kinetics of the reactivation of wruses Bacter~ol Rev 28, 150-163 HlllW F , J r , A k l n E W , B e n t o n W H, Mayhew C J & Jakubowske W (1974a) Apparatus for condmonmg unhm~ted quantmes of finished waters for enteric virus detection Appl M~crob~ol 27, 1177-1178 HlllW F . Jr, Akln E W, Benton W H, Mayhew C J & Metcalf T G (1974b) Recovery of pohovlrus from turbid ¢stuarme water on microporous filters by the use of cehte Appl M~crobwl 27, 506-512 Hill W F, Jr, Akin E W, Benton W J & Metcalf T J (1972) Virus m water II Evaluation of membrane cartridge filters for recovery low mulhphc~t~es of pohovlrus from water Appl M~crobwl 23, 880888

1099

Hdl W F . Jr. Hamblet F E & Benton W H (1969) lnachvation of pohovlrus type 1 by the Kelly Purdy ultraviolet seawater treatment unit Appl Mwrobwl 17, I-6 Kalter S S & Mdlstem C H (1974) Efficacy of methods for the detection of wruses m treated and untreated sewage Vtrus Survival m Water mid Wastewater S~tems Edited by Mahna J F, Jr & Sagik G P Center for Research m Water Resources, Umvers~ty of Texas Austin Texas Konowalchuck J & Splers J I (1971) An evaluation of three agents for elutmg adsorbed enterovlrus from mdhpore mcmbrane filters Can J Microbtol 17, 1351-1355 Mahna J F, Ranganathan K R, Saglk B P & Moore B E (1975) Pohovlrus reactivation by actwated sludge J War Pollut Control Fed 47, 2178-2183 Melnlck J L & Wenner H A (1969) Enterovlruses Dtaqnostlc Procedures for Viral and Rtckettstal h~ectwns Edited by Lennettee E H & Schmldt N J American Pubhc Health Assoc~aUon. 4th EdlUon, pp 529-602 MooreB E Saglk B P & MahnaJ F, Jr 0975) Viral association with suspended sohds Water Res 9. 197-203 Remington R D & Schork M A 0970) Statistics with Apphcatlons to the Biological and Health Sciences Prentice Hall, New Jersey Ruttner F (1973) Fup~damentah oJ Lomloloyy 3rd Edition Umverslty of Toronto Press, Toronto Schaub S A, Sorber C A & Taylor G W (1974) The assoctatlon of enteric wruses with natural turbidity m the aquatic enwronment Vtrus Surwval m Water a~d Wastewater Systems Edited by Mahna J F, Jr & Saglk B P Center for Research m Water Resources Umverslty of Texas, Austin, Texas pp 71-83 Walhs C, Henderson M & Mclmck J L (1972) Enterowrus concentration on cellulose membrane Appl Mttroblol 23, 476-480 Walhs C & Melmck J L (1967) Concentration of enteroviruses on membrane filters J V~rol !, 472-477 Wclhngs F M, Lewis A L & Mountain C W (1974) Virus surwval following wastcwater spray ~rngatlon of sandy soils Varus Surt'wal m Water and Wastewater S~'sterns Edited by Mahna J F, Jr & Saglk B P Center for Research m Water Resources Umverslty of Texas Austin, Texas pp 253-260 Young D C & Sharp D G (1977) Pohowrus aggregates and their surwval m water Appl Mt~roblol 33. 168 177