Effects Environmental Variables and Soil Characteristics ... · 1068 HURST, GERBA, ANDCECH minimum essential medium (14) supplemented with 10%fetal bovineserum,5%tryptose-phosphatebroth,

Vol. 40, No. 6APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Dec. 1980, p. 1067-10790099-2240/80/12-1067/13$02.00/0

Effects of Environmental Variables and Soil Characteristics onVirus Survival in Soil

CHRISTON J. HURST,' CHARLES P. GERBA,'* AND IRINA CECH'Department of Virology and Epidemiology, Baylor College ofMedicine,1 and University of Texas School of

Public Health,2 Houston, Texas 77030

Because of the increasing emphasis placed upon land application as a means ofwastewater disposal, it is important to evaluate the influences of different factorsupon virus survival in soil. The objective of this study was to measure the effectsof various environmental variables on virus persistence. Test samples of soil wereplaced in vials, and the soil was wetted with suspensions of virus in either distilledwater, unchlorinated secondary sewage effluent, or mixtures of effluent and water.The viruses used were coxsackieviruses A9 and B3, echovirus 1, poliovirus 1,rotavirus SAll, and bacteriophages T2 and MS2. The rate of virus inactivationwas evaluated statistically with regard to conditions under which the vials wereincubated and to the soil characteristics. The factors that were found to influencevirus survival were temperature, soil moisture content, presence of aerobic micro-organisms, degree of virus adsorption to the soil, soil levels of resin-extractablephosphorus, exchangeable aluminum, and soil pH. Overall, temperature and virusadsorption to soil appeared to be the most important factors affecting virussurvival.

Viruses conta;ined in wastewater that is ap-plied to soil can persist in the environment forprolonged periods of time. Indigenous enterovi-ruses have been isolated from soils of rapidinfiltration basins receiving primary and second-ary sewage effluents (7). Indigenous enterovi-ruses have also been shown to survive for 28days in soil of a cypress dome after applicationof secondary sewage effluent onto the dome soil(17). Seeded polioviruses were shown by Tierneyet al. (16) to survive in soil after irrigation witheffluent for 11 days during the summer and 96days during the winter. Because viruses canpersist in soil, the land disposal of wastewaterposes a potential risk of resultant human illness.To evaluate the extent of this risk, survival ofviruses in soil must be carefully studied and thefactors affecting virus persistence in soil must beelucidated.A few studies have been conducted concerning

the effects of environmental conditions on sur-vival of viruses in soil. Murphy et al. (11) foundthat survival of mouse encephalomyelitis viruswas greater in soil at neutral pH than in soil thathad been adjusted to either pH 3.7 or pH 8.5,and Bagdasaryan (3) determined that the sur-vival time of enteroviruses was shorter in loamysoil than in sandy soil. Lefler and Kott (9) ob-served that poliovirus and bacteriophage f2 sur-vived equally well in sand saturated with dis-tilled water, with distilled water containing cat-ions, with tap water, and with oxidation pondeffluent.

In general, poliovirus appears to persist forlonger periods of time at 4 to 80C than at 20 to370C (9, 13, 18). Virus survival also appears tobe generally greater under sterile conditionsthan under nonsterile conditions, although thesetests have so far been run only under aerobicconditions (3, 11). No definite reason has beengiven for the apparent difference in virus sur-vival in soil aerobically under sterile versus non-sterile conditions; however, Bagdasaryan (3)concluded that the difference was so small as tohardly be attributable to the antagonistic effectof soil microflora. Published reports concerningthe effects of soil moisture upon virus survivalare conflicting and indicate no clear trend (3, 9,13, 18).The purpose of this study was to compare and

statistically evaluate the effects of these factorsupon virus survival in one system. Experimentswere performed to compare the possible antag-onistic effects of soil and sewage effluent micro-flora upon virus survival under anaerobic as wellas aerobic conditions, to examine the effects oftemperature and of soil moisture upon virussurvival, and to examine the possible role(s)which soil properties play in persistence of vi-ruses.

MATERIALS AND METHODSCells and cell cultures. Buffalo green monkey

(BGM) kidney cells, a continuous green monkey kid-ney cell line, were used for growth of virus stocks andassay of viruses. These cells were cultured in Eagle

1067

on Novem

ber 1, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

http://aem.asm.org/

1068 HURST, GERBA, AND CECH

minimum essential medium (14) supplemented with10% fetal bovine serum, 5% tryptose-phosphate broth,0.03% glutamine, 100 U of penicillin per ml, 100 ,ug ofstreptomycin per ml, and 0.15% sodium bicarbonate.Maintenance medium was the same as supplementedminimum essential medium but with a reduction infetal bovine serum to 2% and an increase in sodiumbicarbonate to 0.25%.MA-104 cells, a continuous fetal rhesus monkey

kidney cell line, were used in the propagation andassay of simian rotavirus SAll. These cells were grownin Eagle minimum essential medium supplementedwith 10% fetal bovine serum, 5% tryptose-phosphatebroth, 2% essential vitamins, 0.25% glucose, 0.03% glu-tamine, 0.075% sodium bicarbonate, 100 U of penicillinper ml, 100 ug of streptomycin per ml, and 25 ug ofgentamicin per ml.Virus and virus assays. The enteroviruses used

included plaque-purified strains of poliovirus 1 (strainLSc-2ab), echovirus 1 (strain Farouk), coxsackievirusA9 (strain Bozek), and coxsackievirus B3 (strainNancy). Enterovirus stocks were grown in BGM cells.Thirty-two-ounce (ca. 960-ml) bottles of BGM cellswere inoculated at a multiplicity of infection of 0.2 to1.0 and incubated for 1 h at 37°C, and maintenancemedium without serum was added back to the bottles.Upon the appearance of 4+ cytopathic effect, bottlescontaining virus were rapidly frozen and thawed threetimes, and cell debris was centrifuged out. Virus stockwas stored at -20°C until use. Enterovirus assays weredone by the plaque-forming unit method as describedby Melnick and Wenner (10). When necessary, enter-ovirus samples were diluted in Tris buffer [containing(per liter) 3 g of tris(hydroxymethyl)aminomethane, 8g of NaCl, 0.38 g of KCI, 0.1 g of Na2HPO4, 1 g ofdextrose, 0.1 g of MgCl2, 0.1 g of CaCl2, and 0.002 g ofphenol red and (per ml) 100 U of penicillin and 100Mg of streptomycin].

Simian rotavirus SAll stocks were MA-104 infectedcell lysates prepared by freeze-thawing, sonication,and low-speed centrifugation. SAll virus was assayedby the plaque-forming unit method of Smith et al.(15). Samples containing rotavirus were diluted in a5% tryptose-phosphate broth solution (BBL Microbi-ology Systems, Cockeysville, Md.). When necessaryfor preventing growth of contaminating microorga-nisms, nystatin (E. R. Squibb & Sons, Princeton, N.J.)was added to the agar overlay medium used for bothenteroviruses and rotavirus at a final concentration of100 U/ml. Neomycin sulfate (E. R. Squibb & Sons)was also added to the agar overlay medium whennecessary at a final concentration of 10 Mg/ml.Two different coliphages were used: MS2 and its

host Escherichia coli ATCC 15597, obtained from theAmerican Type Culture Collection, Rockville, Md.,and T2 along with its host E. coli B, obtained from G.Schaiberger (University of Miami, Miami, Fla.). Bothhost bacteria were maintained in culture on Trypticasesoy broth (BBL).

Bacteriophage stocks were prepared by the follow-ing procedure. Agar plates confluent with plaques wereoverlaid with 5 ml of Trypticase soy broth. After 3 hat room temperature, the plates were gently swirledseveral times and the broth was decanted. The sus-

APPL. ENVIRON. MICROBIOL.

pension of free bacterial cells was then centrifuged at1,600 x g for 15 min to sediment the bacterial cells.The supernatant fluid was passed through an 0.45-,um-pore-size HA membrane filter (Millipore Corp., Bed-ford, Mass.) which had previously been washed withsterile Trypticase soy broth. Samples (0.4 ml) of thefiltrate were stored in ampoules at -70°C until usedin the experiments.The bacteriophages were assayed essentially by the

procedure described by Adams (1). The bottom agarfor the phage assays consisted of Trypticase soy brothcontaining 15 g of agar per liter. The top agar for thephage assays consisted of Trypticase soy broth con-taining (per liter) 7 g of agar and 1 ml of 1 M CaCl2.The plating procedure consisted of mixing together 3ml of top agar, 0.1 ml of virus sample, and 1 ml of log-phase host cell suspension. This mixture was thenspread over a layer of bottom agar. Bacteriophageplaques were counted after 24 h of incubation at 37°C.

Sofls. The main soil used in the virus survivalstudies was a loamy sand obtained from the Salt Riverbed near Phoenix, Ariz. (FM soil). This soil is identicalto that used in the rapid infiltration beds at theFlushing Meadows experimental wastewater landtreatment site and has been used previously for study-ing virus removal during filtration of wastewater bysoil (8). Several other soils were also used for a portionof the experiments. These soils were Vernon, Clarita,Windhorst, Chigley, unclassified soil sample, Pomellofine sand, Anthony, and Rubicond. The latter eightsoil samples were obtained through the courtesy of J.F. McNabb and C. G. Enfield (U.S. EnvironmentalProtection Agency, Ada, Okla.) and have been char-acterized extensively (4). The characteristics of thenine soils are presented in Table 1.Experimental procedure for determining virus

survival in soil. The procedure used for the labora-tory studies on virus survival was as follows. Two-gram amounts of soil were placed into 4-dram (16-ml),screw-capped glass vials, and the soil was then wettedwith virus suspended in either unchlorinated second-ary sewage effluent, distilled water, or a mixture ofeffluent and water. The vials were loosely capped andeither incubated in GasPak anaerobic jars containinggenerated H2-CO2 atmosphere (GasPak generatorsand anaerobic jars purchased from BBL), with wetpaper towels to provide humidity and prevent desic-cation, or incubated aerobically in humid chambersconstructed of plastic tubs partially filed with wet,crushed rock and covered tightly with aluminum foil.Before placement in the aerobic humid chambers, thecrushed rock was washed with tap water to removeextraneous organic material and autoclaved for 1 h toprevent growth of mold during the experiment.

Three survival experiments were performed. Thedesign for the first study involved soil wetted witheither distilled water, unchlorinated sewage effluent,or a mixture of 25 or 50% sewage effluent in distilledwater. For the first study, half of the water-effluentcombinations were sterilized by autoclaving beforeaddition of the virus, and the vials containing soil werealso sterilized by autoclaving before addition of thevirus suspensions. Incubation of one-third of the vialscontaining virus was done at each of the temperatures

on Novem


.asm.org/

Dow

nloaded from

http://aem.asm.org/

VIRUS SURVIVAL IN SOIL 1069

0

_

.0

D

w_ sr 0 0 CD oD 0ucq~

oe eq .Ot . o q

E- o o Lo

.3e qaeEqV V

be LOs1-i L13too 00

R X = a5 P ~OqCS U: eC oR CO

CO _ q QqO COC LX. 4 o=,, D; o > N N t w 0

g Po

0

4)

v

Lo t-S C)CS

LO

t

o, ot-' oD souo*o e ~o0o<t4i XE O co uoco co _O

o g C* o oeq^o

ueqo

00X0 '- '-4e CO Q 0

mw VDX o vo000s

:^_10 Lesr N 0 Xcoco

c' q;X 6 C'sr s '. X

-E,-e qc-

C6~ ~ ~ ~~Cc

.: z ~~~~88 c; C;8 Qb

m 1-4 LO 00 0C 10moV 00808 00

0 eq~~~~~~1 L

o Co E5CD- CO

0)~~~~~~~AA ,c

0)

0

CO)

.0~~~~

* *~ y

(1, 23, and 37°C). Half of the vials were incubatedunder aerobic conditions, and the other half wereincubated under anaerobic conditions. In all, 48 differ-ent sets of incubation conditions were used. The initialvirus titers for vials of FM soil wetted with eitherdistilled water, sewage effluent, or diluted effluentwere all 1.4 x 104 plaque-forming units per g for thosesamples containing poliovirus and 2.1 x 105 plaque-forming units per g for those samples containing echo-virus. The soil moisture in this first study as well asthe third study was 15% of the soil by wet weight. Thismoisture value was chosen because it was the averagethat we have obtained for soil samples in field studieson virus survival. FM soil was chosen for this studybecause it is currently being used for the rapid infiltra-tion of sewage effluent.

For the second survival study, virus was incubatedaerobically at 230C under nonsterile conditions in soilwetted with a 50% solution of sewage effluent. Fivedifferent sets of vials were used in this study, with thedifference in incubation conditions being the moisturecontent within the vials. The amounts of moisture,expressed as percentages of wet weight of the vialcontents, were 5, 10, 15, 25, and 40%. At soil moisturecontents up to 15%, no freestanding liquid was presentin the vials. Some freestanding liquid was present at25% moisture, and a much greater amount was presentat a moisture content of 40%. Incubation of all vialswas done under nonsterile aerobic conditions at 230C.The third survival study entailed comparisons be-

tween the survivals of poliovirus 1, echovirus 1,coxsackieviruses A9 and B3, rotavirus SAll, and bac-teriophages T2 and MS2 in the nine different soils.For this study, all soil samples were adjusted to 15%moisture by addition of a 50% solution of sewageeffluent, and incubation was done under nonsterileconditions at 23°C.For all of these laboratory studies a group of 10 to

15 replicate vials with identical conditions were incu-bated together so that on each sampling date one ofeach type of vial was removed intact, and the contentswere diluted in Tris buffer and frozen at -20°C untilassayed. Dilution of the vial contents before freezingwas accomplished by the addition of4 ml of Tris buffercontaining 2% fetal bovine serum for soil containingenteroviruses or serum-free buffer for samples con-taining rotavirus or bacteriophages. Assay of the sam-ples was by direct inoculation.Two steps were used in analysis of the survival data

(Fig. 1). The first step was to determine slope valuesof the virus survival curves as a function of time. Virussurvival was expressed as the logarithm of the ratioNT/No (called the survival ratio), where No equals thevirus titer at the beginning of the experiment and NTequals the virus titer at elapsed time T. A log1o NT/Novalue of -1 equals 10% of the initial titer, -2 equals1% of the initial titer, and -3 equals 0.1% of the initialtiter. The slopes of the survival curves were deter-mined by regression analysis, using the StatisticalPackage for the Social Sciences (SPSS) computerprogram (12). The survival data appeared, generally,to be well represented by the linear equation y = at+ b, in which t is time in days and coefficients a andb are, respectively, the slope and the intercept as

0

-

;t

10Cc

cc

0cc

U

cr,.

c4L.

Zs

L-

VOL. 40, 1980

r- m CD C14'3 !4 (D ko U,.o tR li . .

E..4 - 0 0 0 (D -

on Novem


.asm.org/

Dow

nloaded from

http://aem.asm.org/


0-

,:-1I

0-A

z

-32-2o 0 2li

2

DAYS VARIABLE

FIG. 1. Format used for regression analysis. (A)Survival slope determination-survival slopes ex-

pressed as the rate of log1o change in virus titer perday. (B) Survival slope values versus independentvariable-the independent variables used in thisstage of the analysis represented experimental con-

ditions or soil characteristics. The example shown in(B) indicates a positive correlation between the vari-able and virus survival.

determined by least-squares analysis. The second stepwas to evaluate by stepwise regression the effects onsurvival slope value of the experimental variables tem-perature, level ofsewage effluent, aerobic or anaerobicatmosphere, sterile or nonsterile environment, soilmoisture content, virus adsorption to soil, and soilcharacteristics. This step involved the plotting of bi-variate scatter diagrams and computation ofa stepwisemultiple regression. In this analysis the survival slopevalues were taken as the "response" or "dependent"variables (y), which were expected to be functionallydependent on incubation conditions and soil charac-teristics (denoted by xi, x2,. . . x"). The multiple regres-sion model was: y = c + a,x, + a2x2 + . . . a,x,. A Pvalue was calculated in each step of the regression,reflecting the statistical significance of the regressioncoefficients of each entry. For references on compu-tational procedures used, the reader is referred tochapters 18 and 20 of the SPSS manual (12).

RESULTSEffects of environmental conditions on

poliovirus survival in FM soil. The survivalresults for various experimental conditions areshown in Fig. 2 through 5 on a time scale forvarious concentrations of sewage effluent usedto wet the soil. It appeared from these plots thattemperature had a large effect on survival of thevirus, with inactivation rates increasing as tem-perature increased. It also appeared that virussurvival under aerobic nonsterile conditions wasless than that occurring under the other threesets of conditions-aerobic sterile and both an-

aerobic sterile and anaerobic nonsterile. To fur-ther examine the possible difference betweenvirus survival under aerobic nonsterile condi-tions versus the other sets of conditions, the datawere plotted with respect to temperature. Thegraph of virus survival at 230C (Fig. 6) indicatedthat virus survivals under both anaerobic sterileand anaerobic nonsterile conditions as well as

under aerobic sterile conditions were similar.Virus survival under aerobic nonsterile condi-tions was distinctly lower than those under theother three sets of conditions. Thus, the pres-ence of aerobic microorganisms appears to havedeleteriously affected virus survival. Anaerobicmicroorganisms, at least those which existed inthe soil and sewage used in this study and underan H2-CO2 atmosphere, did not appear to affectvirus survival. The influence of aerobic micro-organisms was also evident, but to a lesser de-gree, with samples incubated at 370C. This ap-parently reduced degree of influence of aerobicmicroorganisms at 370C may have been due tothe more pronounced influence of thermal ef-fects upon virus survival at that temperature.There was no discernible effect of aerobic micro-organisms upon virus survival for samples incu-bated at 1°C, possibly as a result of the muchslower growth and metabolism rates of the mi-croorganisms at this lower temperature.

In the first step of regression analysis of thedata from this survival study, the rate of changein virus titer, expressed as logio NT/No, appearedto be linear with time. Exceptions occurred insome instances when virus inactivation was min-imal over a long period of time, i.e., with incu-bation at 100. In these cases random errors invirus sampling and assaying, in addition to pos-sible differences in evolution of the microbialenvironment inside the vials, made it difficult toobtain clearly significant slope values. In suchinstances the calculated slope values were takento be statistically nondifferent from zero.

Bivariate scatter diagrams produced in thesecond step of the analysis evaluating effects ofindividual environmental factors on virus sur-vival indicated the following: (i) the survivalslopes and incubation temperatures (Fig. 7)showed a statistically significant reciprocal cor-relation (Pc 0.01), indicating that virus survivaldecreased with increasing temperature, and (ii)the concentration of sewage effluent used to wetthe soil did not correlate with virus survival (P= 0.38). These two findings held under all con-ditions of aerobic or anaerobic atmosphere andsterile or nonsterile incubation. The observeddifference between survival slope values ob-tained for virus survival under aerobic nonsterileconditions versus those obtained under aerobicsterile conditions was not statistically significant(P= 0.156).

Effects of soil moisture on poliovirus sur-vival inFM soil. Figure 8 is the scatter diagramof average survival slopes derived from the sec-ond virus survival study against soil moisturecontent. Linear correlation of survival slope val-ues with soil moisture appears to be inadequateor inappropriate (P = 0.03) to describe the effect


:

on Novem


.asm.org/

Dow

nloaded from

http://aem.asm.org/


of soil moisture on virus survival. It seems, how-ever, that there was a nonlinear trend inasmuchas virus survival decreased as the soil moisturecontent increased up to 15% and then increasedwith the presence of additional amounts of liq-uid. The saturation level of this soil appeared tobe between 15 and 25%; thus, survival may havebeen at its minimum point near the soil moisturesaturation point.

Effects of soil characteristics on virussurvival. The results obtained from the studyof the effects of soil characteristics on virussurvival are plotted with respect to soil type inFig. 9. In Fig. 9H, the survivals of four additionalviruses, coxsackieviruses A9 and B3, echovirus1, and rotavirus SAll were compared with thoseof poliovirus 1 and the two bacteriophages inFM soil. The survivals of these four additionalviruses were not studied in any of the other soils,

0

z

and, therefore, the survival slope values obtainedfor these viruses were not included in the step-wise multiple regression.The analysis of bivariate scatter diagrams of

the virus survival slopes versus various soil char-acteristics indicated that virus survival corre-lated significantly (P c 0.05) with virus adsorp-tion to the soil and with soil saturation pH. Thesignificance of resin-extractable phosphorus wasmarginal (P = 0.056). None of the other soilcharacteristics demonstrated a statistically sig-nificant correlation with virus survival. Loga-rithmic transfornation of the independent vari-ables (the variable of soil pH was not trans-formed, since it is nornally represented as alogarithm) did not substantially improve the fitbetween the soil characteristics and virus sur-vival.

Stepwise multiple regression analysis of the

0 10 20 30 40 50 60 70 80 90 100

TIME (DAYS)

- 1

-2

0 10 20 30 40 50 60 70 80 90 100

TIME IDAYS)

FIG. 2. Poliovirus 1 survival in FM soil wetted with distilled water under sterile (top) and nonsterile(bottom) conditions. Symbols: 0, 10C, aerobic; 0, 1°C, anaerobic: A, 23°C, aerobic: A, 23°C, anaerobic: E,37°C, aerobic; I, 37°C, anaerobic.

VOL. 40, 1980

z

I-

-j

on Novem


.asm.org/

Dow

nloaded from

http://aem.asm.org/


0

I-z40

.t

J

-1

-2

-3

- 1

-2

-3

40 50

TIME IDAYSI

0 10 20 30 40 50 60 70 80 90 100

TIME DAYS)

FIG. 3. Poliovirus 1 survival in FM soil wetted with 25% sewage effluent under sterile (top) and nonsterile(bottom) conditions. Symbols: see legend to Fig. 2.

virus survival slope values as dependent varia-bles and soil characteristics listed in Table 1 asindependent variables was then performed. Ofthe 19 soil characteristics listed in Table 1, the10 that were ranked highest in the stepwisemultiple regression analysis are presented in Ta-ble 2. Virus adsorption to soil accounted for 23%of the variance, extractable phosphorus ac-counted for an additional 11%, exchangeable alu-minum accounted for about 4%, and pH ac-counted for about 13% (Table 2). In combina-tion, these four variables accounted for a totalof approxiimately 51% of all variance in the ex-periment.None of the six variables listed below pH in

Table 2 were statistically significant, and theyrepresented a total of only 13% change in Rsquare. The final equation derived from theseresults thus contains only the first four variables,i.e., adsorption to soil, extractable phosphorus,

exchangeable aluminum, and soil pH. The bi-variate scatter diagrams of the survival slopevalues obtained for incubation of the differentviruses in the nine soil types versus these foursoil characteristics are shown in Fig. 10.

DISCUSSIONThe results from evaluating effects of environ-

mental variables on virus survival indicated thattemperature was a significant predictor of virussurvival under all sets of conditions. This is inagreement with the work of many researchgroups that have evaluated the survival of vi-ruses in the environment. Virus survival was notsignificantly affected by the concentration ofsewage effluent, a finding which held true forboth sterile and nonsterile conditions, as well asaerobic and anaerobic incubation conditions.Our finding that the presence of sewage did notinfluence virus survival is of interest, since vi-


0

I-

z,

-j

on Novem


.asm.org/

Dow

nloaded from

http://aem.asm.org/


-I

- 2

- 3

TrIl .E DAY'Sl

-3I0 10 20 30 40 50 60 70 80 90 100

TIME {DAYS)

FIG. 4. Poliovirus I survival in FM soil wetted with 50% sewage effluent under sterile (top) and nonsterile(bottom) conditions. Symbols: see legend to Fig. 2.

ruses have been shown to survive longer in sew-age-polluted water than in nonpolluted water(2), possibly due to the presence of organic mat-ter contributed by sewage to the polluted water.Inasmuch as our results showed that sewageeffluent did not substantially influence virus sur-vival, we would like to suggest that the practiceof diluting sewage effluent with other water be-fore land application will, in itself, probably notaffect virus survival in the soil.The presence of aerobic microorganisms ap-

peared to result in decreased virus survival (Fig.2 through 6), although the results of a t-testperformed on data evaluating poliovirus survivalunder aerobic nonsterile conditions, contrastedwith aerobic sterile, anaerobic sterile, and an-aerobic nonsterile conditions, indicated thatthese two groups of conditions were not signifi-cantly different statistically (P = 0.156). Theresults of a t-test to compare poliovirus survivals

under aerobic nonsterile versus only aerobicsterile conditions likewise did not indicate sta-tistical significance.

Virus survival did not correlate linearly withsoil moisture content. The scatter diagram ofvirus survival versus soil moisture content did,however, appear to show a general trend; i.e.,virus survival appeared to decrease as the soilmoisture content increased up to the soil satu-ration point. Virus survival then increased asmore liquid was added to the system beyond thesaturation point. Possible reasons that couldaccount for this apparently enhanced survival ofvirus at both low and high soil moistures wouldinclude soil moisture-level-dependent differ-ences in the extent of virus adsorption to the soiland mechanisms of adsorption. Moisture-level-dependent differences in microbial growth ratesmight also affect virus survival.The multiple regression equation derived from

VOL. 40, 1980

02

zC

J

0

I-

C2

0

on Novem


.asm.org/

Dow

nloaded from

http://aem.asm.org/


-1

-2

TIME IDAYS)

-1I

-2

-30 10 20 30 40 50 60 70 80 90 100

T '.E IDAYS,FIG. 5. Poliovirus 1 and echovirus I survival in FM soil wetted with 100%o sewage effluent under sterile

(top) and nonsterile (bottom) conditions. Symbols: forpoliovirus 1, see legend to Fig. 2; for echovirus 1, O, PC,aerobic; El, 23°C, aerobic; El, 37°C, aerobic.

comparison of virus survival in the nine differentsoils includes virus adsorption to soil, soil resin-extractable phosphorus, soil exchangeable alu-minum, and soil pH. These four variables ac-counted jointly for 51% of all variance in thesurvival data. Sources of variance in the experi-ment not related to soil characteristics includedhuman error in sampling, variation in evolutionof the microbial environment between the dif-ferent vials, variation which naturally exists inthe plaque assay method, and perhaps otherfactors presently unknown to us.The fact that adsorption of viruses to soil

significantly affected virus survival is of greatimportance. This finding indicates a dilemmainsofar as virus inactivation during land treat-ment is concerned. On one hand, concern forpublic health would, of necessity, require that

land treatment sites be developed on soils withhigh virus adsorptive capacity. This is requiredto minimize the possibility of viruses applied tosoil reaching groundwater. On the other hand,virus survival is likely to be greatest in thosesoils that would be most effective in preventinggroundwater contamination. Because resultsfrom this study comparing virus survivals aero-bically and anaerobically under nonsterile con-ditions showed that virus survival is apparentlyprolonged under anaerobic conditions, virusreaching the groundwater in an anaerobic envi-ronment might have a subsequently prolongedlifetime of infectivity.The effect that soil pH was found to have

upon virus survival might be mediated throughvirus adsorption to the soil. It was found in arecent study (6) that soil pH was the major


z

II-

C,

0

13

0

on Novem


.asm.org/

Dow

nloaded from

http://aem.asm.org/


variable affecting adsorption of viruses to soil,with adsorption increasing as pH decreased.Both soil exchangeable aluminum levels and

resin-extractable phosphorus levels are impor-tant affectors of virus survival, as was evidencedby their position in the stepwise regression equa-

O r

0z

Iz

0

- 1

- 2

-3

0,

tion. Exchangeable aluminum only (that whichwould be soluble as cations in soil water), andnot total aluminum, was indicated as being im-portant in the stepwise regression equation. Pre-vious work on concentration of virus from waterhas indicated that the presence of aluminum

o 0% Sterile* 0% Non-sterilea 25% Sterile* 25% Non-sterileo 50% Sterile* 50% Non-sterile* IO% Sterileo IO% Non-sterile

10 20 30 40 50 60 70 80 90 100

a8 t

* *. Ii

A

II

A

Ita

a

o 0% Sterile* 0% Non-sterilea 25% Sterile& 25% Non-sterileo 50% Sterile* 50% Non-sterile* 100% Sterileo 100% Non-sterile

I I I I I I0 1 0 20 30 40 50

TIME IDAYS)

8

A

I

a

a

a

IA

60 70 80 90 100

FIG. 6. Poliovirus I survival in FM soil under aerobic (top) and anaerobic (bottom) conditions at 23°C.

0z

-z0

-1

-2

-3

VOL. 40, 1980

Il

on Novem


.asm.org/

Dow

nloaded from

http://aem.asm.org/


cations increases the adsorption of viruses tofilters (5). In this role of aiding virus adsorptionto filters, aluminum is much more effective thaneither magnesium or calcium salts, and neitherexchangeable magnesium nor exchangeable cal-cium significantly affected the stepwise regres-sion equation. Thus, the correlation betweenvirus survival and soil exchangeable aluminumwas possibly due to an increase in virus adsorp-tion to the soil particles at higher aluminumlevels.The level of resin-extractable phosphorus in

soil is in effect a measure of the amount ofavailable phosphate anions in the soil. The ad-dition of phosphate anions to a soil suspensioncan result in elution of adsorbed virus particlesfrom soil (D. H. Taylor, A. R. Bellamy, and A.T. Wilson, Water Res., in press). The findingthat virus survival increases as the level of resin-extractable phosphorus decreases may also havea relationship to previous work on concentrationof virus from water. Virus has been shown toprecipitate from water during alum and limeflocculation in relation to the precipitation ofphosphorus. In the latter case, phosphate anions,when free in solution, may have acted to inhibitvirus adsorption to the calcium and aluminumprecipitates.The soil variables of exchangeable aluminum,

for which correlation with virus survival had apositive direction, and resin-extractable phos-phorus, for which correlation with virus survivalwas reciprocal, were retained in the final modelequation even though their individual correla-tion with the survival slopes was not statisticallysignificant (P < 0.05). In combination with thetwo other soil characteristics of virus adsorptionand soil pH, the above characteristics provide a

0.00

LUJ 0.25

0.50

0.75

1.00

FIG. 7ues verstvival stuo

-0.08 w

-LJa--0-

-J

0

-0.09 1

.

-0.10 1

-0.11

.

II 1 I I

5 10 15 20 25 30 35 40

SOIL MOISTURE (% WET WEIGHT)FIG. 8. Scatter diagram of the survival slope val-

ues versus soil moisture content from the secondsurvival study; R = -0.31, significant at the 0.31 level.

TABLE 2. Stepwise regression equation for virussurvival slopes versus soil characteristics

Signifi-cance Overallof R signifi-

Variable regres- cancesion square of total

coeffi- modelcients

Adsorption to soil ........ 0.010 0.235 0.010Extractable phosphorus ... 0.056 0.345 0.006Exchangeable aluminum .. 0.257 0.381 0.010Soil saturation pH ........ 0.026 0.508 0.003% Silt ................... 0.349 0.529 0.005% Organic matter ........ 0.265 0.558 0.007Conductivity ............. 0.177 0.599 0.007Surface area ............. 0.619 0.605 0.014Exchangeable calcium .... 0.416 0.620 0.022Total cation-exchange

capacity ............... 0.302 0.646 0.028

15 12

model capable of explaining 51% of the variance_§ between the survival slopes and which is statis-

* tically significant (P c 0.05). Elimination fromthe four-variable equation of either exchangea-

j| ble aluminum separately or of both variables(exchangeable aluminum and resin-extractable

. phosphorus) together resulted in loss of the sig-nificance of soil saturation pH in describing virussurvival in the test soils. Although the results of

_. the survival studies cannot prove this hypothe-, , , , , sis, it would appear that the effects of soil pH1 10 20 30 40 upon virus survival are mediated through chem-ical equilibria, among which are the charge

INCUBATION TEMPERATURE (C) states and relative concentrations of aluminumScatter diagram of the survival slope val- and phosphorus.

us incubation temperature from the first sur- The final stepwise multiple regression equa-dy; R = -0.68, significant at the 0.00O1 level. tion derived from this study is: y = 0.1005 +


on Novem


.asm.org/

Dow

nloaded from

http://aem.asm.org/


0

A aPoloA APhap T2&Phagp MS2

-1

-2

-3 I J0 5 10 15 2 30 35

TIME (DAYS)

0

TIME (DAYS)

z -1

-2

-3

-3

TIME (DAYS)

z -1

-2

-3

TIME (DAYS)

5 10 15 a 2 30 35TIME IDAYS)

0 5 10 15 2D 25 30TIME (DAYS)

31 I I I I E-3 ' -30 5 10 15 20 25 30 0 5 10 15 20 25 30 0 5 10152D 253035

TIME (DAYS) Time IdWs ) TIME IDAYS)

FIG. 9. Comparative survivals of viruses in different soils. The soils tested were: (A) Vernon, (B) Clarita,(C) Windhorst, (D) Chigley, (E) unclassified soil sample, (F) Pomello, (G) Anthony, (H) FM, and (I) Rubicond.

0.0025x, - 0.0008x2 - 0.0007x3 -0.0510x4, where,for a given soil, y is the average of the survivalslope values for the three viruses under theconditions of the experiment, xi is the averagepercent adsorption of all three viruses to thesoil, x2 is the resin-extractable phosphorus value(parts per million) for the given soil, x3 is theexchangeable aluminum value (parts per mil-lion) for the given soil, and x4 is the saturation

pH value for the given soil.It would be strenuous to use regression equa-

tions of the type developed in this study foractually predicting virus survival under naturalfield conditions of constantly changing temper-ature and soil moisture. Instead, the value ofsuch derived equations lies in estimating virussurvival in different soil types based upon theirknown physical properties.

VOL. 40, 1980

z

1-z

2JP

0 I2°zr-

-2

3

0

z -I

z

-2

o

z-

0c

p29

on Novem


.asm.org/

Dow

nloaded from

http://aem.asm.org/


0.0

-0. 14

2 *

:

-0.2 _-j

0--0.3 _

-0.4 _

-0.5 >

25 50 75

VIRUS ADSORPTION TO SOIL 1%)

-0.6 L

0100 10 20 30 40 50 60 70

SOIL EXTRACTABLE PHOSPHORUS Ippml

80

0.0 r

22

* 2

-0.1 22

-0. 2 L--i

-0.3 V

-0.4 F

-0. 5 F

0 50 100 150 200 250 300

SOIL EXCHANGEABLE ALUMINUM (ppm)

-0.6'350 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5

SOIL SATURATION pH

FIG. 10. Scatter diagrams of the survival slope values versus soil characteristics from the third survivalstudy. Virus adsorption to soil: R = 0.48, significant at the 0.0052 level. Soil extractable phosphorus: R =

-0.34, significant at the 0.040 level. Soil exchangeable aluminum: R = 0.08, significant at the 0.34 level. Soilsaturation pH: R = -0.39, significant at the 0.022 level.

ACKNOWLEDGMENTS

We thank Gregory Bogdan for assistance in the use ofpackage computer programs.

This work was supported by research grant R-805,292 fromthe U.S. Environmental Protection Agency.

LITERATURE CITED

1. Adams, M. H. 1959. Bacteriophages. Interscience Pub-lishers, New York.

2. Akin, E. W., W. H. Benton, and W. F. Hill, Jr. 1971.Enteric viruses in ground and surface waters: a reviewof their occurrence and survival, p. 59-74. In V. Griffinand V. Snoeyink (ed.), Virus and water quality: occur-

rence and control. University of Illinois Press, Urbana.3. Bagdasaryan, G. A. 1964. Survival of viruses of the

enterovirus group (poliomyelitis, echo, coxsackie) in soiland on vegetables. J. Hyg. Epidemiol. Microbiol. Im-munol. 8:497-505.

4. Enfield, C. G., C. C. Harlin, Jr., and B. B. Bledsoe.1976. Comparison of five kinetic models for orthophos-phate reactions in mineral soils. Soil Sci. Soc. Am. Proc.40:243-249.

5. Farrah, S. R., S. M. Goyal, C. P. Gerba, C. Wallis, and

J. L. Melnick. 1977. Concentration of enterovirusesfrom estuarine water. Appl. Environ. Microbiol. 33:1192-1196.

6. Goyal, S. M., and C. P. Gerba. 1979. Comparative ad-sorption of human enteroviruses, simian rotavirus, andselected bacteriophages to soil. Appl. Environ. Micro-biol. 38:241-247.

7. Hurst, C. J., and C. P. Gerba. 1979. Development of a

quantitative method for the detection of enterovirusesin soil. Appl. Environ. Microbiol. 37:626-632.

8. Lance, J. C., C. P. Gerba, and J. L. Melnick. 1976.Virus movement in soil columns flooded with secondarysewage effluent. Appl. Environ. Microbiol. 32:520-526.

9. Lefler, E., and Y. Kott. 1974. Virus retention and survivalin sand, p. 84-91. In J. F. Malina, Jr., and B. P. Sagik(ed.), Virus survival in water and wastewater systems.University of Texas, Austin.

10. Melnick, J. L., and H. A. Wenner. 1969. Enteroviruses,p. 529-602. In E. H. Lennette and N. J. Schmidt (ed.),Diagnostic procedures for viral and rickettsial infec-tions, 4th ed. American Public Health Association, NewYork.

11. Murphy, W. H., Jr., 0. R. Eylar, E. L. Schmidt, andJ. T. Syverton. 1958. Adsorption and translocation ofmammalian viruses by plants. 1. Survival of mouse

0.0 r

-0.1 .

-0.2 k

AJ -0.3 V

-0.4 F

-0. 5

-A v0

0.0

-0. 1

-0.2

eVI-,

0-

LA, -0.33 _

-0.4 _

-0. 5

-0.6


0

I

2

.

0

on Novem


.asm.org/

Dow

nloaded from

http://aem.asm.org/


cephalomyelitis and poliomyelitis viruses in soil andplant root environment. Virology 6:612-622.

12. Nie, N. H., C. H. Hull, J. G. Jenkins, K. Steinbrenner,and D. H. Brent. 1975. SPSS statistical package forthe social sciences, 2nd ed. McGraw-Hill Book Co., NewYork.

13. Sagik, B. P., B. E. Moore, and C. A. Sorber. 1978.Infectious disease potential of land application of waste-water, p. 35-46. In State of knowledge in land treatmentof wastewater, vol. 1. U.S. Government Printing Office,Washington, D.C.

14. Schmidt, N. J., and E. H. Lennette. 1965. Basic technicsfor virology, Appendix. In F. L. Horsfall, Jr., and I.

Tamm (ed.), Viral and rickettsial infections of man, 4thed. J. B. Lippincott Co., Philadelphia.

15. Smith, E. M., M. K. Estes, D. Y. Graham, and C. P.Gerba. 1979. A plaque assay for the simian rotavirusSA1l. J. Gen. Virol. 43:513-519.

16. Tierney, J. T., R. Sullivan, and E. P. Larkin. 1977.Persistence of poliovirus 1 in soil and on vegetablesgrown in soil previously flooded with inoculated sewagesludge or effluent. Appl. Environ. Microbiol. 33:109-113.

17. Wellings, F. M., A. L. Lewis, C. W. Mountain, and IV. Pierce. 1975. Demonstration of virus in groundwaterafter effluent discharge onto soil. Appl. Microbiol. 29:751-757.

18. Yeager, J. G., and R. T. O'Brien. 1979. Enterovirusinactivation in soil. Appl. Environ. Microbiol. 38:694-701.

VOL. 40, 1980

on Novem


.asm.org/

Dow

nloaded from

http://aem.asm.org/

Documents

Effects Environmental Variables and Soil Characteristics ... · 1068 HURST, GERBA, ANDCECH minimum essential medium (14) supplemented with 10%fetal bovineserum,5%tryptose-phosphatebroth,