Brookes et al., 2005

8/3/2019 Brookes et al., 2005

http://slidepdf.com/reader/full/brookes-et-al-2005 1/8

Relative Value of SurrogateIndicators for Detecting Pathogensin Lakes and Reservoirs

J U S T I N D . B R O O K E S , * , † , ‡ , | , #

M A T T H E W R . H I P S E Y , §

M I C H A E L D . B U R C H , † , | R U D I H . R E G E L , †

L E O N G . L I N D E N , † , | , #

C H R I S T O B E L M . F E R G U S O N , † , ⊥ A N DJ A S O N P . A N T E N U C C I §

CRC for Water Quality and Treatment,PMB 3 Salisbury 5108, South Australia, SA Water Centre for Science and Systems, The University of South Australia,

Mawson Lakes, South Australia, Centre for Water Research,University of Western Australia, 35 Stirling Highway,Crawley, Western Australia, 6009, The Australian Water Quality Centre, PMB 3, Salisbury 5108, South Australia,Ecowise Environmental, P O Box 1834, Fyshwick,

Australian Capital Territory 2609, and School of Earth and Environmental Sciences, The University of Adelaide,South Australia 5005

This study investigated the relative behavior of pathogens,fecal indicator organisms, and particles of varying sizeduring transport through a reservoir following a storm eventinflow in Myponga Reservoir, South Australia. During theinflow, samples were collected from the river and at variouslocations within the reservoir to determine the fate and transport of microroganisms as they progressed through thewater body. Microbiological analysis included the indicatororganisms Escherichia coli , enterococci, Clostridium perfringens , aerobic spores, and somatic coliphages, theprotozoan pathogens Cryptosporidium spp. and Giardia spp.,

and the potential physical surrogates of pathogencontamination including particle size and turbidity. Of themicrobial indicator groups, C. perfringens spores were themost highly correlated with Cryptosporidium spp. con-centrations (Spearman Rho ) 0.58), closely followed byenterococci (Spearman Rho ) 0.57). Cryptosporidium spp.oocysts were predominantly associated with small sizedparticles (range of 14.3-27.7 µm).All of themicrobial indicatorgroups tested were associated with larger sized particleranges (>63.3 µm) except C. perfringens spores which wereassociated with particles in the size range of 45.5-63.3 µm. Although indicators may rank correlate with Cryptospo-ridium spp., the variation in settling rates of differentmicroorganisms has significant implications for the use of

surrogates to estimate pathogen attenuation withinreservoirs. For example, concentrations of Cryptosporidium spp. oocysts were reduced by a factor of 3 on reaching

the dam wall, whereas enterococci were reduced by afactor of 10.

Introduction

The detection and enumeration of protozoan pathogens in

environmental samples is complex, costly,and incurs lengthy delays due to the nature of the analytical tests. This limitsthe capacity of water utilities to confidently determine thefate and distribution of pathogens in large water bodies. Inturn, this leads to an inability to properly characterize andanticipate the risk associated with the movement of patho-gens through reservoirs to water supply offtakes following storm events, which are typically known to carry significantmicrobialcontaminant loads (1-3).The alternativehas beento consider theuse of surrogates such as microbialindicatororganisms, turbidity, or particle counting within lakes andreservoirs. There is evidence thatbacterialindicatorsof fecalcontamination (fecal coliforms and enterococci), bacte-riophages, Clostridium spp. spores, and turbidity are usefulsurrogates under defined circumstances (4). However, thereis ongoing uncertainty as to the most appropriate surrogate,

or combination of surrogates,for determiningpathogen risk in water supplyreservoirs. Identification of themechanismsthat control pathogen fate and transport in lakes andreservoirs may indicate the most suitable surrogates forprediction of risk of pathogen presence. The principalcontrolling mechanisms includesedimentation,inactivation, which is influenced by temperature, visible and ultraviolet(UV) light, and predation (4).

A small numberof studies have examined therelease andtransport of pathogens from fecal material (5 ) or theirsubsequent transport across land surfaces (6 -9 ). Thesestudies highlightthe importanceof rainfall intensity, duration,slope, and vegetation cover in determining the efficiency of pathogen transport to surfacewaters. Consequently, althoughthe microbiologicalquality of surface water inflows is highly

variable, inflows following storm events represent the majormechanism for pathogen seeding in reservoirs.

Thebehavior of inflows entering a water body is controlledby the density of the inflow water relative to that of the lake,the volume of the inflow, the slope at entry, and the bottomfriction (10 ). Warminflowswill be buoyantsurface intrusionsand cold, dense inflows will intrude as interflows or, in thecase of very dense inflows, will sink beneath the lake water where they will flow along the lake bed toward the deepestpoint. In either case the inflow will entrain water from thelake, increasing its volume, changing its density, anddiluting the concentration of pathogens and other constituents. Thespeed at which an inflowtravels through a lake, entrainment with ambient lake water, dilution, and insertion depth areall important in determining the distribution of pathogens

in lakes and reservoirs. The distribution of microorganismsis also a function of their settling rate andconsequentlymay vary for different organisms. This is further influenced by the potential for pathogens and other microorganisms todifferentially aggregate with organic or inorganic particlesand hence settle at higher rates than those of unattachedmicroorganisms.Furthermore,the inactivationof pathogensand microbial indicators is influenced by environmentalfactors, such as temperature and visible and UV light, and will therefore vary between species.

The aim of this study was to investigate the relativebehavior of pathogens, fecal indicator organisms, andparticles of varying size on entering a reservoir, and to

* Corresponding author phone: 61-8-8259 0222; fax: 61-8-82590228; e-mail: [email protected].

† CRC for Water Quality and Treatment.‡ University of South Australia.§ University of Western Australia.| The Australian Water Quality Centre.⊥ Ecowise Environmental.# The University of Adelaide.

Environ. Sci. Technol. 2005, 39, 8614-8621

8614 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 22, 2005 10.1021/es050821+ CCC: $30.25 © 2005 American Chemical SocietyPublished on Web 10/18/2005

8/3/2019 Brookes et al., 2005


determine their appropriateness as indicators for assessing Cryptosporidiumspp.concentrations.A comprehensivefieldexperiment was conducted to track the passage of a stormevent inflow through Myponga Reservoir in South Australia.During the event, samples were collected from the river andreservoir to determine the concentration of the microbialindicator organisms Escherichiacoli , enterococci, Clostridiumperfringens , aerobic spores, somatic coliphages, and theprotozoanpathogens Cryptosporidium spp.and Giardia spp.,as well as potential physical surrogates of contaminationsuch as particle size and turbidity.

Materials and Methods

Myponga Reservoir. The experiment was undertaken atMyponga Reservoir, an impounded, flooded river valley located 70 km south of Adelaide, South Australia (S 35° 24′13′′, E 138° 25′ 29′′). The reservoir has a capacity of 26.8 ×106 m3 and a maximum depth of 42 m at full supply level.The mean retention time based on abstraction rates isapproximately 3 years and the surface area is 2.8 km2. Thecatchment is approximately 124 km2 of mixed land use,including pasture for dairy, beef, and hay production, with

patchy remnant vegetation. Recent estimates of dominantland uses are 62% grazing and 24% dairying (11).

Sampling Design. A sampling program was structured totrack thepassageof a riverineintrusion, including its physicalproperties and the passage of pathogens and indicators,through the reservoir during a large runoff event following intense rainfall. The experiment was conducted from June26-29, 2003 (days of year: 178-180), during which time thetotal rainfall was 58.4 mm.

Tracking the Riverine Intrusion. The reservoir has twometeorological stations that detect wind andsolar radiationand collect data at 10-min intervals. The stations also havethermistorchains thatare usedto capture the hydrodynamicconditions within the water body. One station (Met1) islocated in the main basin and the other (Met2) is located in

a long narrow sidearm of the reservoir (Figure 1). Flow andtemperature were also measured in Myponga Creek at 10-min intervals. Flow from a second minor tributary (Sub-catchment 2) that enters near Myponga Creek was gaugedusing a logging doppler velocity meter (Starflow, UnidataPty. Ltd.) operating in one of two stormwater pipes of 1.5-minternal diameter which channel the entire creek flow for ashort distance just prior to entry into the reservoir.

The characteristics of the riverine intrusion and theambient reservoir conditions were measured by regularprofiling at several sites along the length of the reservoir with a multi-probe water quality analyzer (Hydrolab withtemperature, conductivity, dissolved oxygen, turbidity and

depth sensors) and a LISST (Laser In Situ Scattering andTransmissometry) particle profiler (12 , 13), which measuresthe relative volume of particles in each of 32 size-classesbetween 2.5 and 500 µm.

Microbiological Analysis. Microbiological sampling wasundertaken at the surface and at depth (2.5 m above thebottom) at three sites in the reservoir (Met 1, Met 2, andadjacent to the dam wall), and in the inflows (Figure 1).Sampling was timed to capture the storm event inflow andto precede and capture it within the reservoir (Table 1).

Multiple samples were collected (×3 for protozoans and×5 for all other microbes) at each location during five timeperiods (denoted T1-T5), and analyzed for the following:E. coli , enterococci, somatic bacteriophages, C. perfringens ,and aerobic spores on all five occasions, and at three times(T1, T3, and T4) for Cryptosporidium spp. and Giardia spp.(Table 1). The Australian Water Quality Centre, a NATA andISO 9000 accredited laboratory, performed the following microbiological analyses.

E. coli. E. coli was determined using Colilert 18 (IDEXX).Coliform positives were recorded as yellow wells and these weretestedfor E. coli by checkingfor bluefluorescence underUV at 365 nm.

Enterococci. Samples were filtered through a 0.45- µmmembrane before placing the membrane on m-Enterro-coccus agar and incubating at 35 °C for 48 h. Confirmationof Enterococcus species was achieved by plating of colonies

from m-Enterococcus agar onto Glucosidase agar, furtherincubation, and counting under UV.

Somatic Bacteriophages.Host E. coli brothwas inoculatedonto predriedphagebottom agar with autoclaved phage topagar containingantifoam B solution (Sigma A-5757). Sample was added to the bacterial host plate and incubated for 18h at 36 °C. Somatic coliphages were counted as clear zonesin the lawn of E. coli . A negative control of sterile diluent water and a positive control of diluted somatic coliphages were analyzed with each batch of samples.

C. perfringens and Sulfite-Reducing Clostridia. An aliquotof sample was heat-treated for 20 min. at 70 °C to eliminatenonspore-forming microorganisms prior to filtration through

FIGURE 1. Myponga Reservoir bathymetry (m) showing the inflow locations, 1 and 2, the dam wall where the offtake is situated and thelocations of the two meteorological stations including thermistor chains (Met1 and Met2).

TABLE 1. Sampling Times (Day of Year in 2003) forMicrobiological Samples

transect inflow Met2 Met1 dam wall

T1 178.042 178.042 178.104 178.146T2 178.625 178.601 178.622 178.649T3 178.702 178.688 178.733 178.875T4 179.800 179.743 179.792 180.042T5 180.571 180.500 180.542 180.604

VOL. 39, NO. 22, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 8615

http://dontstartme.literatumonline.com/action/showImage?doi=10.1021/es050821+&iName=master.img-000.jpg&w=325&h=149

8/3/2019 Brookes et al., 2005


a 0.45- µm membrane and placement of the membrane ona plate of Tryptose Sulfite Cycloserine agar. Incubation for24 h was at 36 °C in an atmosphere of H2 and N2. Samples were Gram-stained, and positive C. perfringens verified withMUP-ONPG (4-methyl-umbelliferyl phosphate + o-nitro-phenyl).

Aerobic Spores. Samples were heat-treated at 80 °C for 30min and then filtered through a 0.45- µm membrane and

placed onto Trytone Soya Agar. After 48 h incubation at 30°C, colonies were selected for Gram staining. Aerobic spores were identified as Gram positive bacilli.

Cryptosporidium spp. and Giardia spp. Primary Concen-tration: Samples (10-L) were concentrated by CaCO3 floc-culation (settled overnight and the supernatant aspirated, with recovery of the settled particulates by dissolving theCaCO3 using sulfamic acid) or by filtration using Envirochek filter capsules. ColorSeed (BTF Decisive Microbiology) wasadded to each sample as an internal control for eachprotozoan assay to allow determination of oocyst/cystrecovery efficiency.

Secondary Concentration: The material collected in theprimary concentration was further concentrated by cen-trifugation to reduce the pellet to approximately 10 mL. All

samples were processed with the Dynal IMS kit following standard instructions.

Staining and Microscopic Detection: The final concen-trates were stained on Dynal spot-on slides using the DNA-staining fluorochrome DAPI and FITC-conjugated mono-clonal antibodies against bothCryptosporidiumand Giardia.The samples were scanned on glass microscope slides.Fluorescence microscopy, using blue-light excitation, wasused to scan the slides for characteristic FITC-fluorescenceof cells with the size and shape characteristics of Cryptospo-ridium or Giardia. Identification of organisms wasconfirmedusing DAPI fluorescence (UV excitation) to judge structuralintegrity and countnuclei, and DIC (differential interferencecontrast) to determine internalstructuressuch as sporozoitesin Cryptosporidium and median bodies in Giardia, particu-

larly in DAPI-negative cells.Data Analysis. Statistical analysis was performed using

the JMPstatisticalpackagefrom the SASInstitute. The degreeof association among Cryptosporidium spp. concentration,the concentration of surrogate microorganisms, and indi-vidual particlesize classes wasdetermined by Spearmanrank correlation using corresponding concentrations of eachparameter for all sampling times, sites, and depths. TheSpearmancorrelation comparesrelative similarity by ranking thevaluefor each sampling anddeterminingthe correlationbetween the ranked observations. Higher Spearman Rhovalues indicate a stronger correlation. The critical Rho for n) 63, significance of R ) (0.05) is 0.248.

Results

Inflow Characteristics. The high rainfall early on day 178(42 mm) generated considerable runoff and elevated flow inMyponga Creek. The measured maximum total flow was 9.8m3s-1 recorded after middayof day178 (Figure2). Theriverine water entered the reservoir as an underflow since it wassubstantially cooler, contained more solutes, and thereforehada greaterdensitythanthe ambientreservoir water (Figure3).Therewas a general warming of theriver water duringtheperiod, which acted to decrease the density of the incoming riverinewater.Nevertheless,the inflow signal moved throughthereservoir as a distinct pulse over 36 h. Theunderflow wasfirst detected at Met2 in the sidearm as a general cooling of the bottom waters at 12:10 on day 178, approximately 9 hafter the Myponga Creek hydrograph began to rise. Cooling at depth occurred subsequently at Met1 in the main basin,approximately 4 h after theintrusion passed Met2. Theinflow pulse was first seen at the dam wall during T4, 36 h after thehydrograph peak in the inflow stream.

Occurrenceand Magnitude of Pathogens and Surrogatesin the Intrusion. Both theMyponga Creek andSubcatchment2 inflows delivered substantial concentrations of Cryptospo-ridium spp. to the reservoir. Concentrations in Myponga

Creek ranged between 16 and 100 oocysts (10 L -1) andconcentrations from Sub-catchment 2 were 19-95 oocysts(10L-1). As theriverine intrusion movedthrough thereservoir,Cryptosporidium spp., the main pathogen of interest, wassequentiallydetected at lower concentrations in the bottomdepth samples at locations further from the inflow (Table2a). Very low concentrations of Giardia spp. were found,although they tendedto also occur in thebottom waters that were influenced by the riverine intrusion (Table 2b).

The other microbiological indicator organisms, (E. coli ,Enterococci , somatic bacteriophages, C. perfringens , andaerobic spores) all displayed elevated concentrations withinthe riverine intrusion (Table 2c-h).As with Cryptosporidiumspp., therewas a relativedecreasein surrogate concentrationas the intrusion travelled through the reservoir; the rate of

dilution, however, differed among the various organisms(Table 2).

Thegeneraltrendevidentfor allpathogens andsurrogates was that they were in high concentrations in the incoming river water and appeared in the reservoir at the sitesinfluenced by theriverine intrusion. The data were analyzedby nonparametric Spearman rank correlation to determine which of the microbial indicators had fate and transportcharacteristicsmost similar to thoseof Cryptosporidiumspp.(Table 3). Giardia spp. was found in the most similar rank order to Cryptosporidium spp. but the very low Giardia spp.concentrations reduce the confidence in these correlations.Of the fecal indicator groups, C. perfringens spores and

FIGURE 2. Flow rate and temperature of Myponga Creek and the Subcatchment 2 tributary monitored during the inflow event in June2003, days 178-182. The average reservoir temperature was calculated from thermistors suspended vertically at 20 depths; on day 178 the reservoir was fully mixed.

8616 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 22, 2005

http://dontstartme.literatumonline.com/action/showImage?doi=10.1021/es050821+&iName=master.img-001.png&w=332&h=133

8/3/2019 Brookes et al., 2005


Enterococci showed the highest correlation with Cryptospo-ridium spp. oocyst concentrations (Spearman Rho ) 0.58and 0.57, respectively). Concentrations of the traditionalindicator organism E. coli showed a weaker correlation withCryptosporidium spp. (Spearman Rho ) 0.48), although this was higher than the correlation with either somatic bacte-riophages or aerobic spores (Spearman Rho ) 0.34 and0.21,respectively).

The correlation analysis also indicates the strength of relationship amongall indicators,to determine whether onesingle surrogate was representative of the other surrogates, within the context of transport, settling, and decay losses within a riverine intrusion. E. coli and Enterococci were themost highly correlated (Spearman Rho ) 0.92). This meansthatthroughoutthe hydrograph,and transportin the riverineintrusion, these two organisms displayed the most similarrelativechanges in concentration dueto settlingand dilution.The organism that showed the most consistently highcorrelation with indicators other than Cryptosporidium spp. was E. coli . C. perfringens spores showed the weakestcorrelation with E. coli (Spearman Rho ) 0.60). This is

interesting given that both C. perfringens and Cryptospo-ridium spp. showed the highest rank correlations.

Thedata forthe microbialindicators were also correlatedagainst particle size to try to deduce their behavior relativeto particles. This analysis correlates each particle size class within the selected bins of the LISST profiler with eachorganism for all sampling times and depths. All organisms

correlated moderately well with particle volume concentra-tions ranging between 4.48 and 259 µm (Figure 4) becausethey shared the intrusion as the dominant controlling mechanism. The strongest correlations of organisms withparticle size classes were aerobic spores (74.7-128 µm),somaticbacteriophages(45.5-74.7 µm), C. perfringens spores(45.5-63.3 µm); E. coli and Enterococci (74.7-157 µm), andsulfite-reducing clostridia(63.3-128 µm).This contrastswithCryptosporidium spp., which wasmost highly correlated withsmaller particles in the range 16.8-27.7 µm (Spearman Rho> 0.70). Cryptosporidium spp.showedonlya weak correlation with turbidity (Spearman Rho ) 0.52) and total particlevolume concentration (SpearmanRho ) 0.55).All surrogates

FIGURE 3. Temperatures (°C) measured during the five transects as indicated by the day of year, preceded by year (yyyyddd). Transectswere constructed by contouring around the profile locations (indicated by the dot-dash lines) using a rectangular grid with a sigmacoordinate transformation.


http://dontstartme.literatumonline.com/action/showImage?doi=10.1021/es050821+&iName=master.img-002.jpg&w=333&h=452

8/3/2019 Brookes et al., 2005


showedvery poor correlation with particlesgreater than 332 µm.

DiscussionPathogen Similarity. This study measured a variety of microbiological and physical parameter concentrations andexamined theircorrelations with each otherwithin a riverineintrusion. The Spearman rank nonparametric correlationprocedure was used to determine how similarly pathogens,surrogates, and particles behaved throughout the course of the experiment. However, a high correlation does not

unequivocally provethat a particular surrogateof a pathogenis universally applicable. In this case the pathogens Cryptospo-ridium spp. and Giardia spp.were the mosthighly correlatedbiological parameters (Spearman Rho ) 0.65). Although thedata and relationships related to Giardia spp. in thisexperimentmust be treated with caution due to thevery low concentrations,the resultswere similar to those of Fergusonet al. (14) who reported a partial correlation coefficient of 0.73 (p < 0.05) between Giardia spp. and Cryptosporidiumspp. in waters of the Georges River, Sydney. Of the othermicrobial indicator groups Cryptosporidium spp. was mostclosely correlated with C. perfringens spores, albeit weakly (Spearman Rho ) 0.58). This was also comparable to theresultsof Fergusonet al. (14), whoreported that C. perfringens spores were the microbial indicator group most highly

correlated with Cryptosporidium spp. and Giardia spp.(CorrelationcoefficientR valuesof 0.27and 0.41,respectively).In their study other fecal indicator groups showed weak correlations with Cryptosporidium spp. (fecal coliforms R )

0.03, fecal streptococci R ) 0.08, F-RNA bacteriophages R) -0.39).

All microorganisms experience the same dilution due toentrainment within the inflow, implying the differencesbetween microrganisms as they progress through the res-ervoir are due to differential inactivation or sedimentationrates. The time for the intrusion to move from the inflow tothe offtake was less than 36 h, and considering the cooltemperatures (<10 °C) and high light attenuation (due tohigh dissolved organic carbonconcentrations, >10mg DOCL-1), there waslikely to belittle inactivationduring that time.

Consequently, the major difference between organisms isdeemed to be sedimentation, and this data set thereforeemphasizes that the existence of similar settling character-istics is important when seeking appropriate surrogates forpathogens.The different settling ratesexhibited by the variousmicroorganismsreflect theirvariedsurface chargeand affinity for attaching to particles, as well as the size and density of the particle aggregate. For example, the present study indicatesthatoocysts tendnot to associate withlarger/heavierparticles and therefore settling provides little additionalattenuation than dilution alone, whereas E. coli was attenu-ated to a much greater extent. Hipsey et al. (15 ) also founda similar high attachment and settling rate for E. coli inSugarloaf Reservoir, near Melbourne, Australia.

Turbidity as a Surrogate for Pathogen Transport. It isfeasible that if the size of the particle and microorganismaggregate is known, then turbidity may well be a usefulsurrogate. As indicated above, the aggregation of particles with organic matter or clay minerals plays a major role inpathogen/surrogate transport,settling (15 ),andsurvival(16 ).Thesurfacecharge of theparticles plays a keyrole in particle-

particle interactions in water (17 ). It appears that Cryptospo-ridium spp. oocysts may be adequately aggregated andflocculated during conventional water treatment but may not adsorb well to natural negatively charged clays withinthe natural environment (18, 19 ). Considine et al. ( 20 , 21)supported this hypothesis but concluded thatprotein linkedtethering between silica and oocysts can occur and may facilitate adhesion.Although thissuggests thatoocystswould

TABLE 2. Mean Microorganism Measurements for the FiveTransects (T1-T5)

T1 T2 T3 T4 T5

(a) Cryptosporidium spp. (oocysts/10 L)inflow Myp Crk 49 100 33 16

Subcatch 2 19 71 95Met2 bottom 0 41 81

surface 0 2 0Met1 bottom 2 2 25

surface 0 0 2

dam wall bottom 0 2 33surface 0 2 1

(b) Giardia (cysts/10 L)inflow Myp Crk 3 3 1 2

Subcatch 2 0 0 2Met2 bottom 0 3 3

surface 0 0 0Met1 bottom 0 0 2

surface 0 0 0dam wall bottom 0 0 1

surface 0 0 0

(c) E. coli (/100 mL)inflow Myp Crk 4067 22667 9000 5233 4433

Subcatch 2 2400 20000 23000 2600 3700Met2 bottom 23 1933 2767 5300 1633

surface 129 25 77 63 567Met1 bottom 26 75 35 1900 547

surface 270 42 37 34 55dam wall bottom 33 116 95 1267 1030surface 27 15 21 34 78

(d) Aerobic Spores (/100 mL)inflow Myp Crk 21667 176667 320000 26333 24667



surface 3867 1433 10667 2150 2433dam wall bottom 5400 3367 19333 21000 4933

surface 4933 2633 92333 2667 2533

(e) Somatic Bacteriophages (/100 mL)inflow Myp Crk 167 630 247 777 363


surface 14 23 11 25 28

Met1 bottom 11 12 7 107 10surface 27 11 10 7 6

dam wall bottom 10 16 12 76 14surface 13 9 7 12 4

(f) C. perfringens spores (/100 mL)inflow Myp Crk 109 107 57 87 40




surface 1 0 2 0 0

(g) Enterococci (/100 mL)inflow Myp Crk 4800 12400 7233 1950 737




surface 27 11 12 27 53

(h) Sulfite-Reducing Clostridia (/100 mL)inflow Myp Crk 173 267 257 141 62




surface 17 19 15 13 17


8/3/2019 Brookes et al., 2005


not adsorb readily to inorganic particles, there is some

evidence of high settling velocities for Cryptosporidium spp.in field monitoring situations ( 22 , 23). This suggests that, atleast in certain situations, oocysts must be associated withlarger particles, andit is possible thatoocystsshow an affinity for organic (e.g., sewage effluent) rather than inorganicparticles. An alternative is that theoocysts maybe physically enmeshed within an organic matrix of fecal material and/orsoil particles during entrainment in surface water runoff,although Brookes et al. ( 24) found that this effect was notsignificant. By contrast, viruses appear to readily attach toparticles ( 25 - 27 ), and this is reflected by their greaterreduction relative to dilution over the course of the experi-ment.

Thecorrelationof Cryptosporidium spp. andthe microbialindicators in this study with different particle size classesindicated that Cryptosporidium spp. tends to be transportedsimilarly to small particles, whereas the bacterial and viralindicators tend to be transported like the relatively largeparticles. The association of microorganisms with particlesis governed by their electrophoretic mobility which varies with strain ( 28) and species, pH, organic matter, and ionicstrength ( 29 ). In field situations these parameters will vary with soil type, as will particle size and soil composition.Consequently, some variations in the relationship betweenmicroorganisms and particle size would be expected indifferent catchments and reservoirs. This studysuggeststhatfor a surrogate to be a useful indicator of pathogen risk inreservoirs it must share similar aggregation and settling characteristics to the pathogen of concern. The use of

nonspecific measures of particle concentration, such as

turbidityor totalparticlevolume concentration, in monitoring may provide potentially erroneous predictions of pathogenrisk dependingon the particle size distribution.Nonetheless,a turbidity pulse will serve as an early warning of a potentialpathogen threat if the origin of the turbidity and pathogeniccontamination is similar.

Implications for Monitoring and Pathogen Risk Assess-ment. The storm event that was monitored as part of thisexperiment delivered a considerable pathogen load toMyponga Reservoir in a distinct intrusion that travelledrapidly through the reservoir. This intrusion arrived at theofftake location at the dam wall within 36 h of the com-mencement of rainfall in the catchment. The response timefor a normal routine microbialmonitoringprogram to collectsamples, return them to a laboratory for analysis, process

the samples, and generate results for interpretation typically approaches 48 h. The detection of the inflow containing pathogens at the offtake point in less than 48 h following amodest rain event in this catchment highlights the need toreassess the ability of monitoring programs to adequately gauge pathogen risks. The use of surrogates has frequently been discussed as a more rapid alternative to measurementof pathogen concentrations to estimate pathogen transport,fate, and ultimately risk. However, for surrogates to be of value they must be rapidly measurable, representative of thepathogen of concern, and display similar fate and transportcharacteristics. For the case presented here, the commonly used coliform indicators would have significantly underes-

TABLE 3. Spearman Rank Correlation Among Microbiological Surrogates

Cryptosporidium aerobicspores

somaticbacteriophages

C. perfringens spores E. coli Enterococci

sulfite-reducingClostridia

Cryptosporidium 1.00 0.21 0.34 0.58 0.48 0.57 0.39aerobic spores 0.21 1.00 0.53 0.41 0.82 0.84 0.86somatic bacteriophages 0.34 0.53 1.00 0.72 0.78 0.66 0.72C. perfringens spores 0.58 0.41 0.72 1.0 0.60 0.60 0.74E. coli 0.48 0.82 0.78 0.60 1.00 0.92 0.84Enterococci 0.57 0.84 0.66 0.60 0.92 1.00 0.91sulfite-reducing Clostridia 0.39 0.86 0.72 0.74 0.84 0.91 1.00

FIGURE 4. Spearman rank correlation of microorganisms with particle size as measured by the LISST profiler.


http://dontstartme.literatumonline.com/action/showImage?doi=10.1021/es050821+&iName=master.img-003.png&w=318&h=254

8/3/2019 Brookes et al., 2005


timated the actual risk of Cryptosporidium spp. concentra-tions.

The best and safest strategy for risk assessment of any given pathogen in a water supply reservoirs is to sample andanalyze for the target organism in a systematic way, whiletaking into account the biological, physical, and chemicalfactors that control transport, distribution, and inactivationin naturalsourcewaters. Modelingtoolshave beendevelopedto assess the inflow characteristics of a riverine intrusion toa reservoir prior to sampling for pathogens or surrogates.The minimum sampling strategy can be designed with

assistance fromthe web-basedmodel INFLOW (30 , availableat http://www.cwr.uwa.edu.au/∼ttfadmin/model/inflow/).This model uses flow and temperature of the river inflow,reservoir bathymetry, and density profiles to calculate thetransport time, dilution, and depth of the inflow. A prioriinformation of where the inflow can be located within thereservoirwill enable samplingto target thewater withhighestpathogen concentrations for quantitative risk assessment.For example, in Myponga Reservoir the routine operationalsampling program has involved the sampling of surface waters between locations Met2 and Met1 (Figure 1). It isclear from this study that the inflow had a negligible impacton the surface waters near this location (see Figure 3), andthese data would be falsely negative and provide incorrectinputs to any quantitative microbial risk assessment.

This study investigated the potential for microbial indica-

tor organisms and particle concentrations (i.e., surrogates)to be utilized for pathogen risk assessments in reservoirs.The results indicate that microorganisms do not behave assimilarly as would be expected based on hydrodynamictransport alone. The different settling rates of the microor-ganisms have significant implications for the attenuation of pathogens within the reservoir. For example, Cryptospo-ridium spp. oocysts were reduced bya factorof 3 on reaching the dam wall, whereas the Enterococcus spp. concentration was reduced by a factor of 10.

It is unlikely that any single water quality indicator willreliably assess the risk of bacterial, protozoan, and viralcontamination in aquatic environments, but it is feasiblethat a suite of surrogates can be used to estimate levels of microbialcontaminationwithin definedcircumstances, such

as within a storage reservoir with well-characterized inputs.However, this study highlights that surrogates should becarefully selected to ensure that they represent an accurateassessmentof pathogen risks. While surrogates mayprovidea useful indicator of fecal contamination, their use forquantitative microbial riskassessmentis complicated by theirdifferent kinetic, and settling characteristics relative topathogens. Although turbidity and total particle volumedisplayeda relativelyweak correlationwith Cryptosporidiumspp. concentrations,they arestillusefulparameters that canbe measured on-line to indicate the presence and locationof a riverine intrusion, enabling water quality managers toreact to catchment rainfall events in near real-time.

AcknowledgmentsThe research has been made possible through funding from

the AmericanWater WorksAssociation Research Foundation(AwwaRF Project 2752),the CooperativeResearch Centre for WaterQuality and Treatment (Project 2.2.0.1), and Melbourne WaterCorporation.The projectteam thanksWarwick Grooby,Kerry Carabretta, and the Microbiology and Protozoology laboratories at the Australian Water Quality Centre, formicrobiological analysis. Rod Boothy, Alan Brown, and Viv Allan provided excellent logistical support at MypongaReservoir.

Literature Cited(1) Atherholt, T. B.; LeChevalier, M. W.; Norton, W. D.; Rosen, J. S.

Effect of rainfall on Giardia and Crypto. J. Am. Water Works Assoc. 1998, 90 , 66-80.

(2) Boyer, D.G.; Kuczynska, E.Stormand seasonal distributions of fecal coliforms and Cryptosporidium in a spring. J. Am. Water Resour. Assoc. 2003, 39 , 1449-1456.

(3) Kistemann, T.; Claben, T.; Koch, C.; Dangendorf, F.; Fischeder,R.; Gebel, J.; Vacata, V.; Exner, M. Microbial load of drinking

water reservoir tributaries during extreme rainfall and runoff. Appl. Environ. Microbiol . 2002, 68, 2188-2197.

(4) Brookes, J. D.;Antenucci, J.;Hipsey, M.R.; Burch,M. D.;Ashbolt,N. J.; Ferguson, C. M. Fate and transport of pathogens in lakesand reservoirs. Environ. Int. 2004, 30 , 741-759.

(5) Bradford, S. A.; Schijven, J. Release of Cryptosporidium andGiardia from dairy calf manure: impact of solution salinity.

Environ. Sci. Technol. 2002, 36 , 3916-

3923.(6) Mawdsley, J. L.; Brooks, A. E.; Merry, R. J.; Pain, B. F. Use of anovel soil tilting tableapparatus to demonstrate the horizontaland vertical movement of the protozoan pathogen Cryptospo-ridium parvum in soil. Biol. Fertil. Soils 1996, 23, 215-220.

(7) Atwill, E.R.; Hou, L.;Karle,B. M.;Harter,T.;Tate,K. W.;Dahlgren,R. A. Transport of Cryptosporidium parvum oocysts throughvegetatedbuffer strips and estimatedfiltration efficiency. Appl.Environ. Microbiol. 2002, 68, 5517-5527.

(8) Oliver, D. M.; Clegg, C. D.; Haygarth, P. M.;. Heathwaite, A. L.Determininghydrologicalpathways forthe transferof potentialpathogens from grassland soils to surface waters. In 7thInternational Conference on Diffuse Pollution and BasinManagement, Dublin, Ireland 2003; pp 3.36-33.41.

(9) Davies, C. M.; Ferguson,C. M.;Kaucner, C.;Altavilla, N.; Deere,D.A.; Ashbolt,N. J. Dispersionand transport of Cryptosporidiumoocysts from fecal pats under simulated rainfall events. Appl.Environ. Microbiol. 2004, 70 , 1151-1159.

(10) Imberger, J.; Patterson, J. C. Physical limnology. InAdvances in Applied Mechanics ; Hutchinson,J. W.;Wu, T. Y.,Eds.; AcademicPress: Cambridge, 1989; 303-475.

(11) Thomas, D.; Kotz, S.; Rixon, S. Watercourse survey and manage-ment recommendations for the Myponga River catchment ;Environment Protection Agency: Adelaide , 1999.

(12) Agrawal, Y. C.; Pottsmith, H. C. Instruments for particle sizeand settling velocity observations in sediment transport. Mar.Geol. 2000, 168, 89-114.

(13) Gartner, J. W.; Cheng, R. T.; Wang, P.; Richter, K. Laboratory andfieldevaluations of theLISST-100instrumentfor suspendedparticle size determinations. Mar. Geol. 2001, 175 , 199-219

(14) Ferguson, C. M.; Coote, B. G.; Ashbolt, N. J.; Stevenson I. M.Relationshipsbetweenindicators, pathogens and water quality in an estuarine system. Water Res. 1996, 30 , 2045-2054.

(15) Hipsey,M. R.;Brookes,J. D.;Regel,R. H.;Antenucci,J. P.;Burch,M. D. In situ evidence for the association of Total Coliforms

and Escherichia coli with suspended inorganic particles in an Australian reservoir. Water Air Soil Pollut. (in press).

(16) LaBelle, R. L.; Gerba, C. P. Influence of estuarine sediment onvirus survival under field conditions. Appl. Environ. Microbiol.1980, 39 , 749-755.

(17) Ongerth, J. E.; Pecoraro, J. P. Electrophoretic mobility of Cryptosporidium oocysts andGiardia cysts. J. Environ. Eng.1996,122 , 228-231.

(18) Dai, X.; Boll, J. Evaluation of attachment of Cryptosporidiumparvum and Giardia lamblia to soil particles. J. Environ. Qual.2003, 32 , 296-304.

(19) Hipsey, M. R.; Antenucci, J. P.; Brookes, J. D.; Burch, M. D.;Regel, R. H.; Linden, L. A three-dimensional model of Cryptospo-ridium dynamics in lakes and reservoirs: a new tool for risk management. Intl. J. River Basin Manage. 2004, 2 , 181-197.

(20) Considine,R. F.; Dixon, D. R.;Drummond, C. J. Laterally-resolvedforce microscopy of biological microspheres-oocysts of

Cryptosporidium parvum. Langmuir 2000, 16 , 1323-1330.(21) Considine, R. F.; Drummond, C. J.; Dixon, D. R. Force of

interaction between a biocolloid and an inorganic oxide:complexity of surface deformation, roughness, and brushlikebehavior. Langmuir 2001, 17 , 6325-6335.

(22) Hawkins, P. R.; Swanson, P.; Warnecke, M.; Shanker, S. R.;Nicholson, C. Understanding the fate of Cryptosporidium andGiardia in storage reservoirs: a legacy of Sydney’s watercontamination incident. AQUA 2000, 496 , 289-306.

(23) Medema, G. J.; Schets, F. M.; Teunis, P. F. M.; Havelaar, A. H.Sedimentation of free and attached Cryptosporidium oocystsand Giardia cysts in water. Appl. Environ. Microbiol. 1998, 64,4460-4466.

(24) Brookes, J. D.; Davies, C. M.; Hipsey, M. R.; Antenucci, J. P.Determining the size of Cryptosporidium-associated particles


8/3/2019 Brookes et al., 2005


from cowfaeces undersimulated rainfall andrisk reductionby settling within watersupply reservoirs. Water Health(in press).

(25) LaBelle,R. L.;Gerba, C. P. Influenceof pH,salinity, andorganicmatter on the adsorption of enteric viruses to estuarinesediment. Appl. Environ. Microbiol. 1979, 38, 93-101.

(26) Gantzer, C.; Gillerman, L.; Kuzetsov, M.; Oron, G. Adsorptionand survival of faecal coliforms, somatic coliphages andF-specific RNA phages in soil irrigated with wastewater. Water Sci. Technol . 2001, 43, 117-124.

(27) Lipson, S. M.; Stotzky, G. Adsorption of reovirus to clay minerals: effects of cation-exchange capacity, cationsaturation,and surface area. Appl. Environ. Microbiol. 1983, 46 , 673-682.

(28) Lytle, D. A.; Rice, E. W.; Johnson C. H.; Fox, K. R. Electro-

phoretic mobilities of Escherichia coli O157:H7 and wild-type

Escherichiacoli strains. Appl.Environ. Microbiol.1999, 65 ,3222-

3225.

(29) Lytle, D.A.; Johnson,C. H.;Rice, E.W. A systematic comparisonof the electrokinetic properties of environmentally importantmicroorgansims in water. Colloids Surf., B 2002, 24, 91-101

(30) Antenucci, J. P.;Brookes,J. D.;Hipsey M. R. A simplemodel fordeterminingpathogen transport anddilution throughlakes andreservoirs. J. Am. Water Works Assoc. 2005, 97 , 86-93.

Received for review April 29, 2005. Revised manuscript re-ceived September 1, 2005. Accepted September 9, 2005.

ES050821+


Documents

Brookes et al., 2005