Printed in Great Britain 0043-1354/00/$ - see front matter PII: S0043-1354(00)00291-8
EVALUATION OF TWO CONCENTRATION METHODS FOR
DETECTING GIARDIA AND CRYPTOSPORIDIUM IN WATER
BING-MU HSU1, CHIHPIN HUANG1*M, YEONG-FUA HSU2, GUO-YING JIANG1, AND CHIH-LI LILIAN HSU3
1Institute of Environmental Engineering, National Chiao Tung University, 75, Po-Ai street, Hsinchu,
Taiwan, ROC;2National Institute of Environmental Analysis Environmental Protection Administration, Taiwan, ROC and 3Department of Parasitology, Medical College National Cheng Kung University,
Tainan, Taiwan, ROC
(First received 24 August 1999; accepted in revised form 23 May 2000)
Abstract}The cartridge filtration method and membrane filtration method based on the fluorescent antibody procedure were evaluated for their recovery efficiencies and detection limits of Giardia and Cryptosporidium. We assessed the performances of the two concentration methods for water samples collected from Taiwan water treatment plants. The membrane filtration method was characterized by higher recovery rate and detection limit comparing with the cartridge filtration method. The occurrences of both parasites, and the relationships of parasite concentrations with indicator microorganisms show inconsistency between the two methods. It was discovered that water turbidity reduced the recovery efficiencies, and raised the detection limits for both parasites regardless of the method used. # 2000 Elsevier Science Ltd. All rights reserved
Key words}Giardia, Cryptosporidium,water treatment, ICR
INTRODUCTION
Protozoan Parasites, Giardia and Cryptosporidium, have been recognized as common pathogenic proto-zoa of the gastrointestinal tract (Cook, 1995). Many outbreaks of giardiasis and cryptosporidiosis have been reported in the last few decades (Frost et al., 1996; SoloGabriele and Neumeister, 1996; Kramer et al., 1996). Water is perhaps the major route for massive outbreaks of pathogen infection, as a result of contamination of either raw or treated water (Teunis et al., 1997; Eisenberg et al., 1998; Perz et al., 1998). The occurrence of Giardia and Cryptospor-idium in drinking water has aroused attention on detection of the protozoa at levels concerning human health.
Methods available for the detection of Giardia and Cryptosporidium in water sample generally contains three stages: sample concentration and elution; separation of cysts and oocysts from other debris; staining and identification of the protozoan parasites. Information Collection Requirement Rule (ICR) proposed a standard method for detecting Giardia and Cryptosporidium in water samples by the
fluorescent antibody procedure (USEPA, 1995), which, however, has been heavily scrutinized by many researchers, such as Clancy et al. (1994) and LeChevallier et al. (1995). Investigation shows that the cartridge filters concentration procedure and Percoll-sucrose flotation technique are the primary causes for the poor recovery performances. Method 1622 (USEPA, 1997,1999a) and Method 1623 (USE-PA, 1999b), with an improved concentration proce-dure and adoption of immunomagnetic separation (IMS), are expected to have higher recovery and lower detection limit. However, most water utilities still rely on the ICR method and its forerunner} ASTM P229 or modification regardless of its high variability and low sensitivity. It is because that Method 1622 and 1623 is only applicable at low turbidity as well as its high cost of equipment and materials (Clancy and Hansen, 1999; Atherholt and Korn, 1999).
The aim of this study was to evaluate the two concentration methods, the cartridge filtration pro-posed in ICR protozoan method and the membrane filtration recommended by Method 1622 draft, based on their sensitivities, recovery efficiencies and detec-tion limits on cysts and oocysts. The correladetec-tion between the densities of protozoan parasite and some indicator microorganisms were determined, and the effect of turbidity on these two methods was also evaluated.
*Author to whom all correspondence should be addressed. Tel.: +886-3-5726-463; fax: 886-3-572-5958; e-mail: [email protected]
MATERIALS AND METHODS
Preparation of protozoa-contained sample and collection of water sample
The cysts and oocysts were obtained from Waterborne, Inc. (Louisiana, USA) and the Pleasant Hill Farms (Idaho, USA), respectively, and were diluted to desired concentra-tions by 0.1% PBS as the stock solution. The numbers of cysts and oocysts in the stock solution were counted before each seeded experiment using the immunofluorescence assay technique. To do this, samples were mixed thoroughly and pipetted directly from the stock preparation vial and then onto the glass slides (Dynal1
Spot-On), stained with fluorescent-labelled antibodies (HydrofluorTM Combo Giardia/Cryptosporidium; Ensys, Inc., NC, USA), and counted. Water samples for seeding were taken from the Hsin-Chu water treatment plant, located in northern Taiwan. The waters were seeded with 103–104/l of cysts
and oocysts to determine the recovery efficiencies. To detect the occurrence of parasites in Taiwan, 20 raw water samples and 20 treated water samples from five public water treatment plants, as well as 12 raw water samples from six simple water facilities were collected.
Concentration, elution and clarification technique
Cartridge filtration apparatus include a 25.4 cm long and 1.0 mm nominal-pore-size polypropylene yarn-wound car-tridge filter (M39R10A; Commercial Filters Parker Hanni-fin Corp., IN, USA) with LT-10 filter holder. Membrane filtration apparatus include polycarbonate membrane (142 mm diameter, pore size 3 mm; Catalog No. 16565; Osmonics Inc., USA) with 142 mm stainless-steel filter holder (Millipore Corp., MA, USA).
Aliquots of seeded samples and unseeded samples were filtered through cartridge filters. The filters were removed from the cartridge, cut off from the supporting core, and divided into three parts. The filter fibers were eluted with 0.7–1.2 l eluting fluid (phosphate-buffered saline, 1% Tween 80, 1% SDS) in a mechanical stomacher, and then the eluate was centrifuged at 1050 g for 10 min in a 50 ml centrifuge tube in swinging-bucket rotor. The volume of the packed pellet was recorded. After the supernatant was aspirated, the pellet was resuspended in an equal volume of 10% formalin and the eluting fluid was added to a total volume of 20 ml in the same centrifuge tube. The mixture was under-laid with 30 ml of Percoll-sucrose gradient (sp. gravity=1.10) fol-lowed by centrifuging. The top 20 ml and 5 ml below the interface were collected, diluted with eluting fluid to 50 ml, and centrifuged. The upper-layer liquid was then aspirated till only 5 ml of concentrate was left.
The same water samples as those used in the cartridge filtration method were also passed through 142-mm-diameter polycarbonate membranes in a stainless-steel filter holder. Protozoan parasites were removed from the membrane by scraping, followed by washing with the eluting fluid (phosphate-buffered saline, 1% Tween 80, 1% SDS). The eluate was transferred into a 50 ml conical centrifuge tube, and centrifuged for 15 min at 1050 g for 10 min. After aspirated the top 45 ml solution, the remaining pellet was preserved with equal volume of 10% formalin. Clarification was done by Percoll-sucrose gradient, as described previously.
Staining and examination
The cysts and oocysts in the water samples were stained and examined following the procedures described in the ICR protozoan method for detecting Giardia cysts and Cryptosporidiumoocysts in Water by a fluorescent antibody procedure (USEPA, 1995). To do this, the re-suspended sediment samples were applied to each 25 mm diameter cellulose–acetate membrane, stained with
fluorescent-labelled antibodies (HydrofluorTM-Combo Giardia/Crypto-sporidium; Ensys, Inc., NC, USA), and examined with epifluorescent microscope at 200 , 400 or 1000 magnification (Olympus, Japan). The antibody, Hydro-fluorTM-Combo Giardia/Cryptosporidium, can react with all species of Giardia as well as Cryptosporidium parvum, Cryptosporidium muris, and Cryptosporidium meleagridis. Cysts and oocysts were identified using following para-meters: size, shape, surface feature, and staining reaction. The candidates exhibiting right size and shape were further identified by their internal morphological features with epifluorescent microscopy under a bright field.
Data analysis and calculation
To investigate the recovery efficiency of the cartridge filtration method and membrane filtration method, we visually counted the numbers of cysts and oocysts of the seeded water samples before filtration and after clarification through epifluorescent microscope. The detection limit of each water sample was calculated as described in equation (1) (USEPA, 1995).
X¼ 100
VFR ð1Þ
where X is the detection limit of cysts/oocysts per 100 l of the water sample, V is the water volume, F is the fraction of the initial eluate packed-pellet volume subjected to clarifica-tion, R is the percentage of floated sediment examined. The numbers of cysts and oocysts observed under the micro-scope were recorded, multiplied by the detection limit and divided by the recovery efficiency to obtain the final counts. When no protozoa was observed under the epifluorescent microscope, we expressed the protozoa concentration in water samples as less than (5) their detection limits. In the statistical evaluation, data under the detection limits were treated as zero. Water turbidity was measured using a ratio turbidimeter (HACH, Co., USA). Heterotrophic bacteria were measured by the spread method (APHA, 1995), while other indicator microorganisms (total coliforms, fecal coli-forms, E. coli, and Enterococcus) were measured by membrane filtration procedures described by the standard method for the examination of water and wastewater (Methods 9222 B in APHA, 1995). Spearman R correlation coefficients were calculated between the concentrations of cysts/oocysts and the indicator microorganisms, and be-tween the detection results from cartridge filtration method and membrane filtration method, using the STATISTICA software (StatSoft, Inc., USA).
RESULTS AND DISCUSSION
Recovery efficiency of two concentration methods in seeded water
The mean recovery efficiencies of protozoan parasites concentrated by two methods are shown in Fig. 1. The recovery efficiencies of concentration by cartridge filter for 40 l of raw and treated water were 28.4 11.0% (n ¼ 4) and 30.0 11.7% (n ¼ 8) for cysts, 9.3 3.3% (n ¼ 4) and 9.8 4.5% (n ¼ 8) for oocysts, respectively. Those by membrane filter for 20 l of treated water were 40.4 9.7% (n ¼ 5) for cysts and 17.5 3.2% (n ¼ 5) for oocysts. For membrane filter, less than 2 l of raw water were concentrated, and their mean recovery efficiencies are 38.3 18.5% (n ¼ 4) for cysts and 16.0 1.7% (n¼ 3) for oocysts, respectively. LeChevallier et al. (1995) and Clancy et al. (1994) have concluded the
presence of unidentified variation in evaluating recovery efficiencies, and the values from one lab to another can vary by logarithmic factors of 0.31. LeChevallier et al. (1995) also suggested that careful handling in concentration and clarification may improve the recovery efficiency.
Our analysis showed higher protozoa recovery for treated water samples than for raw water samples, although not statistically different, which was similar to that reported in the literature (LeChevallier and Norton, 1995). Our results also showed that mem-brane filtration method had higher recovery efficiency than cartridge filtration method regardless of the types of water samples. Similar results were observed by Nieminski et al. (1995), and Shepherd and Wyn-Jones (1995). Cryptosporidium oocysts showed lower recovery than Giardia cysts from both concentration methods. The results are in agreement with the findings from related studies (Falk et al., 1998; Nieminski et al., 1995; LeChevallier et al., 1995) that oocysts lodge in the deeper area of the cartridge/ membrane filters due to their small size, which makes them difficult to remove during the elution procedure:
Concentrations of Protozoa and their detection limits in water samples
The five large water treatment plants in this study use the traditional treatment processes, which includ-ing coagulation, flocculation, sedimentation, filtra-tion and disinfecfiltra-tion. The average residual chlorine concentration in treated water determined as 0.743 0.169 mg/l (n ¼ 10). The water from six simple water facilities is treated by simple filtration procedure, because these systems located in the secluded area generally have fewer technical and financial resources to cope with new requirements.
Water samples from water treatment plants and simple water facilities in this study were filtrated
through both types of filters. The mean detection limits of parasites for these two concentration methods are shown in Fig. 2. Among these three types of water samples, treated water displayed the lowest mean detection limit (5.2 2.6 organisms/100 l for cartridge filtration method and 20.5 16.0 organisms/100 l for membrane filtration method), while raw water showed the highest mean detection limit (14.4 4.7 organisms/100 l for cartridge filtra-tion method and 259.3 289.6 organisms/100 l for membrane filtration method). The detection limits of cartridge filtration method for all samples were quite consistent while membrane filtration method exhibit much greater variation in detection limits.
Due to high turbidity of raw waters, only a small volume of sample could be processed through the membrane filter. This resulted in a significantly high mean detection limit, 259.3 289.6/100 l, which was 16 times higher than that of treated water and 10.4 times higher than the simple facility water. Such a small sampling volume also caused the error in calculating cysts or oocysts in water samples when converting to the number of parasites per 100 l due to the fact that parasites are not uniformly distributed in water samples (Atherholt and Korn, 1999). The ICR requires that large water utilities monitor the treated water whenever the Cryptosporidium concen-trations in raw water exceed 100 oocysts/100 l (LeChevallier and Norton, 1995). The membrane filtration cannot replace the cartridge filtration, because any number of cyst or oocyst detected on the slide would have generated a concentration that exceeds the ICR regulation.
The concentrations of Giardia and Cryptospori-dium in 52 water samples were listed in Table 1. Although the water samples used in two concentra-tion methods were collected and concentrated simul-taneously, the occurrence of parasites detected by two methods varied significantly. Due to the differ-ences in detection limits, errors in the measurement procedure, and variation in sampling, the percentage samples of conflicting result were 23 and 27% for
Fig. 1. The mean recovery efficiencies of Giardia and Cryptosporidium detected by two concentration methods
from seeded water samples.
Fig. 2. The mean parasite detection limits of two concen-tration methods from different types of water samples.
cysts and oocysts, respectively. The mean parasite concentrations of raw water, treated water and simple facility water samples are shown in Fig. 3. For all three types of water samples, membrane filtration method displayed higher mean concentra-tion of Giardia than cartridge filtraconcentra-tion method. As for Cryptosporidium, the higher mean concentration from the membrane filtration method only occurred in water facility samples. By comparing the parasite concentrations between raw and treated water from five water treatment plants, the mean removal efficiencies of parasites detected by cartridge filtration method were 90.3% for Giardia and 76.8% for Cryptosporidium, and those by membrane filtration method was 96.5% for Giardia and 96.4% for Cryptosporidium. The parasite concentration from two filtration methods were correlated as follows: Spearman R¼ 0:683 and p50:001 for raw water samples; Spearman R¼ 0:882 and p50:001 for treated water samples; Spearman R¼ 0:334; p¼ 0:150 for simple facilities water samples. In conclusion, considerable variation was found be-tween the occurrence of parasites detected by the two concentration methods, while significant correlation of two methods was discovered between the raw and treated water samples.
Table 2 listed the correlation coefficients between protozoan parasites and indicator microorganisms in the raw water samples. While parasites were con-centrated with cartridge filters, the significant corre-lation was observed between Giardia concentrations and the levels of heterotrophic bacteria, total coli-forms, fecal colicoli-forms, and Enterococcus. No sig-nificant relationship was found between Cryptosporidium oocysts and indicator microorgan-isms besides fecal coliforms. When parasites were concentrated with membrane filters, a significant correlation was observed between Giardia concentra-tions and the levels of heterotrophic bacteria and
Table 1 . The concentration o f Giardia and Cryptosporidium in 52 water samples in Taiwan water supplies Raw water/cartridge filter Raw w ater/membrane filter Treated w ater/cartridge filter Treated water/membrane filter Sampling site Giardia (cysts/100 l) Cryptosporidium (oocysts/100 l) Giardia (cysts/100 l) Cryptosporidium (oocysts/100 l) Giardia (cysts/100 l) Cryptosporidium (oocysts/100 l) Giardia (cysts/100 l) Cryptosporidium (oocysts/100 l) B.D. 540 117 4929 5 5887 123 1 89 5 130 5 300 B.D. 3175 9 21 3190 1524 92 157 6 87 5 113 P.H. 119 5 180 3 50 712 63 292 5 108 5 250 P.H. 554 613 11,027 2441 103 5 38 49 5 29 H.C. 5 29 2097 425 2033 5 17 5 52 41 5 95 H.C. 583 888 3268 7813 143 1 83 41 143 F.Y. 28 5 84 5 54 130 5 22 5 67 5 31 5 71 F.Y. 286 871 3 49 3333 31 32 12 86 C.C. 41 5 125 5 622 5 2437 5 9 5 27 5 61 5 141 C.C. 260 1319 545 1302 16 728 1 4 4 75 P.L. a 32 98 448 357 S.S. a 988 322 1046 179 C.X. a 151 5 92 149 119 B.L. a 5 5 5 15 5 22 5 52 S.K. a 5 7 5 21 5 38 5 90 L.F. a 5 24 5 72 5 131 5 313 aSimple w ater facilities.
Fig. 3. The mean parasite concentrations of two concentra-tion methods from different types of water samples.
fecal coliforms. No significant relationship was found between Cryptosporidium and any indicator micro-organism.
The drinking water standard of the USEPA Federal Register (USEPA, 1989) allows one case of microbiologically caused illness per year per 10,000 individuals. To achieve this requirement, Rose et al. (1991) suggested that tap water should not contain more than 7 10ÿ 4 Giardia cysts/100 l on the basis of a geometric mean for one year. Haas and Rose (1995) stated that an outbreak would probably occur if tap water contains more than 10–30 Cryptosporidium oocysts/100 l. According to the calculation, the Taiwan water supplies run a high risk for waterborne diseases. However, few outbreaks of giardiasis and cryptos-poridiosis were reported. The main reason is that people in Taiwan when choosing tap water as their drinking water, they either boil the water before drinking or treat it with a household purification device. Another reason may be these two parasites are not included in the routine analysis of feces for the diarrhea patients. Therefore, it is hard to tell if any outbreak occurred.
The influence of turbidity on two concentration methods
The mean level of water turbidity was 32 NTU for seeded raw water samples and was 1.2 NTU for seeded treated water samples. Detection of parasite recoveries from seeded water samples (Fig. 1) indicated that any increase in water turbidity resulted in a significant decrease in parasite recovery.
Figure 4 presents the distribution of water turbidity, detection limits of both concentration methods and sampling volume of membrane filtra-tion in all water samples. All of the data in Fig. 4 were derived from log (Xþ 1) transformation. Results showed similar trends and significant correla-tion between turbidity and deteccorrela-tion limit for membrane filters (Spearman R¼ 0:776; p50:001) and cartridge filters (Spearman R¼ 0:726; p50:001). Although membrane filters displayed higher recovery for both parasites in raw and treated water samples than cartridge filters, the sampling volume capacity of membrane filters was significantly reduced by the fact that the membrane pores were easily blocked by the particles of raw water samples. Table 2. Nonparametric correlation coefficient between the densities of protozoa to indicator microorganisms (R: Spearman R; P: p-levels;
Sample number: 16)
Concentration method Parasites Heterotrophic bacteria Total coliforms E. coli Fecal coliform Enterococcus
Cartridge filtration Giardia R¼ 0:794 R¼ 0:657 R¼ 0:229 R¼ 0:700 R¼ 0:505 P50:001 P¼ 0:006 P¼ 0:391 P¼ 0:002 P¼ 0:045 Cryptosporidium R¼ 0:418 R¼ 0:454 R¼ 0:271 R¼ 0:555 R¼ 0:354 P¼ 0:107 P¼ 0:077 P¼ 0:309 P¼ 0:025 P¼ 0:178 Membrane filtration Giardia R¼ 0:715 R¼ 0:394 R¼ 0:154 R¼ 0:642 R¼ 0:456 P¼ 0:002 P¼ 0:131 P¼ 0:576 P¼ 0:007 P¼ 0:076 Cryptosporidium R¼ 0:360 R¼ 0:271 R¼ 0:049 R¼ 0:454 R¼ 0:377 P¼ 0:170 P¼ 0:309 P¼ 0:856 P¼ 0:077 P¼ 0:149
Fig. 4. The distribution of water turbidity, detection limits of two concentration methods and sampling volume of membrane filtration in water samples.
Therefore, membrane filters were only suitable for water samples under low water turbidity. The significant opposite trend and negative correlation were found between water turbidity level and sampling volume of membrane filters (Spearman R¼ ÿ0:706; p50:001).
In raw water samples, a significant correlation was discovered between water turbidity and parasites. The correlation coefficients between raw water turbidities and parasite concentrations were 0.435 (p¼ 0:092) for cysts and 0.533 (p ¼ 0:033) for oocysts from cartridge filter method, and were 0.589 (p¼ 0:016) for cysts and 0.421 (p ¼ 0:104) for oocysts form membrane filters.
CONCLUSIONS
Seeded water samples concentrated by membrane filters achieved higher recovery efficiency than cartridge filters. The detection limits of water samples concentrated by membrane filters were higher than cartridge filters. The cartridge filter can be used for concentrating large volume samples but the mem-brane filter is for samples of a few liters, especially when the water contains high turbidity. Results of recoveries and detection limit suggested that car-tridge filter is suitable for high-turbid water. The membrane filtration method, although safer in preserving parasites during the detection procedure, allows less volume of turbid water to be filtered through the membrane filters and is only suitable for waters of low turbidity and high quality (i.e. filtered water). The occurrences of both parasites and the relationships between parasite concentrations with indicator microorganisms suggest that the detection of parasites by cartridge filtration method is quite different from that of the membrane filtration method.
Though Method 1622 and 1623 have been validated for its high recoveries and low detection limit, it is too expensive to adopt in most countries. Therefore, optimizing the ICR method is very important for developing and under-developing countries. Areas of modification include improving the efficiency in the sample concentration and the accuracy in the detection for parasites in water samples.
Acknowledgements}This work was funded by the Environ-mental Protection Administration (EPA-88-J1-02-03-008). We are grateful to Mr. Peang-Rong Wang for his experi-mental work, and to Dr. Jill Ruhsing Pan for her editing.
REFERENCES
APHA (1995) Standard Method for the Examination of Water and Wastewater, 19th ed. Am. Publ. Hlth. Assoc., Washington, DC.
Atherholt T. B. and Korn L. R. (1999) ICR protocol: alternative treatment of parasite sample data. J. Am. Water Works Assoc. 91(3), 95–102.
Clancy J. L., Gollnitz W. D. and Tabib Z. (1994) Commercial labs: How accurate are they?. J.Am. Water Works Assoc. 86(5), 89–97.
Clancy J. L. and Hansen J. (1999) Uses of protozoan monitoring data. J. Am. Water Works Assoc. 91(5), 51–65. Cook G. C. (1995) Entamoeba histolytica and Giardia lambliainfection: current diagnostic strategies. Parasite 2(2), 107–112.
Eisenberg J. N. S., Seto E. Y. W., Colford J. M., Olivieri A. and Spear R. C. (1998) An analysis of the Milwaukee cryptosporidiosis outbreak based on a dynamic model of the infection process. Epidemiology 9(3), 255–263. Falk C. C., Karanis P., Schoenen D. and Seitz H. M. (1998)
Bench scale experiments for the evaluation of a mem-brane filtration method for the recovery efficiency of Giardia and Cryptosporidium from water. Water. Res. 32(3), 565–568.
Frost G. S. Schaefer III F. W., Messer J. W., Dahling D. R. and Stetler, R. E. (1996) ICR Microbial Laboratory Manual. EPA 600/R-95/178. USEPA, Ofce. of Research and Development, Washington, DC.
Haas C. N. and Rose J. B. (1995) Developing an action level for Cryptosporidium. J. Am. Water Works Assoc. 87(9), 81–84.
Kramer M. H., Herwaldt B. L., Craun G. F., Calderon R. L. and Juranek D. D. (1996) Waterborne disease: 1993 and 1994. J. Am. Water Works Assoc. 88(3), 66–80. LeChevallier M. W. and Norton W. D. (1995) Giardia and
Cryptosporidium in raw and finished water. J. Am. Water Works Assoc. 87(9), 54–68.
LeChevallier M. W., Norton W. D., Siegel J. E. and Abbaszadegan M. (1995) Evaluation of the immuno-fluorescence procedure for detection of Giardia cysts and Cryptosporidiumoocysts in water. Appl. Environ. Micro-biol. 61(2), 690–697.
Nieminski E. C., Schaefer F. W. and Ongerth J. E. (1995) Comparison of two methods for detection of Giardia cysts and Cryptosporidium oocysts in water. Appl. Environ. Microbiol. 61(5), 1714–1719.
Perz J. F., Ennever F. K. and Le Blancq S. M. (1998) Cryptosporidiumin tap water}comparison of predicted risks with observed levels of disease. Am. J. Epidemiol. 147(3), 289–301.
Rose J. B., Haas C. N. and Regli S. (1991) Risk assessment and control of waterborne giardiasis. Am. J. Public Health 81, 709–713.
Shepherd K. M. and Wyn-Jones A. P. (1995) Evaluation of different filtration techniques for the concentration of Cryptosporidiumoocysts from water. Water Sci. Technol. 31(5-6), 425–429.
SoloGabriele H. and Neumeister S. (1996) US outbreaks of cryptosporidiosis. J. Am. Water Works Assoc. 88(9), 76–86.
Teunis P. F. M., Medema G. J., Kruidenier L. and Havelaar A. H. (1997) Assessment of the risk of infection by Cryptosporidium or Giardia in drinking water from a surface water source. Water. Res. 31(6), 1333–1346. USEPA (1989) National primary drinking water regulation;
filtration and disinfection; turbidity; Giardia lamblia, viruses, Legionella, heterotrophic bacteria. Federal Reg-ister54, 27486-27541.
USEPA (1995) ICR Protozoan Method for Detecting Giardia cysts and Cryptosporidium oocysts in Water by a Fluorescent Antibody Procedure. EPA/814-B-95/003. USEPA, Ofce. of Ground Water and Drinking Water, Washington, DC.
USEPA (1997) Method 1622: Cryptosporidium in Water by Filtration/ IMS/ FA and Viability by DAPI/PI, May 1997 draft. EPA/821-D-97/001. USEPA, Ofce. of Water, Washington, DC.
USEPA (1999a) Method 1622: Cryptosporidium in water by Filtration/ IMS/ FA. EPA/821-R-99/001. USEPA, Ofce. of Water, Washington, DC.
USEPA (1999b) Method 1623: Cryptosporidium and Giardia in Water by Filtration/IMS/FA. EPA/821-R-99/006. USEPA, Ofce. of Water, Washington, DC.
