程序代写代做代考 Bayesian graph C html case study DNA Water Research 125 (2017) 438e448

Water Research 125 (2017) 438e448
Contents lists available at ScienceDirect
Water Research
journal homepage: www.elsevier.com/locate/watres
Target virus log10 reduction values determined for two reclaimed wastewater irrigation scenarios in Japan based on tolerable annual disease burden
Toshihiro Ito a, Masaaki Kitajima a, Tsuyoshi Kato b, Satoshi Ishii c, Takahiro Segawa d, Satoshi Okabe a, Daisuke Sano e, *
a Division of Environmental Engineering, Faculty of Engineering, Hokkaido University, North 13, West 8, Kita-ku, Sapporo, Hokkaido 060-8628, Japan
b Department of Computer Science, Graduate School of Engineering, Gunma University, Tenjinmachi 1-5-1, Kiryu, Gunma 376-8515, Japan
c Department of Soil, Water, and Climate; BioTechnology Institute, University of Minnesota, 140 Gortner Laboratory of BioChemistry, 14749 Gortner Avenue, St. Paul, MN 55108-1095, USA
d The Center for Life Science Research, Yamanashi University, 1110, Shimogato, Chuo, Yamanashi, 409e3898, Japan
e Department of Civil and Environmental Engineering, Graduate School of Engineering, Tohoku University, Aoba 6-6-06, Aramaki, Aoba-ku, Sendai, Miyagi 980-8579, Japan
articleinfo abstract
Article history:
Received 9 March 2017 Received in revised form
23 August 2017
Accepted 24 August 2017 Available online 26 August 2017
Keywords:
Norovirus
Performance target
Tolerable disease burden
Virus log10 reduction
Wastewater reclamation and reuse
Multiple-barriers are widely employed for managing microbial risks in water reuse, in which different types of wastewater treatment units (biological treatment, disinfection, etc.) and health protection measures (use of personal protective gear, vegetable washing, etc.) are combined to achieve a perfor- mance target value of log10 reduction (LR) of viruses. The LR virus target value needs to be calculated based on the data obtained from monitoring the viruses of concern and the water reuse scheme in the context of the countries/regions where water reuse is implemented. In this study, we calculated the virus LR target values under two exposure scenarios for reclaimed wastewater irrigation in Japan, using the concentrations of indigenous viruses in untreated wastewater and a defined tolerable annual disease burden (104 or 106 disability-adjusted life years per person per year (DALYpppy)). Three genogroups of norovirus (norovirus genogroup I (NoV GI), geogroup II (NoV GII), and genogroup IV (NoV GIV)) in un- treated wastewater were quantified as model viruses using reverse transcription-microfluidic quanti- tative PCR, and only NoV GII was present in quantifiable concentration. The probabilistic distribution of NoV GII concentration in untreated wastewater was then estimated from its concentration dataset, and used to calculate the LR target values of NoV GII for wastewater treatment. When an accidental ingestion of reclaimed wastewater by Japanese farmers was assumed, the NoV GII LR target values corresponding to the tolerable annual disease burden of 106 DALYpppy were 3.2, 4.4, and 5.7 at 95, 99, and 99.9%tile, respectively. These percentile values, defined as “reliability,” represent the cumulative probability of NoV GII concentration distribution in untreated wastewater below the corresponding tolerable annual disease burden after wastewater reclamation. An approximate 1-log10 difference of LR target values was observed between 104 and 106 DALYpppy. The LR target values were influenced mostly by the change in the logarithmic standard deviation (SD) values of NoV GII concentration in untreated wastewater and the reliability values, which highlights the importance of accurately determining the probabilistic distribu- tion of reference virus concentrations in source water for water reuse.
© 2017 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
1. Introduction
Water reuse has been widely employed to compensate for increasing water demand in arid and semi-arid regions (Rijsberman, 2006), as well as to enhance nutrient recycling by using reclaimed wastewater for agricultural irrigation (Zurita and
* Corresponding author.
E-mail address: daisuke.sano.e1@tohoku.ac.jp (D. Sano).
http://dx.doi.org/10.1016/j.watres.2017.08.057
0043-1354/© 2017 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

T. Ito et al. / Water Research 125 (2017) 438e448 439
Abbreviations
BIC Bayesian information criterion
cDNA complementary DNA
DALYpppy disability-adjusted life year per person per year DB disease burden
IAC internal amplification control
LR log10 reduction
MBR membrane bioreactor
MF-qPCR microfluidic quantitative PCR
MgV mengovirus
MNV murine norovirus
NA nucleic acid
NoV GI norovirus genogroup I
NoV GI.1 norovirus genogroup I genotype 1 NoV GII norovirus genogroup II
NoV GII.4 norovirus genogroup II genotype 4
NoV GIV NoV NT PBS
PEG QMRA RT
SD TLN WHO
norovirus genogroup IV
norovirus not typed phosphate-buffered saline polyethylene glycol
quantitative microbial risk assessment reverse transcription
standard deviation truncated log-normal World Health Organization
White, 2014). However, reclaimed wastewater frequently contains waterborne pathogens, most notably enteric viruses (Gibson, 2014). To manage the health risks due to exposure to viruses in reclaimed wastewater, international and domestic guidelines have recom- mended the multiple-barrier system (Sano et al., 2016). According to the World Health Organization (WHO) guidelines (2006), a wastewater reclamation process is designed to achieve a target value of log10 reduction (LR) by combining different types of wastewater treatment units and health protection measures. The WHO guidelines (2006) require a 6- to 7-log10 virus reduction by a wastewater reclamation process if reclaimed wastewater is used for unrestricted irrigation (i.e., irrigation of all crops including vege- tables eaten raw. In Australia, the states of Queensland and Victoria have also proposed that water reclamation plants achieve virus reductions of 6.5- and 7-log10, respectively, for agricultural irriga- tion (Environment Protection Agency Victoria, 2005, 2003; The state of Queensland, 2008).
The LR target values recommended by the established guide- lines can be employed in other countries and regions where wastewater reclamation guidelines have not been prepared. The government of Japan has issued a wastewater reclamation manual, but the multiple-barrier system is not employed as a measure to control microbial safety. It is possible to employ the international guidelines for the LR target values in Japan, but several reports have argued that these values should be determined based on the context of the countries/regions where water reuse is imple- mented. For example, the Victoria state guidelines recommend that proponents of wastewater reclamation or manufacturers of wastewater treatment units must conduct research to identify target viruses in situ if they are unknown (The Victoria Department of Health, 2013). It is of advantage for wastewater engineers to prepare guidelines that instruct how to calculate LR values, based on the virus types of concern/prevalence, virus monitoring data in untreated wastewater, the water reuse scheme, and hygiene prac- tices implemented in specific countries/regions (Environment Protection Agency Victoria, 2005, 2003; The state of Queensland, 2008).
In this study, we constructed a framework to determine LR target values for viruses in wastewater reclamation systems based on the tolerable annual disease burden. We calculated the virus LR target values using virus concentration data in untreated waste- water from routine monitoring under two exposure scenarios of water reuse for agricultural irrigation. We selected norovirus (NoV) as a model virus in this study, because it has been the most sig- nificant cause of gastroenteritis outbreaks among enteric viruses in Japan (Japan’s National Institute of Infectious Disease, 2017). Approximately 3000 cases of NoV-associated gastroenteritis were
reported annually from 2011 to 2015, with breakdowns of the annual incidence of NoV-associated gastroenteritis of 9.2% for NoV genogroup I (NoV GI), 90% for NoV genogroup II (NoV GII), and 0.8% for NoV not typed (NT) (Japan’s National Institute of Infectious Disease, 2017). NoV GI, NoV GII, and NoV genogroup IV (NoV GIV) in untreated wastewater samples collected from a pilot-scale wastewater treatment plant were quantified using reverse transcription-microfluidic quantitative PCR (RT-MF-qPCR) in this study. Since only NoV GII was present in a quantifiable concentra- tion in the wastewater samples, the quantitative data of NoV GII were used to estimate the probabilistic distribution of NoV GII concentration. The NoV GII tolerable concentration in reclaimed wastewater was then calculated so as not to exceed the tolerable annual disease burden values of 104 or 106 disability-adjusted life year per person per year (DALYpppy), which have commonly been used as health-based targets in countries/regions with a low or high disease burden (Mara, 2011; Mara et al., 2010; World Health Organization, 2006). The required LR target value of NoV GII was calculated by determining the difference between the 95, 99, and 99.9%tiles of NoV GII concentration in the influent and the tolerable concentration of NoV GII in reclaimed wastewater. Finally, sensi- tivity analysis was performed to identify the most influential pa- rameters for the LR target values of NoV GII.
2. Materials and methods
2.1. Collection, processing, and analysis of wastewater samples
Wastewater samples were taken in a pilot-scale anoxic/oxic submerged membrane bioreactor (MBR) in Japan. The MBR pilot plant received pre-screened domestic wastewater from a company dormitory (approximately 30 households). At the MBR pilot plant, the influent was collected twice a month in November and December 2013 and from March to October in 2014, and three times a month in the following winter season (January and February of 2014). Wastewater samples were more frequently collected in January and February because a higher concentration of NoV in wastewater samples was expected. All the collected samples (n 1⁄4 26) were immediately frozen and delivered to the laboratory.
NoV GI, NoV GII, and NoV GIV in each wastewater sample were concentrated by the polyethylene glycol (PEG) precipitation method (Lewis and Metcalf, 1988). Briefly, PEG 6000 (Wako Pure Chemical Industries, Osaka, Japan) and NaCl were added to 50 mL of each wastewater sample to obtain a final concentration of 8.0% (wt/ vol) and 0.4 M, respectively. Murine norovirus (MNV, strain S7-PP3) propagated in RAW 264.7 cells (ATCC TIB-71) was used as a process control to evaluate the recovery efficiency of indigenous viruses in

440 T. Ito et al. / Water Research 125 (2017) 438e448
the wastewater samples (Ishii et al., 2014). The wastewater samples were seeded with 10 mL of MNV suspension (approximately 108 copies) and used for virus concentration. The mixtures were slowly stirred overnight at 4 C, followed by centrifugation at 10,000  g for 30 min at 4 C. The pellet was suspended in 5 mL of 1 M phosphate-buffered saline (PBS) to obtain the virus concentrate.
To check the inhibition of viral nucleic acid (NA) extraction, mengovirus (MgV, vMC0 strain) propagated in Hela cells (ATCC CCL2) was added to each wastewater concentrate (Costafreda et al., 2006). One milliliter of each wastewater concentrate was seeded with 10 mL of MgV suspension (approximately 106 copies). Viral NA was extracted using a NucliSENS® miniMAG® (bioMerieux, Tokyo, Japan) according to the manufacturer’s instructions, to obtain a final NA extract volume of 100 mL. The complementary DNA (cDNA) was synthesized using the PrimeScript RT reagent Kit (Takara Bio, Japan). Forty microliters of the RT reaction mixture containing 8 mL of the extracted NA, 8 mL of Primer Buffer, 2 mL of PrimeScript RT Enzyme Mix I, 2 mM oligo dT primer, and 20 mM random hexamers were incubated at 37 C for 15 min, 42 C for 15 min, and 85 C for 5 s.
Prior to the qPCR, an internal amplification control (IAC) plasmid DNA, including a gene sequence of Pseudogulbenkiania sp. NH8B- 1D2 (Ishii et al., 2014), was added to each aliquot of cDNA prod- ucts to check the inhibition in gene amplification. Previously vali- dated TaqMan qPCR assays were performed in MF-qPCR format for simultaneous quantification of NoV GI, NoV GII, NoV GIV, MgV, MNV, and IAC plasmid DNA (Ishii et al., 2014). Ten-fold serial di- lutions ranging from 2  100 to 2  106 copies of a mixture of the 6 plasmid DNA were used to generate standard curves for MF-qPCR. The standard curve was composed of seven-point 10-fold serial dilutions of the plasmid DNA as positive controls, and a one-point nuclease-free water as a negative control. To increase the target genes, a specific target amplification (STA) reaction was performed prior to MF-qPCR (Ishii et al., 2014). The STA reaction mixture (10 mL) containing 5 mL of 2  TaqMan PreAmp master mix (Applied Bio-systems, Foster City, CA), 0.2 mM each primer, 2.25 mL of DNA/ cDNA template, and 0.25 mL of the IAC plasmid (1⁄4 approximately 105 copies) was incubated at 95  C for 10 min, 14 cycles of 95  C for 10 s, and 60 for 4 min. The STA products were diluted six-fold with Tris-EDTA (TE) buffer and used as a template for MF-qPCR. The MF- qPCR reaction mixture contained 1  TaqMan Universal PCR master mix (Applied Biosystems), 400 nM each of forward and reverse primers, and 200 nM probe. MF-qPCR was performed in triplicate using a BioMark HD reader with a Dynamic Array 48.48 chip or 96.96 chip (Fluidigm, South San Francisco, CA) under the following conditions: 50 C for 2 min, 95 C for 10 min, and 40 cycles of 95 C for 15 s and 60 C for 1 min. The results were analyzed using Real- Time PCR Analysis software version 3.0.2 (Fluidigm).
MF-qPCR assays were carried out twice with 20 samples collected from November 2013 to July 2014, and 6 samples collected from August 2014 to October 2014. Standard curves were generated for the quantification cycle (Cq) values versus the amounts of plasmid DNA (log10 copies/L) in each MF-qPCR assay (Table S2) (Bustin et al., 2010, 2009). The quantification limits of the target genes in the untreated wastewater samples were calculated by multiplying the amplification of the lowest concentration of the plasmid DNA by the concentration factors, given by sample pro- cessing, NA extraction, and cDNA synthesis. The quantitative data of the indigenous viral concentrations below the quantification limits were regarded as negative in the present study. Virus recovery ef- ficiency, viral NA extraction efficiency, and PCR efficiency were determined based on the observed gene copy numbers of seeded and recovered MNV, MgV, and bacterial plasmid DNA, respectively. When these efficiencies were lower than 1%, the quantitative
analysis was redone (Da Silva et al., 2007). The virus recovery ef- ficiency, viral NA extraction efficiency, and PCR efficiency were not used to adjust the observed copy numbers of NoV GI, NoV GII, and NoV GIV for each wastewater sample.
2.2. Bayesian estimation of the virus concentration distributions
The previously published Bayesian model (Ito et al., 2015; Kato et al., 2013) was employed to estimate the virus concentration distributions for datasets containing non-detects. In the Bayesian model, a truncated log-normal (TLN) distribution is adopted to divide the data into two groups, detects and non-detects, by setting values on the limit of quantification. The fitness of virus concen- tration datasets to the probabilistic distributions (normal, log- normal, and gamma distributions) was tested based on the Bayesian Information Criterion (BIC) prior to use of the Bayesian model (Vrieze, 2012). The BIC statistics are defined as Eq. (1):
BIC1⁄4 2lnðLÞþklnðNÞ (1)
where lnðLÞ is the logarithmic maximum likelihood value, k is the number of parameters, and N is the total number of virus concen- tration data, including detects and non-detects.
The previously published Bayesian estimation model was used to estimate the NoV GII concentration distribution in untreated wastewater, because only NoV GII among the three genotypes was present in a quantifiable concentration in the wastewater samples. The likelihood function is written as Eq. (2):
 N 1yi 2 Y ðqi  mÞ
p Xjm;s 1⁄4 4 pffiffiffiffiffi
i1⁄41 
s2
ðqi  mÞ pffiffiffiffiffi
s2

2 TLN xi;m;s ;qi
yi
(2)
14
The virus concentration dataset X is expressed by a tuple X 1⁄4 fðxi ; yi ÞgNi1⁄41 , where xi is the i-th quantitative result of viruses in the total virus concentration data N, and yi is a Bernoulli variable based on the quantification limit 10qi ; yi 1⁄4 1 if xi ! 10qi , and
yi 1⁄4 0, otherwise. The two model parameters of mean m and variance s2 are obtained from posterior distributions, respectively,
~1g
given m 1⁄4 Nð0; 100Þ and s2 1⁄4 Gam ð0:01; 0:01Þ as prior distribu-
tions (Paulo et al., 2005). The posterior predictive distribution of the virus concentration is then obtained by Eq. (3):
x
2.3. Exposure scenarios
; m; s2pm; s2jXdmds (3)
P
pred
log10
jX 1⁄4
Nx
log10
Z
Two exposure scenarios, based on the Japanese context, were assumed (Table 1): Scenario I assumes accidental ingestions of reclaimed wastewater by farmers, whereas Scenario II assumes daily consumption of salad vegetables grown with reclaimed wastewater irrigation.
In Scenario I, accidental ingestion of 1 mL of reclaimed waste- water per person per day was assumed for farmers (Ottoson and Stenstro€m, 2003). In this scenario, reclaimed wastewater was assumed to be sprayed or sprinkled on agricultural land. For intensive farming in Japan, the typical work week was assumed to be at least 5e6 days per person, with a total of 300 days of irrigation as used in other QMRA studies (Mara et al., 2007).
In Scenario II, we assumed that reclaimed wastewater was

Table 1
Quantitative microbial risk assessment model input parameters.
Parameter Parameter name
c Concentration of norovirus genogroup II (NoV GII) in untreated wastewater
Units Categorya
log10-copies/ D mL
Value or distribution
pffiffiffiffi Lognorm(5.0ð1⁄4 10mÞ , 51ð1⁄4 10 s2 Þ)
1 300
Reference
This study
(Ottoson and Stenstro€m, 2003) (Mara et al., 2007)
(Japan’s Ministry of Agriculture Forestry and Fisheries, 2013) (Japan’s Ministry of Agriculture Forestry and Fisheries, 2013) (Japan’s Ministry of Agriculture Forestry and Fisheries, 2013) (Japan’s Ministry of Agriculture Forestry and Fisheries, 2013) (Japan’s Ministry of Agriculture Forestry and Fisheries, 2013)
(Asano et al., 1992; Shuval et al., 1997)
(Shuval et al., 1997)
(Japan’s Ministry of Agriculture Forestry and Fisheries, 2013)
(World Health Organization, 2006)
(Dawson et al., 2005) (Dawson et al., 2005)
(Barker et al., 2013; Messner
et al., 2014; Teunis et al., 2008)
(Teunis et al., 2008)
(Mok et al., 2014)
l1⁄410cV Exposure parameter (Scenario II: salad vegetables consumption by Japanese customers)
Daily consumption of salad vegetables WL Lettuce (leafy vegetable)
WCa Cabbage (leafy vegetable)
WT Tomato (non-leafy vegetable) WCu Cucumber (non-leafy vegetable)
Volume of irrigation water remaining on
salad vegetables VL Leafy vegetable
VNL Non-leafy vegetable
n Days of consumption per year
NoV GII reduction by various health protection measures R NoV GII reduction from last irrigation to
consumption
Die-off after last irrigation
Reduction with washing using water Reduction with washing using chlorinated water
l Daily dose of NoV GII in salad vegetables consumption
Dose-response parameter
g/day g/day g/day g/day g/day
mL/100 g
mL/100 g days
log10-copies log10-copies
log10-copies log10-copies
copies
C
F
F
F
F
F
F
F
F F
F F
F F
C
C
F F F
pfp(inf j exp)1⁄4P (1-exp(-(l/ m)))
F F C F F C
C
C D
28.7 4.6 7.8 11.0 5.3
10
0.36 365
1.69 0.5
0.3 0.89
l1⁄410c((WLþWCa)VLþ(WTþWCu)VNL)/ 10R
phg(inf j exp) 1⁄4 1-(2F1(b, l  (1-a)/a, a-b; a)  (1/(1-a))^-(l(1-a)/a))
a 1⁄4 0.0044
b 1⁄4 0.0020
a 1⁄4 0.9999
(Messner et al., 2014)
P 1⁄4 0.722
m1⁄41106
p(ill j inf, exp) 1⁄4 1 – (1þh l)-r
h 1⁄4 0.00255
r 1⁄4 0.086
p(ill j exp) 1⁄4 p(ill j inf, exp)  p(inf j exp)
pill 1⁄4 1-(1-p(ill j exp)) n DALY1⁄4pillDB
unif(3.71  104; 6.23  103)*
phg(inf j exp) a
b
a
pfp(inf j exp)
(fractional Poisson model)
P
m
p(ill j inf, exp) h
r
pðill j expÞ pill
The probability of infection per exposure event (hypergeometric model)
The probability of infection per exposure event
Illness rate among infected subjects
The probability of illness per exposure event
Annual probability of Illness
Disability-adjusted life year (DALY) calculation DALY DALY per person per year
DB Disease burden per case of illness
DALYpppy DALY per cases of illness
T. Ito et al. / Water Research 125 (2017) 438e448
441
Exposure parameter (Scenario I: intensive farming in Japan)
V Volume of irrigation water ingested per mL/day F
day
n Days of exposure per year days F l Daily dose of NoV GII in intensive farming log10-copies C
*The maximum value in the range of uniform distribution for DALY per cases of illness (6.23 x 103) was used in the calculation. a C, calculation; F, fixed; D, distribution.

442 T. Ito et al. / Water Research 125 (2017) 438e448
ingested by consuming salad vegetables (lettuce, green cabbage, tomatoes, and cucumbers) grown with reclaimed wastewater irri- gation, because there has been an increase in the consumption of salad vegetables in Japan (Japan’s Ministry of Agriculture Forestry and Fisheries, 2013). To assess microbial risks by salad vegetable consumption, 0.36 or 10 mL of irrigation water was assumed to remain in 100 g of non-leafy (tomatoes and cucumbers) or leafy (lettuce and cabbage) vegetables, respectively (Asano et al., 1992; Shuval et al., 1997). We assumed that the salad vegetables were sold at wholesale and retail markets, and that mean daily con- sumption of salad vegetables was 28.7 g per person per day, ob- tained by the sum of the mean daily consumption of lettuce, green cabbage, tomatoes, and cucumbers from 2013 to 2016 in Japan (Japan’s Ministry of Internal Affairs and Communications, 2017). The breakdown of the mean daily consumption of salad vegetables was 4.6 g for lettuce, 7.8 g for cabbage, 11.0 g for tomatoes, and 5.3 g for cucumbers.
The WHO guidelines show 0.5- to 2.0-log10 natural die-off after the last irrigation (World Health Organization, 2006). We also assumed an additional 1.19-log10 reduction of norovirus occurred by washing salad vegetables with tap water and disinfectants (Dawson et al., 2005), which is a widely used hygiene practice for handling foods in Japan (Japan’s Ministry of Education Culture Sports Science and Technology, 2009). In this study, a total of 1.69-log10 norovirus reduction, the sum of the log10 reduction from natural die-off and hygiene practices, was assumed from the last irrigation to consumption.
2.4. The virus tolerable concentration and LR target value calculations
A QMRA model developed in a previous study (Symonds et al., 2014) was used to calculate the NoV GII tolerable concentration in reclaimed wastewater and the LR target value in wastewater reclamation systems, given the tolerable annual disease burden associated with NoV GII infections under each exposure scenario of 104 or 106 DALYpppy as per the Mara et al. and WHO recom- mendations, respectively (Mara, 2011; Mara et al., 2010; World Health Organization, 2006). The tolerable annual disease burden values of 104 or 106 DALYpppy have commonly been suggested as health-based targets in countries/regions with high or low disease burden, respectively (Mara et al., 2010; World Health Organization, 2006). Mara also noted that 104 DALYpppy would provide an adequate margin of public-health safety in relation to waterborne- diarrheal disease in all countries (Mara, 2011). In this study, the two different values of the tolerable annual disease burden were then used in the water reuse irrigation scenarios in Japan.
DALYpppy can be estimated as the product of the annual proba- bility of illness and the estimated burden of disease per case of illness (i.e., DALYs per case), each of which is denoted by pill and DB, respectively, hereinafter. Namely, DALYpppy is expressed as
DALYpppy 1⁄4 pill  DB (4) Although the parameter DB was given by a uniform distribution
in a previous study (Mok et al., 2014), the maximum value in the range of the disease burden per case of illness (6.23  103) was used in the calculation to take the worst case into account.
Let us denote the event of illness and exposure by ill and exp, respectively. For example, the notation p(ill j exp) represents the probability of illness under the condition of an exposure event. Letting n be the number of exposure days per person per year, the annual probability of illness can be written as
pill 1⁄4 1  ð1  pðilljexpÞÞn (5)
Let inf denote the event of infection. The illness probability under exposure is derived as
pðilljexpÞ1⁄4 pðilljinf;expÞpðinfjexpÞ (6)
In this study, the following model for the dose-dependent conditional probability of illness was employed (Teunis et al., 2008):
pðilljinf; expÞ1⁄41ð1þhlÞr (7)
where h and r are the pre-defined parameters of this model (Messner et al., 2014), and the parameter l is a daily dose of NoV (log10-copies). The parameter l is used again in the model of infection pðinfjexpÞ.
Two models, the hypergeometric model and the fractional Poisson model– denoted by phg(inf j exp) and pfp(inf j exp), respectively–were compared as the model of infection p(inf j exp) in this study. The hypergeometric model phg(inf j exp), introduced
by Teunis et al. (2008), has three parameters, a, b, and a (0 and a probability function given by

a 1),
(8)
0
phgðinfjexpÞ1⁄41@2F1 b; a ;aþb;a
 lð1aÞ  1
 1  lð1aÞ a
 1a A
This equation is the Pfaff transformation of the 2F1 hypergeo- metric model (Barker et al., 2013). For the three parameters a, b, a, this study used the values determined by Messner et al. (2014). These dose-response parameters for the 2F1 hypergeometric model were obtained by fitting with multiple dose data of NoV GI geno- type 1 (NoV GI.1) and NoV GII genotype 4 (NoV GII.4) in human challenge studies (Atmar et al., 2014; Frenck et al., 2012; Seitz et al., 2011; Teunis et al., 2008). The Pfaff transformation of the model was used as a close approximation here, assuming all doses
33,323, because the fit value for parameter a provided by Teunis et al. (2008) exceeds a constraint of the Gauss hypergeometric model used in Eq. (8) (Barker et al., 2013).
The fractional Poisson model pfp(inf j exp) was proposed by Messner et al. (2014). The fractional Poisson model assumes the secretion status of the histo-blood group antigen that has been suggested as an infection factor for human NoV (Lindesmith et al., 2003). In the fractional Poisson model, it is assumed that secretor positive individuals (Seþ) are perfectly susceptible and secretor negative individuals (Se) are protected from NoV infection. The probability mass function of this model is given as
   ml 
1  e (9)
pfpðinfjexpÞ 1⁄4 P 
where l is the daily dose of NoV (log10-copies) described above, m is the mean aggregate size, and P is the fraction of secretor positive individuals (Messner et al., 2014). These dose-response parameters for the fractional Poisson model were obtained by fitting with the multiple dose data of NoV GI.1 and NoV GII.4 obtained from human challenge studies (Atmar et al., 2014; Frenck et al., 2012; Seitz et al., 2011; Teunis et al., 2008).
The daily dose (l) for NoV was calculated using Eqs. (10) and (11) for Scenarios I and II, respectively:

l1⁄4 10cV (10)
and reliability were fixed to 106 DALYpppy, 6.23  103 DALY per case of illness, and 95%, respectively (Table 2).
The impact of variations in reliability values on the LR target value was evaluated by graphing a curve for the LR target values versus the reliability values (from 90 to 99.9%) on each logarithmic SD value (from 1 to 4 at a 1 interval) of NoV GII concentration in untreated wastewater. To examine how the variations of reliability values would influence the LR target value calculation, all the assumed values, except for the logarithmic SD, were fixed to the base values in Scenario I and Scenario II (Table 2).
3. Results
3.1. Quantification of NoV GI, NoV GII, and NoV GIV in untreated wastewater samples
To identify the virus type of concern in wastewater reclamation, NoV GI, NoV GII, and NoV GIV genes were quantified in 26 un- treated wastewater samples by RT-MF-qPCR. Since the virus re- covery efficiencies, viral NA extraction efficiency, and PCR efficiency of all samples were higher than 1% (data not shown), all quantita- tive data of NoV GI, NoV GII, and NoV GIV were regarded as acceptable in the present study. The NoV GII gene was detected in 46% (12/26) of the untreated wastewater samples, while the NoV GI and NoV GIV genes were not detected by RT-MF-qPCR in all sam- ples (Table S4). Thus, only the quantification datasets of NoV GII were used in the following calculation.
3.2. Estimation of the NoV GII concentration distribution in untreated wastewater
To determine the most appropriate probabilistic distribution for
the NoV GII concentration data, BIC statistics for the three candi-
date probabilistic distributions (normal, log-normal, and gamma
distribution) were calculated. The log-normal distribution was
identified as the best fit for all datasets among the probabilistic
distributions tested, as it showed the lowest BIC value (Table S5).
The logarithmic mean and the logarithmic SD of the NoV GII con-
.
l1⁄4 10c ððWL WCaÞVL þðWT WCuÞVNLÞ 10R
(11)
where c is logarithmic tolerable concentrations in reclaimed wastewater, V is the volume of irrigation water (mL/day) ingested by intensive farming; , VL and VNL are the volume of irrigation water remaining on leafy (lettuce and cabbage) or non-leafy (tomatoes and cucumbers) vegetables (mL/100 g); WL, WCa, WT , and WCu are the daily consumption of lettuce, cabbage, tomatoes, and cucum- bers, respectively; and R is total NoV log10 reduction from the last irrigation to consumption.
The NoV GII tolerable concentration in reclaimed wastewater was calculated using iterative analysis according to Symonds et al. (2014). Briefly, the value of the tolerable concentration was increased from zero in increments of 0.1 for at most 10,000 itera- tions. Once the value of the tolerable concentration with a maximum DB (6.23  103) produced the DALY value above the tolerable annual disease burden values of 104 or 106 DALYpppy, its operation could break out of the loop. At the end of the loop, the value of the tolerable concentration decreased by 0.1 to meet the tolerable annual disease burden values of 104 or 106 DALYpppy.
The virus LR target values were then calculated to ensure less
than or equal to the annual disease burden values of 104 or 106
DALYpppy at 95, 99, and 99.9% reliability values. Briefly, a defined
percentile (95, 99, and 99%tile) of NoV GII concentration was ob-
tained from the probabilistic concentration distribution of NoV GII
with the logarithmic mean ð1⁄4 mÞ and the logarithmic standard pffiffiffiffiffi
deviation (SD) ð1⁄4 s2Þ. The required LR target value of NoV GII was calculated based upon the 95, 99, and 99.9%tile of the NoV GII concentration in untreated wastewater and the NoV GII tolerable concentration in reclaimed water. All calculations were performed in “R” version 3.1.0. The R codes used in this study are provided in the Supplementary Information.
2.5. Sensitivity analysis
A spider plot of the LR target values was created with variations of assumed values of general parameters, and exposure parameters in Scenario I and Scenario II as inputs (Table 2), in order to observe which of the inputs impacted the LR target values as outputs (Tian, 2013). The base values of each input were alternately swung between 50% and þ50% as possible variations in the inputs, while tolerable annual disease burden, disease burden per case of illness,
Table 2
Input parameters used for sensitivity analysis. Category
General parameter
Exposure parameter in Scenario I Exposure parameter in Scenario II
centration (log10-copies/mL) in untreated wastewater were esti-
T. Ito et al. / Water Research 125 (2017) 438e448 443
pffiffiffiffiffi
mated to be 0.7 ð1⁄4 mÞ and 1.7 ð1⁄4 s2Þ, respectively, based on the
Bayesian model posterior predictive distribution (Fig. 1).
3.3. Calculation of the tolerable NoV GII concentration in reclaimed
wastewater
The NoV GII tolerable concentration in reclaimed wastewater
Base value of input
(range of values between 50% and þ50%)
2 (1e3)
2 (1e3)
2 (1e3)
200 (100e300) 50 (25e75)
50 (25e75)
10 (5e10)
0.36 (0.18e0.54) 200 (100e300) 1 (0.5e1.5)
Parameter
Logarithmic mean of NoV concentration in untreated wastewater (m)
pffiffiffiffiffi Logarithmic SDa of NoV concentration in untreated wastewater ( s2)
Volume of water ingested per day (V)
Days of irriation exposure per year (n)
Daily consumption of leafy vegetables (WL þ WCa)
Daily consumption of non-leafy vegetables (WT þ WCu)
Volume of irrigation water remaining on leafy vegetables (VL) Volume of irrigation water remaining on non-leafy vegetables (VNL) Days of consumption per year (n)
NoV GII reduction from last irrigation to consumption (R)
*To create the spider plot, tolerable annual disease burden, disease burden per case of illness, and reliability were fixed to 106 disability-adjusted life year per person per year (DALYpppy), 6.23  10e3 DALY per case of illness, and 95%, respectively.
**The base values of all the inputs, except for SD.
a SD, standard deviation.

444 T. Ito et al. / Water Research 125 (2017) 438e448
3.4. CalculationoftheLRtargetvaluesofNoVGIIinthewastewater
Fig. 1. Posterior predictive distribution and histogram of norovirus genogroup II (NoVGII) concentration in untreated wastewater. X-axis represents logarithmic virus concentration (log10-copies/mL), and Y-axis represents probabilistic density.
was calculated using two different dose-response models (hyper- geometric and fractional Poisson models) with the tolerable annual disease burden values of 104 or 106 DALYpppy for two water reuse irrigation scenarios in Japan (Scenario I and II) (Table 3). In Scenario I, the tolerable concentrations of NoV GII were 1.3- and 0.28-log10- copies/mL at the annual tolerable disease burden values of 104 and 106 DALYpppy, respectively, when the hypergeometric model was used (Table 3). The fractional Poisson model gave slightly lower values than the hypergeometric model: 1.3- and 0.26-log10-copies/ mL at the annual disease burden values of 104 and 106 DALYpppy, respectively (Table 3).
The tolerable concentrations of NoV GII in reclaimed wastewater in Scenario II were higher than the respective tolerable concen- trations in Scenario I. The tolerable concentrations for the annual tolerable disease burden values of 106 DALYpppy in Scenario II were calculated to be 1.8- and 1.8-log10 copies/mL based on the hyper- geometric model and the fractional Poisson model, respectively (Table 3). The tolerable concentrations for the annual tolerable disease burden value of 104 DALYpppy were 2.8- and 2.8-log10- copies/mL based on the hypergeometric model and the fractional Poisson model, respectively (Table 3). An approximate 1-log10- copies/mL difference of the tolerable concentrations of NoV GII was observed between the annual disease burden values of 104 and 106 DALYpppy in Scenario I and Scenario II.
Table 3
The tolerable concentration of norovirus genogroup II (NoV GII) in reclaimed wastewater (Scenario I: accidental ingestion of reclaimed wastewater by intensive farming in Japan; Scenario II: salad vegetable consumption by Japanese consumers).
reclamation process
The LR target values of NoV GII were calculated based on its concentration distributions in untreated wastewater and the tolerable concentration in reclaimed wastewater. In Scenario I, the LR target values of NoV GII corresponding to the annual disease burden of 106 DALYpppy were 3.2-, 4.4-, and 5.7-log10 at 95, 99, and 99.9% reliability, respectively, for both dose-response models (Table 4). The LR target values of NoV GII were also calculated based on the annual disease burden value of 104 DALYpppy, which were 2.2-, 3.4-, and 4.7-log10 at 95, 99, and 99.9% reliability, respectively, for both dose-response models (Table 4). In Scenario II, the LR target value of NoV GII was likewise calculated based on the annual disease burden value of 106 DALYpppy, which were 1.7-, 2.8-, and 4.1-log10 at 95, 99, and 99.9% reliability, respectively (Table 4). When the annual disease burden value of 104 DALYpppy was employed, 95, 99, and 99.9% reliability corresponds to 0.7-, 1.8-, and 3.1-log10, respectively (Table 4). No substantial difference in these LR values was observed between the dose-response models. Meanwhile, an approximate 2.5-log10 difference of the LR target values of NoV GII was observed between 95 and 99.9% reliability in Scenario I and Scenario II.
3.5. Sensitivity analysis of the LR target values of NoV GII
Spider plots of the LR target values of NoV GII were presented in Figs. 2a and 3a, in order to examine how possible variations of in- puts (NoV GII concentration in untreated wastewater and exposure parameters) influenced outputs (LR target values of NoV GII) in Scenario I and Scenario II. The hypergeometric model and the fractional Poisson model were used to calculate the LR target values of NoV GII for sensitivity analysis; however, no substantial differ- ence in the LR target values of NoV GII was observed between the dose-response models (data not shown). Steeper slopes indicated larger influences on the variations of inputs, while a positive- negative slope indicated positive-negative correlations. In Figs. 2a and 3a, the variations of logarithmic SD values of NoV GII in un- treated wastewater showed steeper slopes than the variations of the other inputs in Scenario I and Scenario II, which indicated that the logarithmic SD values had the greatest impact on the LR target value calculations with all the inputs tested.
To examine the impact of variations of reliability values on the LR target value calculations, the curves for the LR target values of NoV GII versus the reliability values (from 90 to 99.9%) were depicted with various logarithmic SD values (from 1 to 4 at a 1
Table 4
The log10 reduction (LR) target value of norovirus genogroup II (NoV GII) in waste- water reclamation systems (Scenario I: accidental ingestion of reclaimed waste- water by intensive farming in Japan; Scenario II: salad vegetable consumption by Japanese consumers).
Dose-respnse model
Hypergeometric Fractional Poisson Hypergeometric Fractional Poisson
Tolerable disease Tolerable
burden (DALYpppy) concentration(log10-
copies/mL)
burden (DALYpppy)
10e4 10e6 10e4 10e6 10e4 10e6 10e4 10e6
95%a 99%a 99.9%a
2.2 3.4 4.7 3.2 4.4 5.7 2.2 3.4 4.7 3.2 4.4 5.7 0.7 1.8 3.1 1.7 2.8 4.1 0.7 1.8 3.1 1.7 2.8 4.1
Dose-response model
Hypergeometric Fractional Poisson Hypergeometric Fractional Poisson
Tolerable disease Reliability
Scenario I
Scenario II
1.3 0.28 1.3 0.26 2.8
10e4
10e6
10e4
10e6
10e4
10e6 1.8 10e4 2.8 10e6 1.8
Scenario I
Scenario II
a
untreated wastewater below tolerable annual disease burden of 104 or 106 disability-adjusted life year per person per year (DALYpppy).
95, 99 and distribution in
99.9% represents the cumulative probability of NoV GII concentration

T. Ito et al. / Water Research 125 (2017) 438e448 445
Fig. 2. Sensitivity analysis of norovirus genogroup II (NoV GII) log10 reduction (LR) target values in Scenario I (accidental ingestion of reclaimed wastewater by intensive farming in Japan): (a) spider plots of NoV GII LR target values based on percent change in inputs (see Table 2 for the base values of general parameters and exposure parameters in Scenario I); (b) curves for NoV GII LR target values versus reliability values (from 90% to 99.9%) on various logarithmic standard deviation (SD) (from 1 to 4 at a 1 interval) of NoV GII con- centration in untreated wastewater. All the inputs, except for the SD, were fixed to the base values in Scenario I (see Table 2).
Fig. 3. Sensitivity analysis of norovirus genogroup II (NoV GII) log10 reduction (LR) target values in Scenario II (salad vegetable consumption by Japanese consumers): (a) spider plot of NoV GII LR target values based on percent change in inputs (see Table 2 for the base values of general parameters and exposure parameters in Scenario II; (b) curves for NoV GII LR target values versus reliability values (from 90% to 99.9%) on various logarithmic standard deviation (SD) (from 1 to 4 at a 1 interval) of NoV GII concentration in untreated wastewater. All the inputs, except for the SD, were fixed to the base values in Scenario II (see Table 2).
interval) of NoV GII concentration in untreated wastewater (Figs. 2b and 3a). Steeper slopes in Figs. 2b and 3a indicated larger influences as well as the spider plots. The curves in Figs. 2b and 3a were nonlinear, and larger logarithmic SD values of NoV GII concentra- tion in untreated wastewater led to a remarkable increase in the LR target values of NoV GII between 99% and 99.9% reliability.
4. Discussion
In this study, we constructed an entire process for calculating target LR values under two reclaimed wastewater irrigation sce- narios in Japan, which included the estimation of the probabilistic distribution of virus concentration in untreated wastewater and the deviation of tolerable virus concentration in reclaimed water based on the tolerable annual disease burden. Since a waterborne virus of concern and its concentration in untreated wastewater differed among countries/regions, data obtained from monitoring the virus in untreated wastewater should be used to calculate the LR target values in a wastewater reclamation system. We performed RT-MF- qPCR to quantify NoV GI, NoV GII, and NoV GIV as model viruses, which can cause infections in humans (Vinje, 2015), in 26 untreated wastewater samples collected from a pilot scale wastewater treat- ment plant in Japan. Only NoV GII among the three genogroups
could be detected in 12 of the 26 untreated wastewater samples; therefore, NoV GII was used as the target virus in the LR calculation. Since RT-MF-qPCR allows us to simultaneously quantify multiple RNA viruses in water samples, it may be one of the options to rapidly identify which viruses are present in quantifiable concen- trations in water samples.
We used two different values for the tolerable annual disease burden: 104 and 106 DALYpppy. The tolerable annual disease burden value of 106 DALYpppy for waterborne disease due to a certain pathogen is recommended by the World Health Organization (2006), while 104 DALYpppy is proposed as the tolerable disease burden for countries/regions with a high disease burden (Mara, 2011; Mara et al., 2010). The 2011 WHO guidelines for drinking water quality stipulate that waterborne disease may be considered to have little impact on the overall disease burden from all exposure routes (World Health Organization, 2011), which supports the employment of the tolerable annual disease burden value of 104 DALYpppy in the virus LR calculation. Mara argued that overly small values of the tolerable annual disease burden may raise other issues due to the excess costs in precautionary regula- tions (Mara, 2011). In fact, the actual annual disease burden of diarrheal disease in developing countries is estimated to be ~102 DALYpppy (Mara and Sleigh, 2010); therefore, the tolerable annual

446 T. Ito et al. / Water Research 125 (2017) 438e448
disease burden value of 104 DALYpppy is a more reasonable target reflecting the context of countries/regions with a high disease burden. Under exposure scenarios in this study, there was a 1-log10 difference in calculated LR target values of NoV GII between the tolerable disease burden values of 104 and 106 DALYpppy. Kitajima et al. experimentally demonstrated that an approximate 1-log10 difference in human norovirus reduction by chlorination led to an order increase in its CT values (expressed by disinfectant concen- tration and contact time) (Kitajima et al., 2010), which may have an economic impact on the operation of a wastewater reclamation process. Our suggestion is that the tolerable annual disease burden for wastewater reclamation and reuse should be subject to change based on the local context, including socio-economic background.
Several dose-response models and parameters for NoV GI and NoV GII were available for QMRA (Messner et al., 2014; Schmidt, 2015; Teunis et al., 2008). In order to calculate tolerable concen- trations of NoV GII in reclaimed wastewater, we used two different dose-response models, the hypergeometric model and the frac- tional Poisson model, according to recommendations in previous studies (Messner et al., 2014; Teunis et al., 2008; Van Abel et al., 2016). Van Abel et al. (2016) reported that different infection models for NoV GII gave different estimates of health risks. How- ever, our results showed little difference between the calculated tolerable concentrations using the two dose-response models because the calculated single-dose values corresponding to the annual disease burden were within a range not to show substantial difference in the infection probability. Dose-response models give sigmoid-like curves in general, and the lower dose range provides less divergence in the infection probability, as depicted in Fig. A1(b) by Van Abel et al. (2016). Since the single-dose value depends on the virus concentration, exposure volume, and frequency, it is crucial to test the sensitivity of these exposure parameters to the infection probability and DALYpppy under a local context of water reuse.
We performed sensitivity analyses to examine how uncertain and variable inputs (NoV GII concentration in untreated waste- water, exposure parameters, and reliability) would influence the LR target values of NoV GII as outputs in Scenario I and Scenario II in this study. Morgan et al. (1985) have noted that the consideration of uncertainty and variability in risk assessment is critical for decision-making. Spider plots were used in the sensitivity analysis to observe the changes in the calculated LR target values with variations of inputs; such plots have the advantage of making it possible to examine whether the relationship between inputs and outputs is linear or non-linear (Tian, 2013). The spider plots showed that the LR target values of NoV GII were influenced mostly by the change in logarithmic SD values of NoV GII in untreated wastewater among all the parameters tested. Furthermore, the larger loga- rithmic SD values led to a remarkable increase in the LR target values of NoV GII between 99% and 99.9% reliability. These emphasize the importance of accurate estimation of virus con- centration distribution used in the LR calculation. Water reuse operators also need to quantitatively monitor the target virus in source water even during the water reuse operation. If the peak- loading of a target virus is found during monitoring to be sub- stantially higher than that expected from the estimated probabi- listic distribution, it is necessary to implement additional protection measures in the wastewater reclamation process, such as the increment of disinfectant dose. These options for emergency operation need to be included in water reuse guidelines.
The LR target values of pathogens, including viruses, are indi- cated in the 2006 WHO guidelines, in which a 6- to 7-log10 reduction (including wastewater treatment and various health protection measures) is needed for unrestricted irrigation in order to achieve the tolerable annual disease burden value of 106
DALYpppy. The Queensland guidelines state that a 6.5-log10 reduc- tion is needed for the highest quality reclaimed water (Class Aþ), which can be used for agricultural irrigation (The state of Queensland, 2008). In the state of Victoria, reclaimed wastewater used for irrigation of vegetables eaten raw is classified as the highest quality reclaimed water (Class A). The Class A standard in Victoria requires a 7-log10 reduction in untreated wastewater (Environment Protection Agency Victoria, 2005, 2003). These LR values, including the WHO-recommended values, are approxi- mately 2- to 3-log10 higher than the calculated LR target values in this study when the reliability of the LR target value calculation was 99.9%. The difference in the LR target values might be caused by a difference in the virus type of concern, its concentration in un- treated wastewater, and the exposure parameters. The 2006 WHO guidelines listed the steps to calculate the LR target values for reclaimed wastewater irrigation of vegetable crops as follows: (1) 5000 rotavirus per liter of untreated wastewater, (2) 10 mL of treated wastewater remaining on 100 g of lettuce after irrigation, and (3) 100 g of lettuce consumed per person every second day throughout the year (World Health Organization, 2006). As mentioned above, the virus type of concern, its concentration in untreated wastewater, and the water reuse scheme should be based on the social and epidemiological background in the countries/re- gions where water reuse is implemented. One of the options in terms of pathogen concentration in untreated wastewater is to employ non-parametric (boot-strap) generation of a concentration dataset based on a systematic literature review and the results of the meta-analysis (Eftim et al., 2017), although the left-censored data issue cannot be overcome when the number of detected samples is small. The other option is to use the maximum virus concentration obtained from literature, which must give conser- vative LR values (Gerba et al., 2017). If countries/regions do not have the resources for monitoring the virus type of concern, a literature review of data for virus concentration is useful to determine the LR target values. In any case, we advocate that the guidelines for wastewater reclamation need to disclose the calculation steps of the LR target values along with the virus monitoring procedures. One of the options how to disclose the calculation steps is to pro- vide R codes for the calculations, such as those in the supplemen- tary information of this study, which enables a scheme proponent of wastewater reclamation to calculate the LR target values.
In this study, we calculated the LR target value based upon a QMRA framework. The currently available practice for determining the LR target value in each country/region is as follows: (1) accurate estimation of the initial concentration distribution of target viruses in untreated wastewater (Kato et al., 2013) and the LR values in each wastewater treatment unit prior to the design and installation of a wastewater reclamation system (Ito et al., 2015), and (2) pre- determination of the exposure scenarios for water reuse based on a local context. We concluded that the LR target values of virus of concern were influenced mostly by the change in the logarithmic SD values of virus concentration in untreated wastewater and reliability, compared to the other exposure parameters. It is also necessary to verify that the virus concentration in untreated wastewater does not depart from the estimated concentration distribution used in the calculation of target LR values during the operation of a wastewater reclamation system.
5. Conclusions
– When the accidental ingestion of reclaimed wastewater by Japanese farmers was assumed, the LR target values of NoV GII were 2.2, 3.4, and 4.7 at 95, 99, and 99.9% reliability, respectively, to meet the annual disease burden value of 104 DALYpppy.

– When daily consumption of salad vegetables grown with reclaimed wastewater was assumed, the LR target values of NoV GII were 0.7, 1.8, and 3.1 at 95, 99, and 99.9% reliability, respectively, to meet the annual disease burden value of 104 DALYpppy.
– An approximate 1-log10 difference of LR target values was observed between 104 and 106 DALYpppy.
– The LR target values of NoV GII were influenced mostly by the change in the logarithmic SD values of NoV GII concentration in untreated wastewater and reliability values, which indicated the importance of accurately determining the probabilistic distri- bution of reference virus concentrations in source water for water reuse.
Acknowledgements
This study was supported by the Japan Sewage Works Promo- tion Fund.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.watres.2017.08.057.
References
Asano, T., Leong, L.Y.C., Rigby, M.G., Sakaji, R.H., 1992. Evaluation of the California wastewater reclamation criteria using enteric virus monitoring data. Water Sci. Technol. 26, 1513e1524.
Atmar, R.L., Opekun, A.R., Gilger, M.A., Estes, M.K., Crawford, S.E., Neill, F.H., Ramani, S., Hill, H., Ferreira, J., Graham, D.Y., 2014. Determination of the 50% human infectious dose for norwalk virus. J. Infect. Dis. 209, 1016e1022. http:// dx.doi.org/10.1093/infdis/jit620.
Barker, S.F., Packer, M., Scales, P.J., Gray, S., Snape, I., Hamilton, A.J., 2013. Pathogen reduction requirements for direct potable reuse in Antarctica: evaluating hu- man health risks in small communities. Sci. Total Environ. 461e462, 723e733. http://dx.doi.org/10.1016/j.scitotenv.2013.05.059.
Bustin, S.A., Beaulieu, J.-F., Huggett, J., Jaggi, R., Kibenge, F.S., Olsvik, P.A., Penning, L.C., Toegel, S., 2010. MIQE precis: practical implementation of mini- mum standard guidelines for fluorescence-based quantitative real-time PCR experiments. BMC Mol. Biol. 11, 74. http://dx.doi.org/10.1186/1471-2199-11-74.
Bustin, S.A., Benes, V., Garson, J.A., Hellemans, J., Huggett, J., Kubista, M., Mueller, R., Nolan, T., Pfaffl, M.W., Shipley, G.L., Vandesompele, J., Wittwer, C.T., 2009. The MIQE guidelines: minimum information for publication of quantitative real- time PCR experiments. Clin. Chem. 55, 611e622. http://dx.doi.org/10.1373/ clinchem.2008.112797.
Costafreda, M.I., Bosch, A., Pinto, R.M., 2006. Development, evaluation, and stan- dardization of a real-time TaqMan reverse transcription-PCR assay for quanti- fication of hepatitis A virus in clinical and shellfish samples. Appl. Environ. Microbiol. 72, 3846e3855. http://dx.doi.org/10.1128/AEM.02660-05.
Da Silva, A.K., Le Saux, J.C., Parnaudeau, S., Pommepuy, M., Elimelech, M., Le Guyader, F.S., 2007. Evaluation of removal of noroviruses during wastewater treatment, using real-time reverse transcription-PCR: different behaviors of genogroups I and II. Appl. Environ. Microbiol. 73, 7891e7897. http://dx.doi.org/ 10.1128/AEM.01428-07.
Dawson, D.J., Paish, A., Staffell, L.M., Seymour, I.J., Appleton, H., 2005. Survival of viruses on fresh produce, using MS2 as a surrogate for norovirus. J. Appl. Microbiol. 98, 203e209. http://dx.doi.org/10.1111/j.1365-2672.2004.02439.x.
Eftim, S.E., Hong, T., Soller, J., Boehm, A., Warren, I., Ichida, A., Nappier, S.P., 2017. Occurrence of norovirus in raw sewage e a systematic literature review and meta-analysis. Water Res. 111, 366e374. http://dx.doi.org/10.1016/ j.watres.2017.01.017.
Environment Protection Agency Victoria, 2005. Guidelines for Environmental Management: Dual Pipe Water Recycling Schemesdhealth and Environmental Risk Management. Southbank, Victoria [WWW Document]. http://www.epa. vic.gov.au/~/media/Publications/1015 1.pdf (Accessed February 20 2017).
Environment Protection Agency Victoria, 2003. Guidelines for Environmental Management: Use of Reclaimed Water [WWW Document]. http://www.epa.vic. gov.au/~/media/Publications/464 2.pdf (Accessed February 20 2017).
Frenck, R., Bernstein, D.I., Xia, M., Huang, P., Zhong, W., Parker, S., Dickey, M., McNeal, M., Jiang, X., 2012. Predicting susceptibility to norovirus GII.4 by use of a challenge model involving humans. J. Infect. Dis. 206, 1386e1393. http:// dx.doi.org/10.1093/infdis/jis514.
Gerba, C.P., Betancourt, W.Q., Kitajima, M., 2017. How much reduction of virus is needed for recycled water: a continuous changing need for assessment? Water Res. 108, 25e31. http://dx.doi.org/10.1016/j.watres.2016.11.020.
Gibson, K.E., 2014. Viral pathogens in water: occurrence, public health impact, and
available control strategies. Curr. Opin. Virol. 4, 50e57. http://dx.doi.org/
10.1016/j.coviro.2013.12.005.
Ishii, S., Kitamura, G., Segawa, T., Kobayashi, A., Miura, T., Sano, D., Okabe, S., 2014.
Microfluidic quantitative PCR for simultaneous quantification of multiple vi- ruses in environmental water samples. Appl. Environ. Microbiol. 80, 7505e7511. http://dx.doi.org/10.1128/AEM.02578-14.
Ito, T., Kato, T., Takagishi, K., Okabe, S., Sano, D., 2015. Bayesian modeling of virus removal efficiency in wastewater treatment processes. Water Sci. Technol. 72, 1789e1795. http://dx.doi.org/10.2166/wst.2015.402.
Japan’s Ministry of Agriculture Forestry and Fisheries, 2013. The Japanese Market of Processed Vegetables (in Japanese).
Japan’s Ministry of Education Culture Sports Science and Technology, 2009. Washing and Disinfection of Food (in Japanese).
Japan’s Ministry of Internal Affairs and Communications, 2017. Statistics Japan (in Japanese) [WWW Document]. http://www.region-case.com/list-rank-consume/ (Accessed June 13 2017).
Japan’s National Institute of Infectious Disease, 2017. Detection of Gastroenteritis Viruses in Japan, 2014/15e2016/17 Season (As of January 26, 2017 [WWW Document]. http://www.nih.go.jp/niid/en/2013-03-15-04-55-59/2100-disease- based/na/norovirus/idsc/iasr-noro-e/5700-iasr-noro-e-150529.html (Accessed January 11 2017).
Kato, T., Miura, T., Okabe, S., Sano, D., 2013. Bayesian modeling of enteric virus density in wastewater using left-censored data. Food Environ. Virol. 5, 185e193. http://dx.doi.org/10.1007/s12560-013-9125-1.
Kitajima, M., Tohya, Y., Matsubara, K., Haramoto, E., Utagawa, E., Katayama, H., 2010. Chlorine inactivation of human norovirus, murine norovirus and poliovirus in drinking water. Lett. Appl. Microbiol. 51, 119e121. http://dx.doi.org/10.1111/ j.1472-765X.2010.02869.x.
Lewis, G.D., Metcalf, T.G., 1988. Polyethylene-glycol precipitation for recovery of pathogenic viruses, including hepatitis-a virus and human rotavirus, from oyster, water, and sediment samples. Appl. Environ. Microbiol. 54, 1983e1988.
Lindesmith, L., Moe, C., Marionneau, S., Ruvoen, N., Jiang, X., Lindblad, L., Stewart, P., LePendu, J., Baric, R., 2003. Human susceptibility and resistance to Norwalk virus infection. Nat. Med. 9, 548e553. http://dx.doi.org/10.1038/nm.
Mara, D., 2011. Water- and wastewater-related disease and infection risks: what is an appropriate value for the maximum tolerable additional burden of disease? J. Water Health 9, 217e224. http://dx.doi.org/10.2166/wh.2010.109.
Mara, D., Hamilton, A., Sleigh, A., Karavarsamis, N., 2010. More Appropriate Toler- able Additional Burden of Disease Improved Determination of Annual Risks. Norovirus and Ascaris Infection Risks. Extended Health-Protection Control Measures. Treatment and Non-Treatment Options. WHO-FAO-IDRC-IWMI, Geneva. Discussion paper: options for updating the 2006 WHO guidelines. http://www.who.int/water_sanitation_health/wastewater/guidance_note_ 20100917.pdf.
Mara, D., Sleigh, A., 2010. Estimation of norovirus and Ascaris infection risks to urban farmers in developing countries using wastewater for crop irrigation. J. Water Health 8, 572e576. http://dx.doi.org/10.2166/wh.2010.097.
Mara, D.D., Sleigh, P.A., Blumenthal, U.J., Carr, R.M., 2007. Health risks in wastewater irrigation: comparing estimates from quantitative microbial risk analyses and epidemiological studies. J. Water Health 5, 39e50. http://dx.doi.org/10.2166/ wh.2006.055.
Messner, M.J., Berger, P., Nappier, S.P., 2014. Fractional poisson – a simple dose- response model for human norovirus. Risk Anal. 34, 1820e1829. http:// dx.doi.org/10.1111/risa.12207.
Mok, H.F., Barker, S.F., Hamilton, A.J., 2014. A probabilistic quantitative microbial risk assessment model of norovirus disease burden from wastewater irrigation of vegetables in Shepparton, Australia. Water Res. 54, 347e362. http:// dx.doi.org/10.1016/j.watres.2014.01.060.
Morgan, M.G., Universiry, C., Morris, S.C., Deborah, A.L., 1985. Uncertainty in risk assessment. Environ. Sci. Technol. 19, 662e667.
Ottoson, J., Stenstro€m, T.A., 2003. Faecal contamination of greywater and associated microbial risks. Water Res. 37, 645e655. http://dx.doi.org/10.1016/S0043- 1354(02)00352-4.
Paulo, M.J., Van Der Voet, H., Jansen, M.J.W., Ter Braak, C.J.F., Van Klaveren, J.D., 2005. Risk assessment of dietary exposure to pesticides using a Bayesian method. Pest Manag. Sci. 61, 759e766. http://dx.doi.org/10.1002/ps.1060.
Rijsberman, F.R., 2006. Water scarcity : fact or fiction . Agric. water Manag. 80, 5e22. http://dx.doi.org/10.1016/j.agwat.2005.07.001.
Sano, D., Amarasiri, M., Hata, A., Watanabe, T., Katayama, H., 2016. Risk manage- ment of viral infectious diseases in wastewater reclamation and reuse: Review. Environ. Int. 91, 220e229. http://dx.doi.org/10.1016/j.envint.2016.03.001.
Schmidt, P.J., 2015. Norovirus dose-response: are currently available data infor- mative enough to determine how susceptible humans are to infection from a single virus? Risk Anal. 35, 1364e1383. http://dx.doi.org/10.1111/risa.12323.
Seitz, S.R., Leon, J.S., Schwab, K.J., Lyon, G.M., Dowd, M., McDaniels, M., Abdulhafid, G., Fernandez, M.L., Lindesmith, L.C., Baric, R.S., Moe, C.L., 2011. Norovirus infectivity in humans and persistence in water. Appl. Environ. Microbiol. 77, 6884e6888. http://dx.doi.org/10.1128/AEM.05806-11.
Shuval, H., Lampert, Y., Fattal, B., 1997. Development of a risk assessment approach for evaluating wastewater reuse standards for agriculture. Water Sci. Technol. 35, 15e20.
Symonds, E.M., Verbyla, M.E., Lukasik, J.O., Kafle, R.C., Breitbart, M., Mihelcic, J.R., 2014. A case study of enteric virus removal and insights into the associated risk of water reuse for two wastewater treatment pond systems in Bolivia. Water Res. 65, 257e270. http://dx.doi.org/10.1016/j.watres.2014.07.032.
T. Ito et al. / Water Research 125 (2017) 438e448 447

448 T. Ito et al. / Water Research 125 (2017) 438e448
Teunis, P.F.M., Moe, C.L., Liu, P., Miller, S.E., Lindesmith, L., Baric, R.S., Le Pendu, J., Calderon, R.L., 2008. Norwalk virus: how infectious is it? J. Med. Virol. 80, 1468e1476. http://dx.doi.org/10.1002/jmv.21237.
The state of Queensland, 2008. Water Quality Guidelines for Recycled Water Schemes [WWW Document]. https://www.dews.qld.gov.au/__data/assets/pdf_ file/0019/45172/water-quality-guidelines.pdf (Accessed February 20 2017).
The Victoria Department of Health, 2013. Guidelines for Validating Treatment Processes for Pathogen Reduction – Supporting Class a Recycled Water Schemes in Victoria.
Tian, W., 2013. A review of sensitivity analysis methods in building energy analysis. Renew. Sustain. Energy Rev. 20, 411e419. http://dx.doi.org/10.1016/ j.rser.2012.12.014.
Van Abel, N., Schoen, M.E., Kissel, J.C., Meschke, J.S., 2016. Comparison of risk pre- dicted by multiple norovirus dose-response models and implications for quantitative microbial risk assessment. Risk Anal. 37, 245e264. http:// dx.doi.org/10.1111/risa.12616.
Vinje, J., 2015. Advances in laboratory methods for detection and typing of nor- ovirus. J. Clin. Microbiol. 53, 373e381. http://dx.doi.org/10.1128/JCM.01535-14. Vrieze, S.I., 2012. Model selection and psychological theory: a discussion of the differences between the Akaike information criterion (AIC) and the Bayesian information criterion (BIC). Psychol. Methods 17, 228e243. http://dx.doi.org/
10.1037/a0027127.
World Health Organization, 2011. Guidelines for Drinking-water Quality, fourth ed. World Health Organization, 2006. Guidelines for the Safe Use of Wastewater,
Excreta and Greywater-volume 2: Wastewater Use in Agriculture [WWW Document]. http://www.who.int/water_sanitation_health/wastewater/ wwuvol2intro.pdf (Accessed February 20 2017).
Zurita, F., White, J.R., 2014. Comparative study of three two-Stage hybrid ecological wastewater treatment systems for producing high nutrient, reclaimed water for irrigation reuse in developing countries. Water 6, 213e228. http://dx.doi.org/ 10.3390/w6020213.