Spatial Variation Of Fipronil And Its Metabolites In Drinking Water And The Removal Efficiency Of Them Through Drinking Water Treatment Process

Background:Fipronil（FP）is widely used across the world as an emerging broad-spectrum phenylpyrazole insecticide and veterinary drug.FP degrades mainly into fipronil sulfide（FP-SFI）,fipronil sulfone（FP-SFO）,fipronil desulfinyl（FP-DES）,and fipronil amide（FP-AM）through photolysis,reduction,and oxidation.Recent studies indicate that fipronil and its metabolites（FPs）have been detected in aqueous environment of different lakes and rivers worldwide.However,their occurrence in drinking water and source water is limited in China.There is also lack of data on human exposure to fipronil via water ingestion.Meanwhile,FPs in source water also bring new challenges to the conventional drinking water treatment process,and pose a hidden threat to drinking water safety.Objective:We aimed to investigate spatial distribution characteristics of fipronil and its metabolites in drinking water throughout China and estimate daily intake of FPs via water ingestion for health risk assessment.We also aimed to examine the removal rates of FPs from the source water to the treated water and reveal the potential seasonal variation of FPs in taking drinking water from Wuhan.Methods:A total of 884 water samples,including tap water（n=789）and ground water samples（n=95）,were collected from Mainland China（including 31 provinces）and Hong Kong in June 2019.FPs,including FP,FP-SFI,FP-SFO,FP-DES and FP-AM in the extracted samples were determined by a Waters liquid chromatography coupled electrospray ionization（ESI,negative mode）triple quadrupole mass spectrometer（I-Class and Xevo TQ-S,LC-ESI-MS/MS;Waters,Milford,MA,USA）.For skewed distributions,the first and third quartiles were used with median to describe the variability of the distribution and the Wilcoxon or Kruskal-Wallis test was used to estimate the regional difference of FPs’levels in tap water and ground water.To assess health risk from FPs via consumption of drinking water,the estimated daily intake ofΣFPs（the sum of mass-volume concentrations of FP,FP-SFO,FP-SFI,FP-DES,and FP-AM）via drinking water ingestion(EDI_water)was calculated for all age groups under the median-and high-exposure（95th percentile）scenarios and compared with recommended reference dose（Rf D）.And 10 source water,10 treated water and 39 tap water samples were collected in Wuhan during July 2019（summer）to evaluate the removal efficiencies of FPs during drinking water treatment.Then,42 tap water samples were collected from Wuhan in October 2019（autumn）in which FPs were detected to compare with those in39 tap water samples collected from Wuhan in summer to reveal the seasonal variations of FPs in tap water.Results:（1）FP was the most common target analyte detected both in tap water and ground water,with detection frequencies of 55.3%for tap water（n=789）and 75.5%for ground water（n=95）.In ground water,ΣFPs varied from ND（not detected）to 3.67ng/L（median:0.04ng/L）,which was similar to tap water（range:ND-2.00ng/L;median:0.03ng/L）.Among the seven sub-regions,ΣFPs in the tap water samples from East China（median:0.03ng/L,75th percentile:0.31ng/L）were significantly（P<0.05）higher than those found from North（median:0.02ng/L,75th percentile:0.04ng/L）and Northwest China（median:0.03ng/L,75th percentile:0.04ng/L）.Moreover,the FP levels in the tap water from rural areas（median:0.03ng/L,75th percentile:0.04ng/L）were slightly but significantly（P<0.05）higher than those found for the capital cities[median:<MDL（method detection limit）,75th percentile:0.03ng/L]and middle/small cities（median:0.03ng/L,75th percentile:0.03ng/L）.The rural-urban distribution ofΣFPs were similar but the significant difference was only found between the rural areas and capital cities（P<0.05）.（2）Among different age groups,the highest daily intake ofΣFPs through water consumption was found in the infant group with EDI_water values of 2.70 and 82.8pg·kg^-1·d^-1 under the median-and high-exposure scenario,respectively.The EDI_water values for adult group(0.60 and 18.40pg·kg^-1·d^-1 under the median-and high-exposure scenario,respectively)were much higher.Overall,all the values of EDI_water were negligible compared to its recommended Rf D(0.2μg·kg^-1·d^-1).（3）Compared the concentrations of FPs from the source water with those from the treated water,all of the target analytes were mostly removed during the water treatment（with a decrease of 87.5-100%for FP,94.4-100%for FP-SFO,and complete removal of FP-SFI,FP-DES,and FP-AM）.It indicated that current drinking water treatment technologies could remove most FPs in source water.However,there were minute amounts of FPs in the corresponding tap water samples and their detection frequencies were higher than those for the treated water samples,implying the possibility of micro re-contamination after drinking water treatment.（4）FP levels in tap water samples from Wuhan in summer（detection frequency:71.8%,median:0.02ng/L,75th percentile:0.02ng/L）were much higher than those found in autumn（detection frequency:9.52%,median:<MDL,75th percentile:<MDL）.Similarly,ΣFPs in tap water samples in summer（median:0.02ng/L,75th percentile:0.04ng/L）were much higher than those found in autumn（median:ND,75th percentile:0.02ng/L）.Conclusions:This work is the first study on the residue levels of the FPs in drinking water throughout China.In this study,FP was the most common target analyte detected both in tap water and ground water.Among the seven sub-regions,FPs in water from East China were significantly higher than those found for North and Northwest China and required more attention than those in the other regions.For all age groups,the highest daily intake ofΣFPs through water consumption was found in the infant group while all the calculated EDI_watervalues ofΣFPs were much lower than the recommended Rf D values,suggesting that daily intake level ofΣFPs through water consumption were safe.Besides,current drinking water treatment technologies could remove most FP residue in source water but there was possibly occurrence of micro re-contamination in tap water after drinking water treatment.Additionally,the detection frequency decreased strongly in autumn relative to that in summer,possibly because of the higher usage of fipronil insecticide during summer.