Environmental risk assessment of triclosan and triclocarban from personal care products in South Africa

Environmental risk assessment of triclosan and triclocarban from personal care products in South Africa

Environmental Pollution 242 (2018) 827e838 Contents lists available at ScienceDirect Environmental Pollution journal homepage: www.elsevier.com/loca...

838KB Sizes 0 Downloads 68 Views

Environmental Pollution 242 (2018) 827e838

Contents lists available at ScienceDirect

Environmental Pollution journal homepage: www.elsevier.com/locate/envpol

Environmental risk assessment of triclosan and triclocarban from personal care products in South Africa* N. Musee Emerging Contaminants Ecological and Risk Assessment (ECERA) Research Group, Department of Chemical Engineering, University of Pretoria, Private Bag X20, Hatfield, 0028, Pretoria, South Africa

a r t i c l e i n f o

a b s t r a c t

Article history: Received 8 February 2018 Received in revised form 15 June 2018 Accepted 30 June 2018 Available online 12 July 2018

Trends in the widespread use of personal care products (PCPs) containing triclosan (TCS) and triclocarban (TCC) have led to continuous emissions of these chemicals into the environment. Consequently, both chemicals are ubiquitously present at high concentrations in the aquatic systems based on widely reported measured environmental concentration (MECs) data in different environmental systems (e.g. freshwater) worldwide, especially in developed countries. In developing countries, however, lack of MECs data is a major issue, and therefore, inhibits effective risk assessment of these chemicals. Herein, TCS and TCC releases from personal care products (PCPs) were quantified, using a modelling approach to determine predicted environmental concentrations (PECs) in wastewater, freshwater, and soils, and likely risk(s) were estimated by calculating risk quotient (RQs). TCS and TCC in freshwater had RQs >1 based on estimated PECs with wide variations (z2e232) as performed across the three dilutions factors (1, 3, and 10) considered in this study; an indicator of their likely adverse effect on freshwater organisms. In untreated and treated wastewater, TCS RQs values for bacteria were >1, but <1 for TCC, implying the former may adversely affect the functioning of wastewater treatment plants (WWTPs), and with no plausible impacts from the latter. In terrestrial systems, RQ results for individual chemicals revealed no or limited risks; therefore, additional investigations are required on their toxicity, as effects data was very limited and characterised by wide variations. Future national monitoring programs in developing countries should consider including TCS and TCC as the results suggest both chemicals are of concern to freshwater, and TCS in WWTPs. Potential risks of their metabolites remain unquantified to date. © 2018 Elsevier Ltd. All rights reserved.

Keywords: Risk quotient Aquatic organisms Ecological risk Antibacterial Hazard

Quantitative risk assessment of TCS and TCC from consumer products to the environment were found to exhibit wide variations in freshwater systems (RQs z 2e232) dependent on factors like spatial, temporal, income per capita, concertation of chemical per article, water flows, etc. whereas TCS showed potential to adversely influence the functioning of WWTPs, and no risk to the terrestrial systems as RQs « 1. For the first time this study has quantified risks of TCC to the environment using bottom-up approach, and in a region with high resolution as opposed to entire country or continent. 1. Introduction Triclosan (2,4,4 -trichloro-2 -hydroxydiphenyl ether; TCS), and

* This paper has been recommended for acceptance by Dr. Harmon Sarah Michele. E-mail addresses: [email protected], [email protected].

https://doi.org/10.1016/j.envpol.2018.06.106 0269-7491/© 2018 Elsevier Ltd. All rights reserved.

(non-phenolic) carbanilide triclocarban (1-(4-chlorophenyl)-3(3,4-dichlorophenyl) urea; TCC) are widely used polychlorinated, binuclear, and aromatic antimicrobials (Halden and Paull, 2005; Carey and McNamara, 2015) estimated to be contained in over 2000 different personal care products (PCPs), household, and medical products in the USA market (Young, 2013; Smith, 2013), and perhaps even higher (Halden, 2014). Similarly, high usage trends for these antimicrobials have been observed in other markets globally (APUA, 2011; Bedoux et al., 2012). TCS and TCC are high volume usage chemicals with published production quantities characterised by significant variations globally. For example, estimated annual global production volumes for TCC ranged from 0.0006 to 10 000 tonnes (TCC Consortium, 2002; TSCA, 2003; Miller et al., 2008); whereas about 1500 tonnes per year of TCS enter consumer markets globally (Singer et al., 2002). Therefore, the increasing use and wide production of PCPs and concomitant incomplete removal in wastewater treatment plants (WWTPs) of TCS (Chalew and Halden, 2009; Chen et al., 2011;

828

N. Musee / Environmental Pollution 242 (2018) 827e838

Narumiya et al., 2013; Sun et al., 2014), and TCC (Heidler et al., 2006; Lozano et al., 2013; Narumiya et al., 2013) have resulted in their associated high presence and concentrations in different environmental compartments such as soil, freshwater, etc., including from diffuse sources such as run-offs. Numerous environmental monitoring studies, for example, indicate widespread occurrences of TCS and TCC in treated wastewater (Ying and Kookana, 2007; Amdany et al., 2014; Madikizela et al., 2014), freshwater systems (Halden and Paull, 2005; Ying et al., 2007; Lv et al., 2014; Pintado-Herrera et al., 2014), soils (Cha and Cupples, 2009; Al-Rajab et al., 2015; Healy et al., 2017), sediments (Singer et al., 2002), groundwater (Gottschall et al., 2012), and sludge (Cha and Cupples, 2009; Fu et al., 2016; Sherburne et al., 2016; Healy et al., 2017) reported as measured environmental concentrations (MECs). Most of these environmental monitoring investigations, however, were done in developed countries such as USA (Sapkota et al., 2007; Yu and Chu, 2009), Canada (Hua et al., 2005), United Kingdom (Sabaliunas et al., 2003), Switzerland (Lindstrom et al., 2002), and China (Zhao et al., 2013) to mention a few, with not many studies in developing world, including South Africa, as recently highlighted by Ebele et al. (2017). Due to limited MECs of TCS and TCC data in the developing countries make it impossible to quantify their likely risks in the environmental systems. Moreover, TCS and TCC environmental discharges in the developing world are from treated and untreated wastewater due to limited coverage of WWTP facilities since a considerable portion of the population lack sanitation services, as well as from run-off. As such, this raises the need to understand and quantify the potential risks of TCS and TCC implications to the aquatic systems; and the approach proposed herein seek to achieve this objective. TCS usage was estimated to be about 85%, 5%, and 10% of the total annual production, respectively, in PCPs, textiles, and plastics and food contact materials (SCCP, 2009). Recently, PCPs were estimated to account for 68% of the global TCS use; with the rest 16%, 8%, and 8%, respectively, in disinfectants and medical supplies, paints, and plastics (QYResearch, 2015). About 84% of antimicrobial bar soaps in the USA contained TCC (Perencevich et al., 2001). Although TCS and TCC have short useful lifespans of the order of seconds during the use phase (Borchgrevink et al., 2013), they exhibit long lifespans in different environmental systems, from a few days to years (Ying et al., 2007). For example, each agent has a half-life of 60 days in water; and in soils the half-lives of TCS and TCC are 120 and 180 days, respectively (Clarke et al., 2016); and hence are categorized as biopersistent (Ying et al., 2007; Daughton and Ternes, 1999; Higgins et al., 2011; Ebele et al., 2017) and bioaccumulative in aquatic organisms (Coogan et al., 2007; Coogan and La Point, 2008). TCS and TCC ecotoxicological data show they can induce adverse effects such as; antibiotics resistance by various microbial communities (Levy, 2002; Yazdankhah et al., 2006; Oggioni et al., 2013), alterations to microbial community structures (Carey and McNamara, 2015; Carey et al., 2016), and induction of high toxicity to the aquatic organisms compared to other disinfectants (Brausch and Rand, 2011). In addition, TCS and TCC can potentially induce adverse human health impacts such as endocrine disruption effects (Witorsch and Thomas, 2010; Pollock et al., 2014; Lee et al., 2014), birth defects (Geer et al., 2016), increased risk to obesity (Lankester et al., 2013), reduction in quality of sperm in men (Zhu et al., 2016), and likelihood for the proliferation of cancer cells (Dinwiddie et al., 2014; Kim et al., 2014; Winitthana et al., 2014). Three studies in South Africa have reported TCS MECs in wastewater, one in freshwater, but none in sediments, or soils (Table S1). Amdany et al. (2014) quantified TCS in the Johannesburg area with the concentrations in the influent and effluent found to

be extremely high compared to other regions globally, and similar findings in the same region have been documented by Lehutso et al. (2017). Also, Madikizela et al. (2014) detected TCS in the influent and effluent in Durban (Table S1). The high TCS MECs reported in South Africa were attributed to wide use of PCPs, incorporation of high TCS concentrations in PCPs compared to other countries, and WWTPs’ low removal efficiency for chemicals since many operate above design capacity. However, in South Africa only a single study for TCC MECs values was found in the literature (Table S1) (Lehutso et al., 2017), although it has higher affinity to accumulate in the environment compared to TCS (Brausch and Rand, 2011). To fill these data gaps, mathematical models can prove to be valuable and powerful tools, useful for the estimation of these antibacterial emissions, environmental concentrations, and potential risks to the environment. Such results are essential to provide an indication of which environmental compartment(s) may experience unacceptable risk levels even if currently their experimental and monitoring data is missing. Previously, predicted environmental concentrations (PECs) of TCS in the environment were estimated using substance flow analysis (SFA) (Huang et al., 2014). To the author's knowledge, however, none has to date estimated TCC flows into the environment using life-cycle based approaches. Thus, this study presents first results of this kind for TCC. In the present study, the aim was to refine the estimates of TCS and TCC flows using material flow analysis (MFA) into natural (soil and freshwater) and technical (wastewater, sludge, and landfills) systems. Next, risk estimation was performed by calculating risk quotient (RQ) at regional level. The RQs values were used in this paper as an indicator on the levels of ecotoxicological risks posed by TCS and TCC in different environmental compartments. Overall, the specific objectives of this work were: (i) through retail market survey and on-line marketing websites develop an inventory for PCPs (containing TCS and TCC), and use patterns in South Africa; (ii) use MFA to quantify TCS and TCC releases from PCPs annually into natural and engineered systems by calculating the PECs; (iii) to collect ecotoxicological data (acute and chronic toxicity) for organisms in different environmental systems from published literature to determine predicted no effect concentrations (PNECs); and (iv) estimate TCS and TCS risk (expressed as RQ) towards different organisms (algae, fish, etc.), and in different compartments by comparing PECs to PNECs, and subsequent implications to the environment.

2. Material and methods 2.1. Defining system and system boundary In this study, we adopted a well-defined system boundary that considered a given contaminant: (i) to flow from grave to cradle; (ii) to have several ways in which its waste streams are managed; and (iii) to have probable eventual impacts on natural and technical systems. The provincial administrative boundaries, in South Africa, were considered as system boundaries for the MFA modelling. The Gauteng Province (GP), one of the nine provincial administrative regions in South Africa was chosen as the system boundary to estimate TCS and TCC risks; motivation for the choice is outlined in supporting information (Section S1). A granular approach was adopted to develop the model reported herein to allow adequate flexibility to make further modifications as new data become available, or to achieve a certain purpose and need in a specific case study. A temporal boundary of one year (2014) was used; which can easily be extended to previous and future years subject to data accessibility and purpose.

N. Musee / Environmental Pollution 242 (2018) 827e838

2.2. Inventory of PCPs containing TCS and TCC The necessity to develop an inventory for PCPs containing TCS and TCC in South Africa was due to several reasons, viz.: paucity of MECs data for TCS save three studies (Amdany et al., 2014; Madikizela et al., 2014; Lehutso et al. (2017), and one for TCC (Lehutso et al. 2017) in various locations, regions, or exposure media systems; lack of annual production or imported quantities of TCS and TCC into South Africa; and absence of use data on distribution per different PCPs categories by consumers in different income categories. Hence, product inventory developed and reported herein served as model input data, and was limited to PCPs as they account for the highest use of TCS and TCC as reported elsewhere (Perencevich et al., 2001; SCCP, 2009; QYResearch, 2015). Data for TCS- and TCC-containing consumer products were sourced from websites, retail stores, and published literature where the following procedure was followed. First, information on PCPs known to contain TCS or TCC (or both) were collected from published peer-reviewed and grey literature. Secondly, through visits to retail stores, product categories containing TCS or TCC based on product labels were identified, and data captured in a database. Through targeted online searches and use of key words (single or combined) additional products were identified e and collaborated with evidence of their availability/presence in the retail market. The search terms used are defined in the supporting information (Section S2). 2.3. Estimation of TCC and TCS flows The MFA approach was used to quantify TCS and TCC concentrations in the environment following releases during products use and disposal phases. The model results aided to gain better understanding on the fate processes of these antibacterial agents into different environmental compartments as final sinks. The static MFA (Brunner and Rechberger, 2004), which assumes the system to be in a steady state was used within a system defined in space (GP) and time (1 year) to track the flows of TCS and TCC from PCPs. Importantly, the steady state assumed allowed the estimation of each chemical's concentration in different environmental compartments. TCS and TCC releases into the environment were quantified using bottom up-based approach, where the key model input parameters were: market share of a given sub-product category (%), TCS or TCC concentration incorporated in a given product (w/w %), and daily usage per given product and/or product category (g/ capita/d). The following expression was used in estimating TCS or TCC from a specific product type using the expression (Zhang et al., 2015; Musee, 2017):

Mi;j;totali ¼

X

Mi;j X  ¼ Ci;j *DQi;j *MSj ðThe three equations in this paper have the font so large

they stand out disprotinately larger than the text: Please consider to reduce to be the same font as textÞ: (1) where Mi;j is quantity of chemical i (i ¼ TCS or TCC) from PCPj (j ¼ deodorants, toothpaste, soaps, etc.) released into GP region annually (kg), Mi;j;totali as total mass of chemical i, Ci;j concentration of active chemical i in PCPj, DQi;j the daily usage per person of PCPj containing chemical i (g/ca/d) based on ECB (2003) and USEPA (2011) data, and MSj represents market penetration of a given PCPj in South Africa (dimensionless quantity). Therefore, the total

829

quantity of chemical i released into the environment was a summation of masses for a given chemical i from different consumer product categories. How each model input parameter (daily release rate, market share, and concentration) was derived is described in supporting information (Section S3.1e S3.3 ). 2.4. Risk assessment Risk assessment entails the identification of potential adverse consequences along with their severity and likelihood (Jenkin et al., 2007); and consists of three key steps, namely; exposure assessment, hazard assessment, and risk characterization. 2.4.1. Environmental exposure assessment Previously, exposure environmental concentrations used to estimate the potential risks of TCS (Tamura et al., 2013; Huang et al., 2014; Guo and Iwata, 2017), and TCC (Tamura et al., 2013) were based on MECs. However, in this study, environmental concentrations the PECs were estimated using model input data described in supporting information in Sections S2.2, S2.3, and S4 by combining different types and sources of data into an analytical structure. Due to uncertainties of model input data, three release scenarios were considered, viz.: minimum, probable, and maximum for TCS and TCC into different environmental systems following the approach described in Musee (2011). TCS and TCC are known to undergo several processes, like fixation, degradation, inactivation and transportation once released into the environment (Petrie et al., 2014; Verlicchi and Zambello, 2015). However, in this study transformation processes were not considered due to lack of fate data for TCS and TCC, for example, in the agricultural soils and WWTPs. 2.4.2. Hazard characterization PNECs were derived from the experimental dose-response assessment ecotoxicological data published in scientific literature following prescribed procedures in Technical Guidance Document (TGD) (ECB, 2003). PNEC is the ratio of the no observed effect concentration (NOEC) to the assessment factor (AF) for any given environmental compartments expressed as:

PNECi ¼

NOECi AF

(2)

where i refers to freshwater, soils, and wastewater environmental systems. In cases where NOECs are lacking, EC50 or LC50 values were used instead in Eqn. (2), and AF adjusted accordingly (ECB, 2003). Owing to wide variability of ecotoxicological data collected for three organisms (fish, Daphnia magna, and algae) in different environmental compartments, AF applied in a given case is dependent on overall quality of database and the endpoint covered (ECB, 2003); and varies between 10 and 1000. A factor of 1000 was used on acute toxicity data (EC50 or LC50) as a protective factor, and designed to identify chemicals with potential to induce adverse effects during effects assessment process. Conversely, AF of 10 was used for chronic toxicity data (NOEC or LOEC). Lowest value for NOEC, EC50 or LC50 was chosen for each case organism under consideration to estimate plausible worst-case scenario environmental risk. Ecotoxicological data for organisms in soil (plants, earthworms, and microorganisms in soil), wastewater (microorganisms in WWTPs), and freshwater (fish, D. magna, and algae) were used in this study. 2.4.3. Risk characterization Risks of TCS and TCC in a specific compartment were estimated by calculating risk quotient (RQ) following the TGD on risk

830

N. Musee / Environmental Pollution 242 (2018) 827e838

assessment (ECB, 2003) as means of expressing risk posed by a particular chemical to a specific organism, and expressed as:

RQi ¼

PECi PNECi

(3)

According to European Chemicals Bureau (2003), risk is classified based on a binary system where for RQ  1, a likely appreciable risk exists to the environment whereas for RQ < 1, ecotoxicological risk is considered to be unlikely. The binary classification system, however, is limited and so broad; hence several categorization schemes for RQ have been proposed (Table S2). For data interpretation and categorization of TCS and TCC ecotoxicological risks, the framework by Lemly (1996) was followed.

2005); with bar soaps as most dominant product type (69.2%) followed by hand wash products (13.5%), liquid soaps (7.7%), and remaining 9.6% being used in shower gels and aftershave. In the USA retail market survey, TCC was found predominantly in liquid soaps and bar soaps (Perencevich et al., 2001). In this study, only 88 (68.8%) and 45 (86.5%), respectively, of the TCS- and TCC-containing products (Table S3) were used for the estimation of their likely risks to the environmental systems. The products excluded from this model was either because they were non-PCPs (e.g. kitchenware or office accessories), or due to inaccessibility of data to determine their market penetration in South Africa. 3.2. Quantification of TCS and TCC flows in GP

3. Results and discussion 3.1. Inventory analysis The inventory database developed in this study had 128 and 52 products that contained TCS and TCC, respectively (Table S3). The TCS- and TCC-containing products had each 16 and 5 subcategories as shown in Table S3. For a product to be included in this study, it had not only be found online but also confirmed being in the retail shops at the time of data collection phase of this study. Broadly, TCS-containing products were in three broad categories, and distributed as follows: 115 were PCPs (89.8%), 8 were kitchenware (6.3%), and 5 were classified as others (pet care, office accessories, and first aid) (3.8%). The results indicate the distribution of products containing TCS in South African retail market were similar to those reported elsewhere (SCCP, 2009; QYResearch, 2015) where the PCPs accounted for the highest use of TCS predominantly in deodorants (44.5%), and liquid hand soaps (14.8%). For TCC-containing products, all were in PCPs category (Table S3), and results herein are in good agreement to previous inventories reported in the USA (Perencevich et al., 2001; Halden and Paull, 2005), and European Union (European Commission,

Table 1 Estimated daily usage rates for different PCPs in GP, South Africa. Product category

Different income groups usage (g/p/d) Very low

Low

Medium

High

Bar soaps Liquid soaps Hand wash Shower gel Aftershave Toothpaste Body lotion Hand lotion Deodorants Face treatments Hair products

0.223 0.225 0.084 0.464 0.074 0.130 0.711 0.024 0.371 0.251 0.464

0.282 0.286 0.107 0.589 0.094 0.165 0.902 0.031 0.471 0.319 0.589

0.671 0.679 0.254 1.399 0.224 0.392 2.145 0.075 1.119 0.757 1.399

3.600 3.640 1.360 7.500 1.200 2.100 11.500 0.400 6.000 4.060 7.550

To avoid overestimation of TCS and TCC quantities into the GP environment from different PCPs; first, the daily usage rates were estimated for specific product category per income group in South Africa. This is due to lack of published data on daily usage per capita for PCPs in South Africa. Using Eqs. (1) and (2) in supporting information (Section S3.1), and through integration of complex data for daily usage rates published by ECB (2003) and USEPA (2011) (Table S4), population (Table S5), and income per capita (Table S6), daily usage rates for different products per income groups were estimated for GP, and the results summarised in Table 1. Next, following the approach outlined in Section 2.3; quantities of TCS and TCC from PCPs in GP into wastewater, landfills, and freshwater (via run-off and incompletely treated effluent) were estimated. Estimates of TCS and TCC quantities from different product categories (e.g. deodorants and associated sub-categories) are given in Table 2. Estimated TCS and TCC under most probable scenario released into the environment were 332.2 kg and 1763.7 kg, respectively (Table 2). Notably, bar soaps accounted for 68.5% (Table 2) of the total TCC released into environment since they are the dominant TCC-containing products in South Africa retail market (account for 69.2% (Table S7)), and also are generally more affordable compared to other product categories e.g. liquid soaps and shower gel. Moreover, the high per capita consumption of TCC compared to TCS was attributed to the high allowable concentration of TCC incorporated in PCPs (in South Africa) with a maximum of 1.5% (Table S7) compared to 0.3% for TCS (Table S8) per product article according to FCDA regulations (see supporting information Section S3.2 ). And, per capita consumption of TCS and TCC in GP was estimated, respectively, as 25.17 and 133.6 mg/c/ pa; with the obtained GP values being lower than those for other regions or countries globally (Table S9). For consistency regarding the inclusion or exclusion of products summarised in Tables S7 and S8, each brand product had meet one or two features, namely: (i) a label stating TCC or TCS as one of the product ingredients, and (ii) mass and/or volume or the percentage of TCC or TCS being specified. Such data were used to estimate the flows of either TCC or TCS

Table 2 Estimated quantities (kg) of TCS and TCC from PCP's into the GP environment (2014). Income group

TCS Product categories

TCC Product categories

T

BL

BS

HL

D

FT

HP

BA

LS

BS

HW

LS

SG

AS

Very low Low Medium High Sub-total

2.73 14.65 34.82 44.18 96.37

0.15 0.80 1.91 2.42 5.28

0.06 0.32 0.76 0.97 2.11

0.01 0.04 0.09 0.12 0.25

5.62 30.15 71.65 90.91 198.33

0.80 4.28 10.18 12.91 28.16

0.02 0.09 0.22 0.28 0.61

0.00 0.01 0.03 0.04 0.09

0.03 0.16 0.37 0.47 1.03

34.27 183.74 436.68 554.05 1208.73

2.16 11.58 27.53 34.93 76.20

12.01 64.43 153.12 194.27 423.83

1.42 7.64 18.15 23.03 50.25

0.13 0.72 1.71 2.16 4.72

*T: toothpaste; BL: body lotion; BS: bar soaps; HL: hand lotion; D: deodorant; FT: face treatment; HP: hair product; BA: bath additive; LS: liquid soaps; HW: hand wash; SG: shower gel; AS: after shave.

N. Musee / Environmental Pollution 242 (2018) 827e838

831

Fig. 1. Calculated PNECs of TCS in freshwater, wastewater, and terrestrial environmental systems for: (a) chronic, and (b) acute toxicity data. Symbol (o) represents PNECs for freshwater and wastewater organisms, and (*) for terrestrial organisms.

from a specific brand product into the environment. Deodorants and toothpastes accounted for 59.7% and 29% of the total TCS released into the environment, respectively. By comparing the quantities of TCS released with reference to the number of products in a given sub-category yielded valuable insights. For example, toothpaste had only seven sub-categories of TCScontaining products compared to 19 and 57 for liquid soaps and deodorants, respectively (Table S3), but released considerably high quantities of TCS into the environment. This points to likelihood of other drivers on the quantities of TCS or TCC released besides the number of products per given sub-category per given product (Tables S3 and 2). The high market penetration, for toothpaste products compared to other product categories (Table S3), for instance, where liquid soaps had 19 sub-category products but only released 1.03 kga1 (Table 2) supports this view point. Overall, highly priced products such as hand lotion, face treatment, hair products, and bath additives, among other accounted for very low quantities of TCS releases into the environment associated with low market penetration. This was further collaborated with low volumes of products per given category in GP in these categories (data not shown here). Estimated quantities of TCS or TCC from various products were found to be dependent on factors, viz.: income per capita, market penetration, and concentration in a given product. TCC, for

example, market penetration varied from 0.03% to 1.86% whereas concentration per article ranged from 0.02% to 1.5%, and was dependent on product category in focus (Table S7). Thus, although liquid soaps had fewer sub-categories (4) compared to hand wash (7), the former accounted for higher quantities of TCC release into the environment. This was plausibly due to on average higher concentration of TCC in liquid soaps per article than in hand wash (Table S1).

3.3. Predicted no effect concentration Results of dose-response acute- and chronic-toxicity data collected from the literature were used to calculate PNEC values for organisms in freshwater (algae, D. magna, and fish), wastewater (bacteria), and terrestrial (earthworms, etc.) systems. PNECs results for the acute and chronic toxicity of TCS and TCC, respectively, are given in Figs. 1 and 2. For TCS (Fig. 1), results for chronic- and acutetoxicity showed algae as the most sensitive organism based on the set of data assessed in this study, and ranged in 5 and 3 orders of magnitude, respectively. Thus, NOECs for TCS yielded the least PNEC of 0.0015 mg/L e an indication of highest toxicity. Our findings indicate good agreement, and are comparable to other researchers’ previous results (Brausch and Rand, 2011; Tamura et al., 2013). In wastewater, the bacteria exhibited high toxicity variability,

Fig. 2. Calculated PNECs of TCC in freshwater, wastewater, and terrestrial environmental systems for: (a) chronic, and (b) acute toxicity data. Symbol (o) represents PNECs for freshwater and wastewater organisms, and (*) for terrestrial organisms.

832

N. Musee / Environmental Pollution 242 (2018) 827e838

with TCS PNEC values spanning over 7 orders of magnitude (acute toxicity: Fig. 1b) plausibly due to high diversity of bacteria strains present in WWTPs. Secondly, the current test systems of measuring the effects of chemicals on microbial activity are based on different endpoints, and hence exhibits high levels of sensitivity (ECB, 2003), which in turn, partly may have contributed to the high toxicity variability. D magna was least sensitive organism with the PNECs spreading over 2 orders of magnitude, and followed by fish. The calculated PNECs in this study showed good agreement with results of Tamura et al. (2013); where TCS was most sensitive to algae, similar to earlier review findings of Brausch and Rand (2011). The high TCS sensitivity in algae is attributed to its inherent antibacterial characteristics through mechanisms such as disruption of lipid synthesis pathways, membrane destabilization, or uncoupling of oxidative phosphorylation (Brausch and Rand, 2011). Conversely, in the terrestrial systems, organisms with the lowest PNECs were algae and plants species. Due to limited acute- and chronic-toxicity data for TCC in the literature, fewer PNECs were estimated (Fig. 2), and the observed variability was significantly lower compared to TCS. For TCC, the most sensitive organism was D magna, and algae as the least sensitive (1 order of magnitude). Only a single acute toxicity of TCC was found for bacteria, and none for chronic exposure. Notably, given the PNECs of TCS and TCC to bacteria in wastewater are low (<0.1), may suggest likely diminished treatment efficacy and microbial diversity in activated sludge units during wastewater treatment, and in turn, lead to potential environmental and human health concerns due to inadequate treatment. The terrestrial organism with the lowest PNEC for both chronic and acute data was found to be Eisenia fetida (red worm). 3.4. Predicted environmental concentrations To assess ecological risks, PECs of TCS and TCC were performed using estimated quantities released into different environmental compartments in GP under three plausible scenarios (Table 3). Estimated PECs were dependent on factors like WWTPs removal efficiency of TCS and TCC, fraction of treated wastewater based on

population that receives sanitation services in GP, dilution factor, and income per capita for different income groups. For example, the removal efficiencies of TCC and TCS are largely similar (Heidler et al., 2006; Lozano et al., 2013); however, in this study lower removal efficiencies for South Africa's WWTPs were used (Musee, 2011) when compared to more developed countries where removal efficiency of about 97% for TCC and TCS was reported (Heidler et al., 2006; Lozano et al., 2013). Due to lack of transformation data for TCS and TCC in the WWTPs, the removed quantities were considered to have accumulated in the sludge. Results in Table 3 show that about 5 times higher quantities of TCC were released into the environment compared to TCS in GP. 3.4.1. PECs in terrestrial systems The sludge generated from GP was calculated using flows and treated effluent data for 48 WWTPs with total daily flow of 2 552.80 ML/d (Table S10) where 350 kg dry solids are produced for every ML treated (Marx et al., 2004). The estimated total sludge was 326 082 t/a. Agricultural application of sludge in South Africa is about 80e97.4% (Musee, 2011), and sludge application rate of 10 000 kg/ha (Synman and Herselman, 2006); thus, in this study we estimated PECs under three application scenarios, and the results are shown in Table 4. In all three scenarios considered; estimated PECs were very low for TCS and TCC with the maximum application worst-case concentrations being 3 mg/kg and 15 mg/kg, respectively. A report by Government of Canada (2016) using the same approach applied in this study estimated PEC of 68 mg/kg (termed as highly conservative) for TCS (with biosolids application rate of 8 300 kg/ha (Environment Canada, 2006)). S anchez-Brunete et al. (2010) reported measured TCS concentrations of 4.7 mg/kg and 1.7 mg/kg in agricultural soil sampled 1 day and 6 months following biosolids application (application rate of 12 000 kg/ha). Hence, estimated PECs in terrestrial systems in this study are in the same order of magnitude with previous studies (Government of Canada, nchez-Brunete et al., 2010), however, the differences may 2016; Sa be attributed to measured and predicted concentrations in addition to application rates used in different regions globally.

Table 3 Estimated quantities of TCC and TCS released into the environment based on model inputs under the minimum, probable, and maximum scenarios. Variable

TCC

Quantity released into Env. fWWTP: fraction of treated WW fR: fraction removed in WWTPs Quantity released into WWTPs Quantity in the sludge Quantity in the effluent from WWTPs Quantity released through run-off Total quantity into freshwater

TCS

MIN

PRO

MAX

MIN

PRO

MAX

1675.54 0.93 0.83 1558.25 1290.99 267.26 117.29 384.55

1763.73 0.91 0.68 1604.99 1097.00 507.99 158.74 666.73

1851.91 0.89 0.58 1648.20 957.05 691.15 203.71 894.86

315.63 0.93 0.84 293.54 245.70 47.84 22.09 69.93

332.25 0.91 0.67 302.01 201.57 100.45 30.23 130.68

348.86 0.89 0.59 310.48 182.63 127.86 38.37 166.23

Abbreviations: Env.: environment, WW: wastewater, WWTPs: wastewater treatment plants. All fractions are dimensionless, and quantities are expressed in kga1.

Table 4 Estimated PECs and RQ values for TCC and TCS in the terrestrial systems. Variable

TCC

TCS

MIN

PRO

MAX

MIN

PRO

MAX

Sludge used in agriculture (%) Total sludge used in farms (t/a) TCC or TCS in sludge (kg) Conc. in sludge (mg/kg) PECsoil (mg/kg) PNECsoil (mg/kg) RQsoil

0.800 2.61  105 1290.99 4950 15 40 0.36

0.887 2.9  105 1083.11 3740 11 40 0.28

0.974 3.18  105 944.94 2980 9 40 0.22

0.800 2.61  105 245.70 940 3 20 0.16

0.887 2.9  105 201.57 700 2.4 20 0.12

0.974 3.18  105 182.63 570 2 20 0.10

N. Musee / Environmental Pollution 242 (2018) 827e838

833

Table 5 Estimated PECs, PNECs and RQ values for TCC and TCS in the freshwater and wastewater systems. Environmental compartment

Freshwater

Wastewater (treated effluent)

PEC, and dilution factor (DF)

DF: 10, (PEC, ng/[) DF: 3, (PEC, ng/[) DF: 1, (PEC, ng/[) PNEC (ng/[) RQ (DF ¼ 10) RQ (DF ¼ 3) RQ (DF ¼ 1) PECinfluent (ng/[) PECeffluent (ng/[) PNEC (ng/[) RQinfluent RQeffluent

TCC

TCS

MIN

PRO

MAX

MIN

PRO

MAX

39.3 131.0 393.1 25 1.6 5.2 15.7 1590 270 2500 0.6 0.1

70.7 235.5 706.6 25 2.8 9.4 28.3 1700 540 2500 0.7 0.2

99.8 332.7 998.2 25 4 13.3 39.9 1840 770 2500 0.7 0.3

4.7 15.6 46.8 1.5 3.1 10.4 31.2 197 30 21 9.4 1.4

14.6 48.6 145.9 1.5 9.7 32.4 97.3 337 110 21 16.1 5.2

34.8 116 348 1.5 23.2 77.3 232.0 650 270 21 31 12.9

3.4.2. PECs for freshwater and wastewater systems Under the three likely emission scenarios considered, the estimated PECs for TCS and TCC were  1 000 ng/L in freshwater, in all three dilution factors (DFs) (1, 3 and 10), however, TCC concentrations were > 1 000 ng/L in the influent (Table 5). According to Keller et al. (2014), the proposed dilution factors for South Africa are between 10 and 40. However, in GP most part of the year have low or no precipitation, and therefore, a large number of river systems are seasonal. As a result, very low dilution factors are possible and exacerbated often by prolonged drought periods. As such, in this model, lower DFs of 1 and 3 were considered to take into account these environmental conditions. Moreover, during certain months, DF can be < 1 due to high evaporation rates, and increased water use from the river systems for irrigation and other purposes (Musee, 2011). Estimated PECs for TCS in the influent and effluent were very low compared to MECs results of Amdany et al. (2014). The high MECs of TCS were based on specific sampling spot grabs, and unlikely to be representative of entire GP. In addition, due to complexity of the model input data defined by both spatial and temporal features made it implausible to make an adequate comparison of PECs and MECs. However, considering PECs derived herein for GP, and others reported elsewhere (Table S1) for TCS, showed in certain instances the values were comparable. For example, comparing TCS results for GP (4.7e348 ng/L) (Table 5) with Zhao et al. (2013) findings in surface water (0.9e478 ng/L) indicate both are in the same order of magnitude since China also experiences dry months (Tamura et al., 2013) as is the case in South Africa (defined by DFs of 1 and 3 in this model). Similar comparisons on MECs and PECs reported herein within the same range were observed for TCS in influent and effluent as reported by Lindstrom et al. (2002) for Switzerland. 3.5. Environmental risk assessment Using PNECs and PECs results reported in sections 3.3 and 3.4, respectively, RQ calculations were performed to estimate TCS and TCC risks to the organisms in freshwater, wastewater, and soils. Two approaches were followed to calculate RQ values. The first approach entailed calculating RQ values where the estimates were based on the minimum toxicity value for the most sensitive organism in a given compartment, and therefore, the PNEC value was calculated using an appropriate AF as prescribed by ECB (2003). Following this approach, and using chronic NOECs of D magna (for TCC as 0.25 mg/L) and algae (for TCS as 0.015 mg/L), and AF of 10 as per ECB (2003) categorization system, results obtained are shown in Table 5, where different DFs were considered in the case of freshwater systems. Although lower quantities of TCS compared to

TCC were released into the environment (Table 3), however, since TCS is more toxic it yielded much higher RQs (Table 5). Results of higher RQs for TCS compared to TCC derived in this study showed similarity to findings of Brausch and Rand (2011) (RQTCS ¼ 19.17; RQTCC ¼ 10.96) and Tamura et al. (2013) (RQTCS > 10; RQTCC z 10 based on algae results) calculated using PNEC values based on chronic toxicity, and MECs. Using the two studies that have reported MECs for TCC (Lehutso et al., 2017) and TCS (Amdany et al., 2014) in Gauteng Province, a compassion of PECs and MECs was done as well as the RQs results derived based on each data type (Table S11). Generally, PECs for both TCC and TCS showed good agreement with the minimum- to middle-values of MECs data except for TCC in freshwater where PECs were higher than MECs. The lower PECs values were attributed to two reasons. First, the collected brand product data used as model input were deemed not to account for all sources of TCC and TCS to the environments considered, and therefore, yielded lower predicted concentrations. And secondly, since MFA approach yields an estimate of emitted average concentration of a given contaminant into the environmental compartment of focus; partly may account why the PECs values were lower than the MECs ones. Results of TCS in freshwater (based on PECs and MECs), influent and effluent revealed RQs > 1; hence indicating TCS poses elevated risks to the environment. The high RQs in freshwater derived using MECs indicate that despite WWTPs having had high removal efficiency for TCS of 67e86% (Amdany et al., 2014; Lehutso et al., 2017) in addition to dilution effect in the river system, TCS risks remained high especially to the algae. In addition, RQs for TCS based on PECs estimates in this study were lower than those reported by Thomaidi et al. (2015) with RQ value of 4914 calculated using MECs data; however, MECs based derived RQs values were in good agreement (587e5813) (Table S11). Results also indicate that during low precipitation season, most flows are wastewater from WWTPs (DF ¼ 1), with consequent risks likely to be significantly high. At DF ¼ 40, TCC risks can only occur under maximum release scenario for TCC (RQTCCMax ¼ 1) whereas for TCS, this is plausible under probable (RQTCSPro ¼ 2.43) and maximum (RQTCSMax ¼ 5.8) release scenarios (data not in Table 5). Nonetheless, there are potentially likely risks from the rest of release scenarios for TCS, and this was also observed in RQ results calculated using MECs values (Table S11) where the later were higher in range of 25e196 times than PECs derived ones. However, MEC- and PEC-based calculated RQs values for TCC suggest low or no risk in the influent and effluent; but only in freshwater (Table S11). However, there is need for caution since as mentioned earlier, results presented in this study did not take into account potential risks associated with metabolites of TCS or TCC likely to be formed during transformation processes (e.g. adsorption into

834

N. Musee / Environmental Pollution 242 (2018) 827e838

sediments, and phototransformation, chlorination, etc.) in freshwater and wastewater systems, or due to mixtures of TCS and TCC at various ratios, or together with their formed respective metabolites. For example, TCS can transform in the aquatic systems into methyl-triclosan e a chemical with similar properties to TCS but with very high bioaccumulation as well as biopersistence potential and aquatic toxicity, but poorly monitored in the environment. In addition, caution is essential in making comparison of RQ values derived using PECs- and MECs-based inputs as the former is for all plants whereas the latter is only for specific sampling locations. Overall, risk results presented herein, or reported elsewhere in previous investigation only provide partial risks of TCS and TCC in the aquatic system, and hence this merits further research in order to bridge those data gaps. In terrestrial systems, TCS and TCC under all release scenarios had RQs <1, and therefore, posed no risk to soil organisms based on

estimated PECs. However, since the highest mass of TCS and TCC were estimated to be in sludge, this implies that as the use and release of these chemicals into the environment continues to increase, the PECs may likely exceed the PNECs, and could pose risk to the soil organisms. The second approach entailed calculation of RQs for organisms in different compartments using acute and chronic toxicity collected in this study (only PECs-based results are presented herein as the MECs-based one are too few to be representative of the region in focus). Ecotoxicological data even for the same species, or on the same trophic level showed marked differences as evident in Figs. 1 and 2. These differences are associated with factors like: intra-species differences in addition to life-history stage, non-uniformity of exposure media chemistry (abiotic factors), presence or absence of other environmental pollutants, testing protocols, among others, thus; each toxicity data was taken to

Fig. 3. RQs calculated using (a) acute and (b) chronic toxicity of TCS to different organisms.

N. Musee / Environmental Pollution 242 (2018) 827e838

represent a certain level of ecological risk in the environment. Results for the TCS and TCC, respectively, shown in Fig. 3 and Fig. 4 summarises estimated RQs under the following conditions: DF ¼ 1, PEC for the probable release scenario, and AFs for acute and chronic toxicities as 1000 and 10, respectively. The results in Figs. 3 and 4 generally suggest, on the basis of acute and chronic toxicities, that both chemicals pose highly variant likely risks to organisms in different compartments with RQs > 1. Also, the results provide valuable insights that raise questions on the appropriateness of current approach of selecting the most sensitive organism based on a single toxicity value as

835

performed in the first approach; which in turn may lead to undesirable scenarios where other organisms are compromised. For example, choice of least toxicity assumes no impacts to organisms with higher toxicity; however, results in Figs. 3 and 4 show wide variability of adverse effects even of the same species, and at different life-history attested with numerous RQs > 1. 4. Environmental implications and research directions Using MFA analysis, the likely risks of TCS and TCC in the aquatic systems were estimated; however, lack of MECs data in South Africa

Fig. 4. RQs calculated using (a) acute and (b) chronic toxicity of TCC to different organisms.

836

N. Musee / Environmental Pollution 242 (2018) 827e838

was a major issue. As a result, it was not feasible to ascertain the accuracy of the estimated PECs based on several proposed ranking criteria frameworks (Coetsier et al., 2009; Verlicchi et al., 2014). Despite this shortcoming, findings presented herein have practicaland policy-related implications. First, due to limited studies on flows of TCC and TCS in many developing countries such as those in Africa, Asia and South America (Ebele et al., 2017) implies that the computer modelled RQ values can be useful in predicting their potential risks given these chemicals increasing presence in the environment. Secondly, scarcity of chronic data in organisms usually used for risk assessment in different environmental compartments render it impossible to estimate TCS and TCC long-term impacts. This means, TCS and TCC can be considered for inclusion in national monitoring programs in different environmental systems including sediments and pore-water as well as studies on their chronic toxicity to different taxa based on the prediction data sets. Such screening models can allow the identification of hot spots, and ultimately aid to develop appropriate and corrective strategies for specific situations and locales. PECs results in this study are valuable for rapid screening of TCS and TCC risks, but not as a substitute for field monitoring data. Until now, divergent and fragmented data on the production and use of TCS and TCC (Table S9) limits our ability to accurately quantify likely loads to be released into different environmental systems, and further compounded by model-parametrization challenges related to exposure modelling. Moreover, lack of market penetration data for products containing TCS and TCC makes it difficult to compare and refine our model results. This challenge is not unique to South Africa, but rather is global. Since most companies that manufacture and use these chemicals in numerous products have a global presence; new and innovative policies and mechanisms at national level should be implemented to acquire market, and toxicity data to bridge such gaps in order to support risk assessment of ECs as experimental data for such chemicals is often missing or highly scarce. Fourthly, there is necessity to link likely implications of both TCS and TCC or their mixtures to human health through the food chain. However, lack of bioaccumulation data for TCS and TCC in edible plants makes this task implausible, and therefore, merits further attention. As both chemicals are used in PCPs, and concurrently released into the environment, TCS and TCC are present in the environment as a mixture, and as individual contaminants. This raises the need for future work to consider interactions between TCS and TCS as a mixture, and the impact on the aquatic organisms where effects may be antagonistic, additive, or synergistic such that individual chemicals effects can either be reduced, or enhanced. Need for mixture studies is also critical because in many previous studies on chemicals toxic effects were observed in concentrations deemed to have minimal or no effect for individual chemicals. 5. Conclusions The modelling exposure framework presented herein based on the RQ approach was performed to estimate the potential risks of TCS and TCC to freshwater, wastewater, and soil. Although, the approach is predictive, it is a valuable tool for estimating the PECs of chemicals in the aquatic environment especially where MECs data is missing. Moreover, given the complexity of environmental systems, and the spatial level targeted in this study, implies the estimated PECs are dependent on the quality and available data. Hence, depending on market penetration, purchasing power, dilution factors, among other variables, RQs were >1 in freshwater ranging from 2 to 232 for both TCS and CC (PECs values); whereas for TCS > 1 in influent and effluent wastewater, but < 1 for TCC. Hence, results

show TCS can pose risks in wastewater and freshwater, whereas TCC poses risks to freshwater but none in wastewater. Using the available ecotoxicological data, both chemicals posed no risk (RQ « 1) to the terrestrial environments in GP, South Africa, although the ecotoxicological data in this compartment is scarce and defined by numerous data gaps. Although in practice the least toxicity is generally selected, and used to screen the potential risks to a given organism in a certain environmental compartment, use of all toxicity values nonetheless provide insights into effects likely to occur even at different life-history stages of the same organism which often are not taken into account; with long-term detrimental impacts. Furthermore, studies on the bioaccumulation of TCS and TCC in edible plants remain a matter of increasing urgency as currently it is not possible to determine these chemicals threat to the human health via the food chain. These antibacterial agents mixture toxicity also need to be determined in future to aid more accurate risk assessment in the environmental systems. Overall, this study's findings are envisaged to find wide application and relevance particularly to countries with similar sludge and wastewater management strategies as is the current practice in South Africa for PCPs as well as in the formulation of research questions and policy interventions. Acknowledgement The author thanks Jaco Erasmus and Karen Viljoen for their assistance in data collection of products market data. The author acknowledges the financial support from the University of Pretoria (UP) (Grant No.: A0Y229), and the Water Research Commission (WRC) (K5/2509/1), South Africa. The author thanks Prof Philip Crause for critical comments and for kindly editing an earlier version of this manuscript. Comments and suggestions of three anonymous reviews are highly acknowledged since they aided to improve the quality of the article. Any opinions, findings, conclusions, or recommendations expressed in this article are solely those of the author and do not necessarily reflect the views of the University of Pretoria and Water Research Commission. Appendix A. Supplementary data Supplementary data related to this article can be found at https://doi.org/10.1016/j.envpol.2018.06.106. References Al-Rajab, A., Sabourin, L., Lapen, D., Topp, E., 2015. Dissipation of triclosan, triclocarban, carbamazepine and naproxen in agricultural soil following surface or sub-surface application of dewatered municipal biosolids. Sci. Total Environ. 512e513, 480e488. Amdany, R., Chimuka, L., Cukrowska, E., 2014. Determination of naproxen, ibuprofen and triclosan in wastewater using the polar organic chemical integrative sampler (POCIS): a laboratory calibration and field application. WaterSA 40, 407e414. APUA, 2011. Triclosan. White Paper Prepared by the Alliance for the Prudent Use of Antibiotics (APUA) Available on the Link. http://emerald.tufts.edu/med/apua/ news/leadership_award.shtml. (Accessed 11 March 2017). Bedoux, G., Roig, B., Thomas, O., Dupont, V., Le Bot, B., 2012. Occurrence and toxicity of antimicrobial triclosan and by-products in the environment. Environ.l Sci. Pollution. Res. 19, 1044e1065. Borchgrevink, C.P., Cha, J., Kim, S., 2013. Hand washing practices in a college town environment. J. Environ. Health 75, 18e24. Brausch, J.M., Rand, G.M., 2011. A review of personal care products in the aquatic environment: environmental concentrations and toxicity. Chemosphere 82, 1518e1532. Brunner, P.H., Rechberger, H., 2004. Introduction. In: Practical Handbook of Material Flow Analysis. CRC Press, Boca Raton, FL, USA. Carey, D., McNamara, P., 2015. The impact of triclosan on the spread of antibiotic resistance in the environment. Front. Microbiol. 5, 1e11. Carey, D., Zitomer, D., Hristova, K., Kappell, A., McNamara, P., 2016. Triclocarban influences antibiotic resistance and alters anaerobic digester microbial community structure. Environ. Sci. Technol. 50, 126e134.

N. Musee / Environmental Pollution 242 (2018) 827e838 Cha, J., Cupples, A., 2009. Detection of the antimicrobials triclocarban and triclosan in agricultural soils following land application of municipal biosolids. Water Res. 43, 2522e2530. Chalew, T.E., Halden, R.U., 2009. Environmental exposure of aquatic and terrestrial biota to triclosan and triclocarban. J. Am. Water Works Assoc. 45 (1), 4e13. Chen, X.J., Nielsen, J.L., Furgal, K., Liu, Y.L., Lolas, I.B., Bester, K., 2011. Biodegradation of triclosan and formation of methyl-triclosan in activated sludge under aerobic conditions. Chemosphere 84, 452e456. Clarke, R., Healy, M.G., Fenton, O., Cummins, E., 2016. A quantitative risk ranking model to evaluate emerging organic contaminants in biosolid amended land and potential transport to drinking water. Human Ecol. Risk Assess: Int. J. 22, 958e990. Coetsier, C.M., Spinelli, S., Lin, L., Roig, B., Touraud, E., 2009. Discharge of pharmaceutical products (PPs) through a conventional biological sewage treatment plant: MECs vs PECs? Environ. Int. 35, 787e792. Coogan, M., La Point, T., 2008. Snail bioaccumulation of triclocarban, triclosan, and methyltriclosan in a north Texas, USA, stream affected by wastewater treatment plant runoff. Environ. Toxicol. Chem. 27, 1788. Coogan, M., Edziyie, R., La Point, T., Venables, B., 2007. Algal bioaccumulation of triclocarban, triclosan, and methyl-triclosan in a North Texas wastewater treatment plant receiving stream. Chemosphere 67, 1911e1918. Daughton, C., Ternes, T., 1999. Pharmaceuticals and personal care products in the environment: agents of subtle change? Environ. Health Perspect. 107, 907e938. Dinwiddie, M.T., Terry, P.D., Chen, J., 2014. Recent evidence regarding triclosan and cancer risk. Int. J. Environ. Res. Publ. Health 11, 2209e2217. Ebele, A.J., Abdallah, M.A.-E., Harrad, S., 2017. Pharmaceuticals and personal care products (PPCPs) in the freshwater aquatic environment. Emerg. Contam 3, 1e16. ECB (European Chemical Bureau), 2003. Institute for Health and Consumer Protection, European Commission: Technical Guidance Document on Risk Assessment in Support of Commission Directive 93/67/EEC for New Notified Substances. Available at: https://echa.europa.eu/documents/10162/16960216/ tgdpart2_2ed_en.pdf. (Accessed 30 March 2017). Environment Canada, 2006. Guidance for Conducting Ecological Assessments under CEPA, 1999: Science Resource Technical Series, Technical Guidance Module: Sludge Amendment. Working Document. Environment Canada, Ecological Assessment Division, Gatineau (QC). European Commission, 2005. Opinion on Triclocarban for Uses Other than as a Preservative, pp. 1e39. Fu, Q., Sanganyado, E., Ye, Q., Gan, J., 2016. Meta-analysis of biosolid effects on persistence of triclosan and triclocarban in soil. Environ. Pollut. 210, 137e144. Geer, L.A., Pycke, B.F.G., Waxenbaum, J., Sherer, D.M., Abulafia, O., Halden, R.U., 2016. Association of birth outcomes with fetal exposure toparabens, triclosan and triclocarban in an immigrant population in Brooklyn, New York. J. Hazard Mater. 323, 177e183. Gottschall, N., Topp, E., Metcalfe, C., Edwards, M., Payne, M., Kleywegt, S., Russell, P., Lapen, D., 2012. Pharmaceutical and personal care products in groundwater, subsurface drainage, soil, and wheat grain, following a high single application of municipal biosolids to a field. Chemosphere 87, 194e203. Government of Canada, 2016. Assessment Report: Triclosan. Environment and Climate Change Canada, Health Canada. Guo, J., Iwata, H., 2017. Risk assessment of triclosan in the global environment using a probabilistic approach. Ecotoxicol. Environ. Saf. 143, 111e119. Halden, R., Paull, D., 2005. Co-occurrence of triclocarban and triclosan in U.S. water resources. Environ. Sci. Technol. 39, 1420e1426. Halden, R.U., 2014. On the need and speed of regulating triclosan and triclocarban in the United States. Environ. Sci. Technol. 48, 3603e3611. Healy, M., Fenton, O., Cormican, M., Peyton, D., Ordsmith, N., Kimber, K., Morrison, L., 2017. Antimicrobial compounds (triclosan and triclocarban) in sewage sludges, and their presence in runoff following land application. Ecotoxicol. Environ. Saf. 142, 448e453. Heidler, J., Sapkota, A., Halden, R.U., 2006. Partitioning, persistence, and accumulation in digested sludge of the topical antiseptic triclocarban during wastewater treatment. Environ. Sci. Technol. 40, 3634e3639. Higgins, C.P., Paesani, Z.J., Chalew, T.E.A., Halden, R.U., Hundal, L.S., 2011. Persistence of triclocarban and triclosan in soils after land application of biosolids and bioaccumulation in Eisenia Foetida. Environ. Toxicol. Chem. 30, 556e563. Hua, W., Bennett, E.R., Letcher, R.J., 2005. Triclosan in waste and surface waters from the upper Detroit River by liquid chromatography-electrospray-tandem quadrupole mass spectrometry. Environ. Int. 31, 621e630. Huang, C., Ma, H., Yu, C., 2014. Substance flow analysis and assessment of environmental exposure potential for triclosan in mainland China. Sci. Total Environ. 499, 265e275. Jenkin, S.R., Armstron, C.W., Monti, M.M., 2007. Health Effects of Biosolids Applied to Land: Available Scientific Evidence. Virginia Department of Health. Available at: http://www.vdh.virginia.gov/epidemiology/DEE/documents/biosolids.pdf. (Accessed 20 September 2017). Keller, V.D.J., Williams, R.J., Lofthouse, C., Johnson, A.C., 2014. Worldwide estimation of river concentrations of any chemical originating from sewage-treatment plants using dilution factors. Environ. Toxicol. Chem. 33, 447e452. Kim, J.Y., Yi, B.R., Go, R.E., Hwang, K.A., Nam, K.H., Choi, K.C., 2014. Methoxychlor and triclosan stimulates ovarian cancer growth by regulating cell cycle- and apoptosis-related genes via an estrogen receptor-dependent pathway. Environ. Toxicol. Pharmacol. 37, 1264e1274. Lankester, J., Patel, C., Cullen, M.R., Ley, C., Parsonnet, J., 2013. Urinary triclosan is

837

associated with elevated body mass index in NHANES. PLoS One 8, e80057. Lee, H.R., Hwang, K.A., Nam, K.H., Kim, H.C., Choi, K.C., 2014. Progression of breast cancer cells was enhanced by endocrine-disrupting chemicals, triclosan and octylphenol, via an estrogen receptor-dependent signalling pathway in cellular and mouse xenograft models. Chem. Res. Toxicol. 27, 834e842. Lemly, A., 1996. Evaluation of the hazard quotient method for risk assessment of Selenium. Ecotoxicol. Environ. Saf. 35, 156e162. Lehutso, R.F., Daso, A.P., Okonkwo, O., 2017. Occurrence and environmental levels of triclosan and triclocarban in selected wastewater treatment plants in Gauteng Province, South Africa. Emerging Contam. 3, 107e114. Levy, S., 2002. Active efflux, a common mechanism for biocide and antibiotic resistance. J. Appl. Microbiol. 92, 65Se71S. Lindstrom, A., Buerge, I.J., Poiger, T., Bergqvist, P.A., Muller, M.D., Buser, H.R., 2002. Occurrence and environmental behavior of the bactericide triclosan and its methyl derivative in surface waters and in wastewater. Environ. Sci. Technol. 36, 2322e2329. Lozano, N., Rice, C.P., Ramirez, M., Torrents, A., 2013. Fate of triclocarban, triclosan, and methyltriclosan during wastewater and biosolids treatment processes. Water Res. 47, 4519e4527. Lv, M., Sun, Q., Xu, H., Lin, L., Chen, M., Yu, C.P., 2014. Occurrence and fate of triclosan and triclocarban in a subtropical river and its estuary. Mar. Pollut. Bull. 88, 383e388. Madikizela, L., Muthwa, S., Chimuka, L., 2014. Determination of triclosan and ketoprofen in river water and wastewater by solid phase extraction and high performance liquid chromatography. SA J. of Chem. 67, 143e150. Marx, C.J., Alexander, W.V., Johannes, W.G., Steinbach-Kanes, S., 2004. A Technical and Financial Review of Sewage Sludge Treatment Technologies. Water Research Commission Report No. 1240/1/04. Water Research Commission, Pretoria. Miller, T.R., Heidler, J., Chillrud, S.N., Delaquil, A., Ritchie, J.C., Mihalic, J.N., Bopp, R., Halden, R.U., 2008. Fate of triclosan and evidence for reductive dechlorination of triclocarban in estuarine sediments. Environ. Sci. Technol. 42, 4570e4576. Musee, N., 2011. Simulated environmental risk estimation of engineered nanomaterials: a case of cosmetics in Johannesburg City. Hum. Exp. Toxicol. 30 (9), 1181e1195. Musee, N., 2017. A model for screening and prioritizing consumer nanoproduct risks: a case study from South Africa. Environ. Int. 100, 121e131. Narumiya, M., Nakada, N., Yamashita, N., Tanaka, H., 2013. Phase distribution and removal of pharmaceuticals and personal care products during anaerobic sludge digestion. J. Hazard Mater. 260, 305e312. Oggioni, M.R., Furi, L., Coelho, J.R., Maillard, J.-Y., Martínez, J.L., 2013. Recent advances in the potential interconnection between antimicrobial resistance to biocides and antibiotics. Expert Rev. Anti Infect. Ther. 11, 363e366. Perencevich, E.N., Wong, M.T., Harris, A.D., 2001. National and regional assessment of the antibacterial soap market: a step toward determining the impact of prevalent antibacterial soaps. Am. J. Infect. Contr. 29, 281e283. Petrie, B., Barden, R., Kasprzyk-Hordern, B., 2014. A review on emerging contaminants in wastewaters and the environment: current knowledge, understudied areas and recommendations for future monitoring. Water Res. 72, 3e27. Pintado-Herrera, M.G., Gonzalez-Mazo, E., Lara-Martin, P.A., 2014. Determining the distribution of triclosan and methyl triclosan in estuarine settings. Chemosphere 95, 478e485. Pollock, T., Tang, B., deCatanzaro, D., 2014. Triclosan exacerbates the presence of 14C-bisphenol A in tissues of female and male mice. Toxicol. Appl. Pharmacol. 278, 116e123. QYResearch, 2015. Global Triclosan Industry 2016 Market Research Report. QYR Chemical & Material Research Centre. Sabaliunas, D., Webb, S.F., Hauk, A., Jacob, M., Eckhoff, W.S., 2003. Environmental fate of triclosan in the river Aire basin, UK. Water Res. 37, 3145e3154. S anchez-Brunete, C., Miguel, E., Albero, B., Tadeo, J.L., 2010. Determination of triclosan and methyl triclosan in environmental solid samples by matrix solidphase dispersion and gas chromatographyemass spectrometry. J. Separ. Sci. 33, 2768e2775. Sapkota, A., Heidler, J., Halden, R., 2007. Detection of triclocarban and two cocontaminating chlorocarbanilides in US aquatic environments using isotope dilution liquid chromatography tandem mass spectrometry. Environ. Res. 103, 21e29. SCCP, 2009. Opinion on Triclosan (COLIPA No. P32). European Commission, Health & Consumer Protection Directorate-General. Scientific Committee on Consumer Products. http://ec.europa.eu/health/ph_risk/committees/04_sccp/docs/sccp_o_ 166.pdf. (Accessed 11 March 2017). Sherburne, J., Anaya, A., Fernie, K., Forbey, J., Furlong, E., Kolpin, D., Dufty, A., Kinney, C., 2016. Occurrence of triclocarban and triclosan in an agro-ecosystem following application of biosolids. Environ. Sci. Technol. 50, 13206e13214. Singer, H., Muller, S., Tixier, C., Pillonel, L., 2002. Triclosan: occurrence and fate of a widely used biocide in the aquatic environment: field measurements in wastewater treatment plants, surface waters, and lake sediments. Environ. Sci. Technol. 36, 4998e5004. Smith, S., 2013. U.S. Disinfectant & Antimicrobial Chemicals Market. PRWeb. http:// www.prweb.com/releases/2013/9/prweb11150188.htm. (Accessed 28 February 2014). Sun, Q., Lv, M., Hu, A., Yang, X., Yu, C., 2014. Seasonal variation in the occurrence and removal of pharmaceuticals and personal care products in a wastewater treatment plant in Xiamen, China. J. Hazard Mater. 277, 69e75. Synman, H.G., Herselman, J.E., 2006. Guidelines for the utilization and disposal of

838

N. Musee / Environmental Pollution 242 (2018) 827e838

wastewater sludge: Volume 2, Requirements for the agricultural use of wastewater sludge. Water Res. Report, No. TT 262/06. Tamura, I., Kagota, K.I., Yasuda, Y., Yoneda, S., Morita, J., Nakada, N., Kameda, Kimura, Y., Tatarazako, K.N., Yamamoto, H., 2013. Ecotoxicity and screening level ecotoxicological risk assessment of five antimicrobial agents: triclosan, triclocarban, resorcinol, phenoxyethanol and p-thymo.l. J. Appl. Toxicol. 33, 1222e1229. Thomaidi, V.S., Stasinakis, A.S., Borova, V.L., Thomaidis, N.S., 2015. Is there a risk for the aquatic environment due to the existence of emerging organic contaminants in treated domestic wastewater? Greece as a case-study. J. Hazard Mater. 283, 740e747. TCC Consortium, 2002. High Production Volume Chemical Challenge Program Data Availability and Screening Level Assessment for Triclocarban, CAS#: 101-20-2, pp. 1e38. Report No. 201-14186A. www.epa.gov/chemrtk/tricloca/c14186cv.pdf. (Accessed 11 March 2017). TSCA, 2003. Chemical substances inventory. U.S. Environmental protection agency, Washington, D.C. Am. J. Infect. Contr. 29, 281e283. USEPA, 2011. U.S. Environmental Protection Agency (EPA). Exposure Factors Handbook: Edition. National Center for Environmental Assessment, Washington, DC; EPA/600/R-09/052F. Available from:. the National Technical Information Service, Springfield, VA. online at: http://www.epa.gov/ncea/efh. (Accessed 5 August 2017). Verlicchi, P., Zambello, E., 2015. Pharmaceuticals and personal care products in untreated and treated sewage sludge: occurrence and environmental risk in the case of application on soil d a critical review. Sci. Total Environ. 538, 750e767.  , D., 2014. Comparison of Verlicchi, P., AlAukidy, M., Jelic, A., Petrovi c, M., Barcelo measured and predicted concentrations of selected pharmaceuticals in wastewater and surfacewater: a case study of a catchment area in the PoValley (Italy). Sci. Total Environ. 470e471, 844e854. Winitthana, T., Lawanprasert, S., Chanvorachote, P., 2014. Triclosan potentiates

epithelial-to-mesenchymal transition in anoikis-resistant human lung cancer cells. PLoS One 9, e110851. Witorsch, R., Thomas, J., 2010. Personal care products and endocrine disruption: a critical review of the literature. Crit. Rev. Toxicol. 40, 1e30. Yazdankhah, S., Scheie, A., Høiby, E., Lunestad, B., Heir, E., Fotland, T., Naterstad, K., Kruse, H., 2006. Triclosan and antimicrobial resistance in bacteria: an overview. Microb. Drug Resist. 12, 83e90. Ying, G.G., Kookana, R.S., 2007. Triclosan in effluents and biosolids from Australian wastewater treatment plants. Environ. Int. 33, 199e205. Ying, G.G., Yu, X.Y., Kookana, R.S., 2007. Biological degradation of triclocarban and triclosan in a soil under aerobic and anaerobic conditions and comparison with environmental fate modelling. Environ. Pollut. 150, 300e305. Young, S., 2013. FDA Examining Antibacterial Soaps, Body Washes. CNN Health. http://www.cnn.com/2013/12/16/health/fdaantibacterial/. (Accessed 16 February 2017). Yu, C.P., Chu, K.H., 2009. Occurrence of pharmaceuticals and personal care products along the West Prong Little Pigeon River in east Tennessee, USA. Chemosphere 75, 1281e1286. Zhang, Y., Leu, Y., Aitken, R., Riediker, M., 2015. Inventory of engineered nanoparticle-containing consumer products available in the Singapore retail market and likelihood of release into the aquatic environment. Int. J. Environ. Res. Publ. Health 12, 8717e8743. Zhao, J.L., Zhang, Q.Q., Chen, F., Wang, L., Ying, G.G., Liu, Y.S., Yang, B., Zhou, L.J., Liu, S., Su, H.C., Zhang, R.Q., 2013. Evaluation of triclosan and triclocarban at river basin scale using monitoring and modelling tools: implications for controlling of urban domestic sewage discharge. Water Res. 47, 395e405. Zhu, W., Zhang, H., Tong, C., Xie, C., Fan, G., Zhao, S., Yu, X., Tian, Y., Zhang, J., 2016. Environmental exposure to triclosan and semen quality. Int. J. Environ. Res. Publ. Health 13, E224.