1. Introduction
Pollution occurs globally in natural and engineered water systems, and contamination of drinking water sources challenges ecological safety and public health. According to a report by the United Nations Children’s Fund (UNICEF) and the World Health Organization (WHO), although 90% of the global population has access to a basic drinking water source, water contamination is widespread, and at least 2 billion people are using drinking water sources that are contaminated with feces [
1]. To protect aquatic environments and safeguard drinking water safety, the potential origins of pollutants in drinking water sources and their resultant contamination occurrences should be recognized [
2], [
3]. Among different water contaminants, harmful chemical constituents in raw water can be primarily classified in terms of their origins, such as naturally occurring chemicals, chemicals of industrial origin, chemicals due to human dwellings, and chemicals from agricultural activities [
2]. Both natural processes (e.g., climate change (a process that is natural in its progression, despite its anthropogenic sources), and natural disasters such as floods, drought, and wildfire) and human activities (e.g., urbanization, mining, domestic/farm/industrial wastewater discharge, and hydraulic fracturing) may affect the availability and quality of source water and even finished water at drinking water treatment plants (DWTPs) [
4], [
5], [
6].
Efforts have been made to develop numerous methods for identifying potential sources of water contamination on a watershed scale, such as inventory analysis of pollutant discharge, hydrological modeling, stable isotope tracing, and multivariate statistics [
7]. Several specific pollutants are viewed as contamination indicators of different pollution origins. For example, the presence of primidone and carbamazepine suggests possible wastewater-discharge-induced pollution [
8]. Caffeine is likely related to fecal and untreated wastewater contamination, while a high ratio of caffeine to carbamazepine may be indicative of a major proportion of raw or insufficiently treated sewage in surface water [
9], [
10]. Sucralose—an artificial sweetener that is extensively found in aqueous environments—is consistently absent in natural waters without wastewater impacts but is constantly detected in wastewater-impacted waters [
11]. Therefore, sucralose has been suggested to serve as a tracer for wastewater contamination [
11], [
12].
Disinfection byproducts (DBPs) are a family of secondary contaminants generated during water disinfection. They are formed by the unintended reactions of chemical disinfectants with certain water matrix constituents, known as DBP precursors (e.g., natural organic matter (NOM), anthropogenic contaminants, and halides) [
13], [
14]. Since the first discovery of DBPs in chlorinated drinking water in the 1970s, DBPs have gained a great deal of attention due to their potential adverse health effects [
15]. A variety of chemical constituents can serve as DBP precursors, with varied origins such as terrestrial or natural aquatic systems, atmospheric systems, wastewater treatment plants (WWTPs), industrial discharge, human dwellings, aquaculture, and agricultural activities [
16], [
17], [
18]. In general, water pollution is directly linked to human activities of urban, industrial, or agricultural origin, and a series of natural events have been found to lead to degradation in surface water quality [
4], [
19], [
20], [
21], [
22]. Thus, considerable interest has been directed to the impacts of natural processes and human activities on DBPs and DBP precursors in drinking water sources. For example, tropical storms can dramatically alter the quantity and quality of dissolved organic matter (DOM) in natural water bodies, including those serving as drinking water sources, and thus change the concentration and speciation of DBPs formed during subsequent disinfection at DWTPs [
23]. Inappropriate disposal and leakage of bromide- and iodide-containing wastewater from hydraulic fracturing can lead to an increase in these unwanted halides in raw water and further promote the production of DBPs, especially highly toxic brominated and iodinated DBPs [
24]. Furthermore, because disinfection typically acts as the last barrier in wastewater treatment, the discharge of WWTP effluent and treated farm and industrial wastewater can potentially introduce the produced DBPs to natural waters [
25], [
26].
Given that source water is a highly complex matrix, it is technically infeasible to identify and quantify all the water constituents. However, studies on DBPs and DBP precursors—where the latter are determined using DBP formation potential (FP) tests—in water sources may open up a new opportunity, due to a greater understanding of the contributions of different pollution origins to specific DBP precursors, the distinct DBP formation patterns from different precursors, and the theoretical absence of formed DBPs in water sources without disinfected wastewater impacts. Here, we hypothesize that DBPs and DBP precursors can potentially serve as alternative indicators to assess the pollution status of drinking water sources, identify the pollution origins, and evaluate changes in source water quality due to natural processes or human activities.
To this end, this study is organized as follows: ① Firstly, we review the origins of DBPs and DBP precursors in drinking water sources and the formation traits of DBPs from different precursors; ② we then summarize natural and human-induced impacts on DBP precursors and produced DBPs in source water; ③ we discuss the feasibility of using formed DBPs and DBP precursors as contamination indicators for drinking water sources; and, finally, ④ we recognize environmental insights and future research needs.
2. DBPs, DBP precursors, and their origins
Since the first discovery of trihalomethanes (THMs) in the mid-1970s, a series of DBPs have been detected in disinfected water all over the world (
Fig. 1 [
13], [
16], [
18], [
27], [
28], [
29], [
30], [
31], [
32], [
33], [
34], [
35], [
36], [
37], [
38], [
39], [
40], [
41], [
42], [
43], [
44], [
45], [
46], [
47], [
48], [
49], [
50], [
51], [
52], [
53]), including in drinking water, reclaimed wastewater, disinfected hospital wastewater, desalinated seawater, treated produced water from oil and gas extraction, discharge from industrial facilities, and farm wastewater [
25], [
37], [
54], [
55], [
56], [
57], [
58]. In general, common organic DBPs include carbonaceous DBPs (C-DBPs), such as haloacetaldehydes (HALs) and widely regulated THMs and haloacetic acids (HAAs); and nitrogenous DBPs (N-DBPs), such as nitrosamines (NAs), haloacetonitriles (HANs), haloacetamides (HAMs), and halonitromethanes (HNMs) [
37]. Among these DBPs, THMs, HAAs, HALs, HANs, HAMs, and HNMs share a similar CX
3R structure (where X = H, Cl, Br or I, and R = functional group) and thus have been classified as CX
3R-type DBPs (Fig. S1 in Appendix A) [
18], [
59]. With regard to DBP precursors, as shown in
Fig. 1, dominant investigations on DBP precursors in source water have gradually evolved from a focus on naturally occurring chemicals to investigations into anthropogenic contaminants, due to increasing concern over human-related activities on the watershed scale, where drinking water sources are situated.
2.1. Existing DBPs in aquatic environments and their origins
Studies investigating DBPs mostly focus on DBPs in disinfected drinking water, since people are directly exposed to drinking water DBPs, which have been reported to have potentially adverse effects on human health [
13]. Moreover, in order to mitigate biofouling and limit the discharge of invasive aquatic organisms and pathogens to receiving water bodies (e.g., surface water, groundwater, and marine water), wastewater is always subject to disinfection treatment, leading to the formation of DBPs [
58], [
60], [
61]. Since DBPs also pose risks to ecological safety [
62], [
63], [
64], the generation of DBPs in disinfected wastewater that is then discharged to the aquatic environment raises considerable concern. Interestingly, after estimating the inputs of DBPs to the marine environment from various industrial activities, based on the volumes of treated water and DBP concentrations in corresponding wastewater types, Grote et al. [
65] concluded that cooling water was the predominant anthropogenic source of DBPs, followed by desalination brine and ballast water.
Table 1 [
60], [
65], [
66], [
67], [
68] presents a summary of DBPs detected in disinfected wastewater, some of which may be important origins of existing DBPs in water sources.
DBP levels in seawater toilet effluents refer to average concentrations. DHALs: dihalogenated acetaldehydes; THALs: trihalogenated acetaldehydes; TBM: tribromomethane; DBAA: dibromoacetic acid; DBAN: dibromoacetonitrile; 2,4,6-TBP: 2,4,6-tribromophenol; N/A: not available.
2.2. DBP precursors in water sources and their origins
Fig. 2 provides a schematic of the origins of DBP precursors that are present in natural water sources, as well as general DBP formation pathways via the reactions of inorganic and organic DBP precursors with different disinfectants. In general, certain halide ions (i.e., bromide and iodide), various fractions of NOM or effluent organic matter (EfOM), and specific anthropogenic contaminants, which are derived from various origins (e.g., natural, domestic, industrial, and agricultural sources), can all serve as DBP precursors. Commonly used disinfectants include free chlorine, chloramines, chlorine dioxide, ozone, ultraviolet (UV) treatment, and disinfectant combinations. The formation of DBPs from various disinfection agents has been previously compared [
69], [
70]. In general, each disinfectant tends to form its own suite of DBPs, with some overlap of DBPs among different disinfectants [
13]. Free chlorine mainly produces THMs and HAAs, and chloramines could also lead to the formation of iodinated DBPs and NAs. Chlorine dioxide does not produce halogenated DBPs to any significant degree, except for chlorite and chlorate, and ozone disinfection is generally associated with issues concerning bromate and aldehydes. The sources, characteristics, and DBP formation patterns of representative DBP precursors in water sources have been systematically reviewed elsewhere [
16], [
71], [
72], [
73], [
74], [
75], and a brief overview is presented below.
2.2.1. Bromide and iodide
Bromide is generally present in fresh water at concentrations ranging from a few micrograms per liter to several milligrams per liter [
74], and iodide levels in source water are reported to range from 0.5 to 200 μg·L
−1 [
76]. Both natural processes (e.g., seawater intrusion and geologic mineral dissolution) and anthropogenic activities (e.g., seawater desalination, mining, discharge of treated wastewater and industrial effluents, and hydraulic fracturing) can increase the levels of bromide and iodide in aquatic environments [
73], [
77], [
78]. As for the impacts of bromide and iodide on DBP formation, the presence of bromide increases the total concentration of halogenated DBPs while shifting the prevailing DBP species to brominated DBPs (
Table 2) [
12], [
17], [
56], [
71], [
72], [
75], [
76], [
79], [
80], [
81], [
82], because bromine species (HOBr/OBr
-) are more active substitution agents than their chlorine counterparts [
72], [
79]. In addition to producing halogenated DBPs, bromide exhibits different effects on the formation of
N-nitrosodimethylamine (NDMA) in different water matrixes (e.g., model compound water, surface water, and wastewater) [
83]. Similarly, the disinfection or chemical oxidation of iodide-containing waters can result in the formation of iodinated DBPs, with the formation characteristics relying on the water types and disinfectant types [
73]. The formation of iodinated DBPs during the chemical oxidation of natural water generally follows this order: chloramines > chlorine > ozone [
84].
2.2.2. NOM
NOM in drinking water sources either originates externally from the water body (i.e., allochthonous NOM) or is generated
in situ (i.e., autochthonous NOM). Allochthonous NOM is commonly produced from the breakdown of terrestrial plants and thus contains a high aromatic content, while autochthonous NOM, which is mainly derived from phytoplankton, macrophyte, and bacterial sources, tends to have a low aromatic but high organic nitrogen content [
72], [
75], [
85].
As can be seen in
Fig. 1, during the 1980-1990, studies related to DBP precursors in source water principally focused on terrestrial-derived NOM (e.g., humic acid and fulvic acids) and microbial-derived NOM (e.g., amino acids and protein). Later, as water demand dramatically increased due to population growth and urbanization, alternative water sources were sought, and water bodies impaired by unwelcome factors, such as municipal wastewater discharge and algal blooms, were embraced [
81]. Consequently, increased attention was paid to wastewater EfOM and algal organic matter (AOM)—both of which principally have microbial origins with abundant dissolved organic nitrogen (DON)—as significant DBP precursors [
71], [
82].
Humic substances are the primary precursors of THMs and HAAs; they are also important in HNMs formation [
72]. Moreover, amino acids are the principal precursors of N-DBPs, such as HANs, HAMs, and HNMs [
80], [
81], and the chlorination of amino acids can generate aldehydes, which are further transformed into chloral hydrate (CH) with subsequent or concomitant chlorine substitution [
44]. Moreover, produced by exudation or cell lysis during algal eutrophication, AOM is rich in nitrogen with little aromatic structure and thus has substantial potential to produce N-DBPs [
49]. Both extracellular and intracellular AOM yield more HAAs than THMs, in which trichloronitromethane (TCNM) and dichloroacetonitrile (DCAN) are two commonly found N-DBPs formed from AOM [
71]. In particular, AOM can serve as a significant precursor of NAs and iodinated THMs during chloramination [
86], [
87].
2.2.3. Anthropogenic contaminants
Various anthropogenic contaminants, such as pharmaceuticals and personal care products (PPCPs), industrial chemicals and their byproducts, fungicides, and pesticides, can enter into drinking water sources through wastewater effluent discharge or agricultural runoff, before being transformed into DBPs during subsequent disinfection treatment in DWTPs [
12], [
17], [
53]. For example, the excretion of individuals using medical drugs and uncontrolled drug disposal (e.g., discarding drugs into toilets) increase the concentrations of pharmaceuticals in municipal wastewater, which may be poorly removed at WWTPs before being discharged into nearby natural water bodies. In a
de facto water reuse route, these residual pharmaceuticals can return to the raw water downstream of the DWTPs [
88]. Furthermore, poor management of veterinary drugs or feed additives in livestock manure at animal feeding operations can lead to the introduction of these substances to source water through agricultural runoff [
88].
Most investigations on DBP formation from anthropogenic contaminants focus on NAs and iodinated DBPs, both of which are highly carcinogenic or toxic [
12], [
76]. Levels of NAs and their precursors are negligible in pristine water sources with little pollution from conventional wastewater effluent, which is a dominant source of NAs and their chloramine-reactive and ozone-reactive precursors [
86]. The identified chloramine-reactive NA precursors mainly include the herbicide diuron [
48], the drug ranitidine [
52], and the drug methadone [
89]. Moreover, among the identified ozone-reactive NA precursors, dimethylsulfamide is a decomposition product of the fungicide tolylfluanid [
90], while 1,1,5,5-tetra-methylcarbohydrazide is typically derived from municipal wastewater containing industrial effluents [
91]. As for iodinated DBPs, iodinated X-ray contrast media, which are widely used at medical centers and cannot be completely removed by WWTPs, are more reactive with chlorine than chloramine to form iodinated DBPs, providing a different pathway for the formation of iodinated DBPs from iodide and NOM [
51], [
76].
3. Factors affecting DBPs and DBP precursors in drinking water sources
Water quality—particularly NOM characteristics—in drinking water sources can be greatly influenced by both natural events and anthropogenic activities.
Fig. 3 presents a series of examples of factors affecting DBPs and DBP precursors in drinking water sources. Special attention should be paid to climate change, which has profound impacts on source water quality, including basic physical and chemical parameters (e.g., temperature, pH, turbidity, and dissolved oxygen), DOM, nutrients, metals, micropollutants, and biological contaminants (e.g., pathogens and cyanobacteria) [
4], [
6]. Besides the constant increase in the global averaged annual surface temperature, a changing climate is strongly associated with the increasing frequency and intensity of climate- and weather-related natural disasters (e.g., droughts, floods, tropical storms, and wildfires) and unwelcome events (e.g., seawater intrusion and harmful algal blooms), which lead to changes in source water availability and quality [
92], [
93], [
94], [
95]. However, it should be noted that climate change is not the only cause of the aforementioned natural disasters or extreme events.
Anthropogenic activities can pollute water sources via both point and non-point routes. Wastewater discharges (e.g., domestic wastewater, industrial wastewater, farm wastewater, and hydraulic fracturing wastewater) receive considerable attention as point sources contributing to source water pollution. In contrast, atmospheric deposition and agriculture/urban runoff are perceived to be nonpoint sources [
96], [
97]. In general, nonpoint pollution sources contribute more contaminants to water sources than point pollution sources.
3.1. Natural factors
3.1.1. Droughts
Droughts can have multiple impacts on drinking water sources. First, the loss of water due to enhanced evaporation in a drought can concentrate the NOM, halides, and anthropogenic contaminants originally present in raw water, thereby increasing the formation of DBPs during drinking water production [
12], [
92]. Second, when droughts are compounded with a
de facto reuse scenario, the fraction of treated wastewater effluent in the receiving water bodies grows accordingly, raising the contributions of halides, as well as EfOM containing soluble microbial products (SMPs) and anthropogenic contaminants, to DBP formation [
50], [
98]. Third, the resultant increase in the nutrient contents in source water during a drought can trigger algal blooms, altering the composition of organic matter and increasing AOM-related DBP precursors [
93], [
99]. Therefore, autochthonous NOM most likely serves as the dominant origin of DBP precursors in eutrophic lakes or in source water experiencing low or intermittent rainfall events [
85].
3.1.2. Storms and floods
Extreme storms and floods can rapidly increase sediment, organic matter, turbidity, and pathogens in source water through urban drainage (e.g., combined sewer overflow), subsurface flow, surface runoff, and the resuspension of bottom sediments [
100], [
101], [
102]. It is also noteworthy that the adverse effects derived from short-term extreme flooding can last for several weeks, even after water levels have returned to base flow conditions.
The effects of storms and floods on DOM and DBP precursors depend on the characteristics of the catchments and the precipitation patterns of the storm events [
103], [
104]. More specifically, terrestrial-derived DOM with high levels of DBP FP was reported to enter a large watershed during storm events, with a higher DBP precursor content resulting from a more intense storm event [
105]. Rainfall events can alter the reactivity of organic matter in drinking water sources with an increase in THM FP, whereas the impact on HAA FP has been shown to be less noticeable [
106]. Decreases in the proportions of brominated THM FP and brominated HAA FP during and after rainfall were also observed, compared with pre-rainfall periods [
106]. In addition, storm drainage was found to contribute to NA precursors, with the FPs of total NAs approaching approximately 100 ng·L
−1 [
80]. Notably, intensive short-term events such as tropical storms, which mainly occur in coastal regions and are connected to extensive rainfall and flooding, have been reported to export DBP precursors into source water and thus significantly impact drinking water quality [
23], [
107].
3.1.3. Seawater intrusion
The occurrence of seawater intrusion is related to local sea level rise and/or prolonged dropdown in coastal groundwater levels as a result of climate change, groundwater over-extraction, land-use change, or sea-level fluctuations [
94]. Apart from elevated levels of total dissolved solids and chloride ions, the formation and speciation of DBPs (e.g., THMs, HAAs, and bromate) as a function of increased bromide and iodide resulting from seawater intrusion are of particular concern when estuary surface water or coastal groundwater is adopted as source water, especially in dry seasons [
108], [
109], [
110].
Bromide in the Pearl River Basin, which is the second largest river area in China, can reach 1975.64 μg·L
−1 in dry seasons as a result of seawater intrusion [
110]. Moreover, an increasing bromide level (up to 0.8 mg·L
−1) in source water exported from the Sacramento-San Joaquin River Delta in the United States was reported, which was related to saltwater intrusion from San Francisco Bay, leading to an increase in the levels of total THMs, a shift toward the formation of more brominated species during chlorination, and the increased production of bromate during ozonation [
111]. It should be noted that even small changes in saltwater intrusion can significantly impact the levels of total THMs and the extent of bromine incorporation in THMs [
108], [
109]. Higher frequencies of regulatory violations of halogenated organic DBPs, especially brominated THMs, and HAAs, in groundwater systems on coastal islands in Washington, DC, USA, have been reported. Plausible reasons include elevated bromide levels originating from seawater intrusion and the different composition and amount of DOM present in groundwater compared with surface water [
112].
3.1.4. Wildfires
In recent years, the amount, scale, and severity of wildfires have gradually increased across the world as a consequence of climate change, fuel load accumulation, extensive droughts, or increased human activities in forests [
95]. Post-fire rainstorms increase surface runoff and erode ash and soil from burned landscapes, thereby posing a great threat to forest watersheds. In addition, elevated nutrient levels accompanied by resultant post-fire algal blooms, which increase suspended sediment and metals, as well as changes in the quantity and composition of DOM in water sources, are reported to be associated with wildfires [
5], [
95]. Wildfires can destroy forest ecosystems within a short time, and changes in source water quality and treatability persist for several years or even decades [
113], [
114].
Changes in the quantity and composition of DOM in drinking water sources resulting from wildfire heating and subsequent rainfall events may influence DBP formation and speciation during water disinfection treatment. Postfire rainstorm samples show elevated DOM reactivity in the formation of HAN and TCNM, whereas their DOM reactivity in forming THM and HAA is similarly to non-storm samples [
115]. Wildfires also lead to more abundant NDMA precursors [
114]. Fire-altered terrestrial DOM is generally found to be more reactive in producing N-DBPs compared with humic acids and even algal and bacterial materials [
115], [
116], [
117]. Notably, increased HAN precursors in post-fire watersheds may also be associated with algal blooms due to increased nitrogen exported from burned catchments [
113]. In addition, the growth of bromide concentrations in raw water and an increase in the bromine incorporation factor in THMs and HAAs after wildfire have been reported [
114], [
116].
3.1.5. Algae blooms
Eutrophication and climate change can promote the proliferation and expansion of harmful algal blooms, which can influence source water quality, water treatment performance, and drinking water safety [
93]. The characterization of algae-derived NOM and the formation and speciation of DBPs have been extensively studied [
71], [
85]. Both AOM and its associated toxins can serve as DBP precursors during disinfection [
118], [
119]. As mentioned above, AOM exhibits a lower specific C-DBP FP (microgram of C-DBP per milligram of dissolved organic carbon (DOC)) than allochthonous NOM, whereas algal-impacted water is considered to have a higher potential for the formation of N-DBPs [
120], [
121]. Interestingly, although the overall DOC in water changes during the algal growth phase, the specific C-DBP FP is not a function of this phase [
121], [
122]. A full-scale study further confirmed that harmful algal bloom events in drinking water sources contribute to the enrichment of DBP precursors [
123]. Algal blooms can also arise from wildfires due to elevated nutrient (e.g., inorganic nitrogen and phosphorus) inputs into water sources [
124]. Moreover, postfire algal blooms have been found to increase the chlorine reactivity of wildfire-affected terrestrial DOM for DBP formation through biogeochemical processes [
125].
3.1.6. Sunlight
Heightened solar UV radiation, likely due to climate change and ozone depletion, can alter organic matter characteristics in fresh water systems via temperature variation and photolysis. The effect may vary depending on the origin, composition, and form of organic matter present in the source water [
126], [
127]. Although the overall DOC in natural water samples does not change significantly after irradiation, solar irradiation causes a shift in the DOC from a humic-like to a hydrophilic character and thus decreases the aromaticity of DOM [
128]. Consequently, the reactivity of DOC with respect to HAA formation tends to decline after irradiation, whereas the relative degree of bromine incorporation into THMs and HAAs shows an increasing trend [
128]. Furthermore, significant increases in specific HAN FP (microgram of HAN per milligram of DOC) and specific CH FP (microgram of CH per milligram of DOC) are reported when leachates from fresh litter are exposed to sunlight [
129]. Unlike NOM, the SMPs in EfOM are relatively amenable to solar photolysis, and the reactive intermediate species formed by the solar irradiation of SMPs favor the formation of THMs, CH, and TCNM [
130]. The effects of sunlight on the HAN FP and TCNM FP of wastewater effluent have also been investigated [
131], [
132]. In addition, solar irradiation can promote the release of DOM from particulate organic matter. The photo-released DOM from soil mainly contains humic-like components exhibiting a high THM FP, whereas the photo-released DOM from litter leaves principally comprises protein-like substances with a high HAA FP [
133], [
134]. Notably, sunlight photolysis is reported to mitigate NDMA formation when methadone serves as the NDMA precursor during chloramination [
135].
3.2. Human factors
3.2.1. Atmospheric deposition
Pollutant emissions to the atmosphere can be generally categorized into anthropogenic emissions (e.g., industrial stacks, municipal waste incinerators, agricultural activities, and vehicle exhaust), natural emissions (e.g., volcanic eruptions, windblown gases and particles from forest fires, windblown dust, soil particles, and sea spray), and re-emitted emissions [
136]. Primary carbonaceous particulate matter (PM) in the atmosphere predominantly originates from the intentional combustion of biomass and fossil fuel [
137]. Thus, atmospheric deposition is generally classified as an anthropogenic factor.
Atmospheric deposition results in significant impacts on surface water quality [
136], [
138]. Both dry (i.e., particulate and gaseous deposition) and wet depositions (i.e., precipitation such as rain and snow) can contribute to an increase in DBP precursors in surface water [
18], [
139], [
140]. Hou et al. [
18] demonstrated that the contribution of PM-derived DOM to the formation of some CX
3R-type DBPs (e.g., trichloroacetic acid (TCAA), DCAN, TCNM, and dichloroacetamide) during the chlorination of rainwater could not be ignored. Moreover, He et al. [
139] reported that atmospheric dry deposition contributed DOM and DBP precursors to surface water and that these were particularly related to the formation of trichloromethane (TCM), dichloroacetic acid, and TCAA. However, information from the literature on PM-derived DBP precursors remains limited, and further research is required.
3.2.2. Domestic effluent discharge
Municipal wastewater, which comprises greywaters (i.e., wastewater from showers, kitchen sinks, bathroom washbasins, and laundry) and blackwaters (i.e., urine and feces), is recognized as an important source of NAs and their precursors [
57], [
86], [
141]. Given that domestic wastewater is generally collected in sewer systems, treated in WWTPs, and discharged into adjacent natural water bodies [
142], and that EfOM can affect the quality of raw water in the context of its
de facto reuse [
50], [
82], this section mainly focuses on the impact of WWTP effluent on DBP precursors and DBP levels in receiving water bodies.
DBPs are formed during disinfection treatment, which is typically the last barrier in water treatment plants. Therefore, the effluents of WWTPs contain both DBP precursors and formed DBPs generated during wastewater disinfection [
14], [
143]. In a US survey, the median concentrations of THMs, HAAs, HANs, dihalogenated HALs, and trihalogenated HALs in chlorinated WWTP effluents were reported to be 57.0, 70.0, 16.0, 7.2, and 16.0 μg·L
−1, respectively [
66]. With regard to surface water, Heng et al. [
26] investigated the occurrence of volatile halogenated DBPs in an urban river supplied by reclaimed wastewater in Beijing, China. Their results indicated that TCM and CH were detected in all river water samples, with the highest levels reaching 485.1 and 30.8 μg·L
−1, respectively.
Wastewater-derived DBP precursors are also of concern [
50], [
144]. In general, EfOM is a combination of NOM, SMPs, and trace harmful chemicals such as anthropogenic micropollutants [
82]; it is characterized by a higher DON content, lower DOC/DON ratio, and higher N-DBP FP than NOM [
145]. Notably, certain PPCPs and NDMA FP in drinking water sources are suggested to serve as potential indicators of the impact of treated wastewater on watersheds [
8].
3.2.3. Farm wastewater or agriculture runoff
Chemicals used in crop production (e.g., fertilizers, herbicides, and pesticides), animal husbandry (e.g., veterinary drugs, feeds, and additives), and aquaculture (e.g., fertilizers, feeds, and additives), as well as constituents of the excretion of livestock and aquatic products, are likely to be present in farm wastewater effluent or agriculture runoff and may enter water sources [
2], [
25], [
146]. In addition to determining the yields of DBPs from specific DBP precursors, such as herbicides (e.g., diuron and bromacil) and veterinary antibiotics (e.g., minocycline) [
48], [
147], [
148], several studies have quantified the levels of different DBPs and DBP FPs in farm effluent or agriculture runoff. The contribution of farm wastewater or runoff to DBPs and DBP precursors generally depends on the feeds, pesticides, and pharmaceuticals used in the farms, as well as the treatment procedures applied to treat farm wastewater, which mainly include biological treatment, chemical oxidation, and disinfection.
Eckard et al. [
149] reported that the THM FP of irrigation runoff from agriculture systems was higher than that of runoff from a natural forested area and that of shrubland/grassland watersheds. However, compared with other domestic and industrial wastewaters, runoff related to crop production is a less significant contributor to NAs and their precursors [
25], [
86]. To be specific, runoff from animal husbandry can serve as a source of NDMA,
N-nitrosopyrrolidine (NPYR), and their chloramine- and ozone-reactive precursors [
86]. In fact, the concentrations of NDMA,
N-nitrosodiethylamine (NDEA), NPYR,
N-nitrosomorpholine (NMOR), and
N-nitrosodibutylamine (NDBA) in swine wastewater can reach hundreds to thousands of nanograms per liter [
25]. Notably, the contribution of livestock-breeding wastewater to NAs and NA precursors is largely dependent on the treatment process. For example, almost all NA precursors can be transformed into NAs when livestock-breeding wastewater is subject to complete anaerobic digestion, whereas residual NA precursors can exist in the case of incomplete anaerobic digestion [
150].
Aside from agricultural activities, wastewater generated from aquaculture can also contribute DBPs and their precursors. For example, DOC and bromide levels in shrimp farm effluents have been reported to be relatively high, and the THM FP of analyzed shrimp farm effluent can range from 810 to 3100 μg·L
−1 [
146]. Chen et al. [
25] reported that NDMA, NDEA, and NDBA were detected in an aquaculture pond, with NDBA being the most abundant NA species. The NA level and NA FP in fishery ponds were found to be lower than those in other wastewaters (e.g., WWTP effluent, industrial wastewater, and livestock-breeding wastewater); however,
N-nitrosomethylethylamine (NMEA) has rarely been detected in wastewaters other than fishery ponds [
150].
3.2.4. Industrial effluent discharge
The compositions of industrial effluents are quite complex. DBPs and their precursors may be present in different treated industrial wastewater effluents, and their levels and constituents largely depend on the nature of the industrial processes. Some examples can shed light on the role of industrial effluent discharge. First, the treated wastewater from shale drilling operations, waste from coal-fired power utilities, brine from oil and gas wells, and discharge from some specific industrial facilities are reported to contain elevated bromide levels, and their discharge can increase the bromide levels in receiving natural water bodies [
151], [
152]. The following two sections will further introduce bromide-related processes and their impacts on DBP formation in detail.
Second, industrial wastewaters, including textile printing and dyeing wastewater, electroplating wastewater, metal finishing wastewater, food processing wastewater, pulping wastewater, fine chemicals wastewater, mechanical industrial wastewater, cosmetics and personal care product manufacturing wastewater, and pharmaceutical manufacturing wastewater, have been found to contain NAs and their precursors. Notably, discharge from metal finishing and electronics-related facilities, as well as the effluent of textile printing and dyeing processes, contain high NDMA levels [
25], [
57], [
150]. Textile printing and dyeing wastewater also possess a high NDMA FP, while electroplating industrial wastewater has significantly high levels of NDEA and NDEA FP [
150].
Third, in addition to the aforementioned industrial wastewaters that contribute bromide or NAs and their precursors, effluent discharge from other industrial facilities can act as a source of DBPs and DBP precursors [
65], [
153]. Interestingly, dairy processing facilities, in which iodine and iodine sanitizers (e.g., iodophors) are commonly used, can be a potential and significant source of iodinated DBPs in receiving water bodies [
154]. The combined use of iodine, other iodine compounds, and disinfectants (i.e., chlorine) within dairy processing facilities or downstream WWTPs contributes to the formation of iodinated DBPs, especially iodinated THMs. In addition, coking wastewater generated from the manufacturing and processing of iron in steel industries has been found to possess a relatively high DBP FP [
155].
3.2.5. Coal-fired power plants
The addition of bromide-containing salts (e.g., calcium bromide) to a coal combustion unit is an effective method to enhance mercury removal in coal-fired power plants operating wet flue gas desulfurization [
152]. In general, the dosed bromide cannot be removed and will ultimately be discharged, leading to the enrichment of bromide in source water and downstream DWTP intakes [
156], [
157]. Although bromide in surface water typically does not pose a threat to ecosystems, a dramatic increase in total DBP formation and a shift to more brominated DBP species at downstream DWTPs can be expected when bromide loadings are elevated in source water [
158], [
159]. It was reported that a wet scrubber installation at a power plant increased bromide levels in raw water and the concentration of brominated DBPs in finished water at four downstream DWTPs [
152].
3.2.6. Gas extraction
Hydraulic fracturing has gradually emerged on a global scale, expanding the extraction of natural gas from geologic deposits. During this process, part of the injected water eventually returns to the surface as flowback water or produced water (fluids that are extracted together with the natural gas during production), both of which typically contain elevated levels of total dissolved solids, salinity, and toxic metals and metalloids [
160], [
161]. Hydraulic fracturing wastewater (HFW) may be reused onsite for gas extraction, injected into underground, or discharged to surface water after treatment [
151]. However, surface water can be adversely impacted by HFW due to surface leaks and spills occurring near drilling locations, the direct disposal of untreated wastewater, and the discharge of inadequately treated wastewater [
78].
The impact of HFW on the DBP formation at downstream DWTPs is mainly associated with bromide enrichment in drinking water sources [
151], [
162]. For example, the discharge of treated HFW was found to contribute to a sharp increase in the bromide level in a US river, ranging from 88 to 1220 μg·L
−1 [
152]. With the use of blended solutions prepared by diluting HFW with fresh water, Parker et al. [
24] found that the impact of HFW on DBP formation and speciation was largely dependent on the dilution factor and the type of disinfectant. More specifically, when the dilution factors were as low as 0.01%-0.1%, HFW promoted the formation of THMs and HANs during chlorination, the production of NDMA and iodinated THMs during chloramination, and the generation of bromate during ozonation [
24]. Moreover, a shift toward increased bromine substitution for the formation of THMs, dihalogenated HAAs, trihalogenated HAAs, and dihalogenated HANs during chlorination has been reported when the water to be disinfected is influenced by shale gas produced water [
163].
Elevated ammonia levels in treated oil and gas wastewater have been found to enhance the formation of N-DBPs during chlorination [
164]. Other non-bromide ions, including metal ions (e.g., magnesium, calcium, and barium) and sulfate, also affect DBP formation due to catalytic effects [
165]. Notably, in addition to DBP precursors (e.g., bromide and phenols), treated produced water from oil and gas extraction, in which the concentration of total THMs and dibromochloronitromethane could reach micrograms per liter level, can be a source of formed brominated DBPs [
56].
4. Discussion
In this section, we will discuss the technical feasibility of adopting DBPs and DBP precursors as promising indicators for assessing pollution in drinking water sources (
Fig. 4). In general, DBP precursors in drinking water sources can be determined using DBP FP tests, in which source water samples are dosed with a stoichiometric excess of specific disinfection agents (e.g., chlorine or chloramines) for a reaction time that is designed to maximize DBP formation under certain reaction conditions [
166], [
167]. The formed DBPs in
Fig. 4 are mainly DBPs derived from disinfected wastewater (e.g., WWTP effluent, farm effluent, and industrial effluent), which is discharged into adjacent natural water bodies and leads to the presence of DBPs in drinking water sources.
DBPs that are originally present in drinking water sources and DBP precursors determined using DBP FP tests can potentially be used to trace the origins of DOM in drinking water sources and evaluate the impacts of natural or anthropogenic events on source water quality. Moreover, DBPs and DBP precursors combined with other general parameters (e.g., DOC, DON, UV absorbance at 254 nm (UV254), and fluorescence excitation-emission matrix spectroscopy) and certain contamination indicators (e.g., primidone, carbamazepine, caffeine, and sucralose) can provide further insightful information. A detailed discussion is presented below.
4.1. Concentration and speciation of existing DBPs derived from disinfected wastewater and DBPs generated during DBP FP tests
Since DBPs are unique secondary contaminants produced during water disinfection, it is expected that the presence of relatively high levels of DBPs in drinking water sources indicates the effect of treated wastewater discharge on water sources [
14]. Reports have frequently identified a suite of DBPs (i.e., THMs, HAAs, and HALs) at high concentrations in source water impacted by WWTP effluents [
26], [
66], [
168]. Furthermore, the detection of iodinated DBPs in drinking water sources might result from the discharge of wastewater with a high iodinated DBP level, such as WWTP effluents that receive dairy waste and wastewater produced during dairy processing [
154]. Remarkably, high NA levels in source water can be associated with the discharge of WWTP effluent, treated farm wastewater, or industrial wastewater, and the special treatment of farm or industrial wastewater may generate specific NA species [
25], [
150]. More specifically, treated wastewater from metal finishing and electronics-related facilities, textile printing, and dyeing typically contain high levels of NDMA, whereas NMEA is barely detected in wastewater except for those from fishery ponds [
25], [
150].
With respect to DBP precursors, EfOM in WWTP effluent and AOM derived from algae blooms are important sources of DON in drinking water sources, which makes a significant contribution to the formation of N-DBPs (e.g., HANs, HAMs, HNMs, and NAs) during disinfection (
Fig. 4) [
71], [
145]. Notably, in addition to eutrophication and climate change, droughts and wildfires can trigger algal blooms. An increase in the overall DBP formation and a greater extent of bromine incorporation in DBPs during DBP FP tests can likely be ascribed to an elevated bromide level in drinking water sources. Similarly, increased iodide levels promote the formation of iodinated DBPs, especially in the case of chloramine disinfection. As mentioned above, increasing bromide and iodide levels in water sources may be caused by seawater intrusion, aquaculture wastewater discharge, coal-fired power plant discharge, and hydraulic fracturing. A series of micropollutants derived from various wastewater discharges (e.g., farm effluent, industrial discharge, and domestic discharge), including PPCPs, industrial chemicals, fungicides, and pesticides, can serve as the precursors of NAs [
17]. Moreover, the increased formation of iodinated DBP during a DBP FP test using free chlorine may be associated with contamination from wastewater containing high levels of iodinated X-ray contrast media [
51], [
169]. Interestingly, the ratio of HAN FP to THM FP has been adopted to assess the impact of wildfires on source water [
115], implying that different DBP FP ratios can be applied as pollution indicators.
4.2. DBPs and DBP precursors combined with other general parameters or certain contamination indicators
Several analytical methods have been used to characterize the organic matter present in source water and evaluate the effectiveness of water treatment processes, including bulk parameter (e.g., DOC and DON) determination, spectroscopic methods (e.g., UV and visible absorption spectroscopy, and fluorescence), chromatographic methods (e.g., high-performance size exclusion chromatography), and mass spectrometric methods (e.g., liquid chromatography-mass spectrometry and gas chromatography-mass spectrometry) [
170], [
171]. This subsection discusses studies that have combined DBPs and DBP precursors with other general parameters or certain contamination indicators to determine the origins of pollution in drinking water sources and to evaluate changes in source water quality arising from natural or human events.
Specific DBP FP (microgram of DBP per milligram of DOC), which is obtained by normalizing the DBP FP via DOC concentration, is considered to indicate the relative abundance of disinfectant-reactive components in organic matter; thus, it can provide insights into the sources of DBP precursors [
20], [
21]. For example, postfire rainfall samples were found to have increased reactivity in forming HAN and TCNM, whereas the impact of wildfire on DOM reactivity toward the formation of THMs and HAAs was reported to be insignificant [
5], [
116]. However, efforts are needed to quantitatively relate DOM origins with the ranges of specific DBP FP values, because this is expected to provide a benchmark for tracing the source of DOM. Wastewater discharge with higher DON content and a lower DOC/DON ratio can result in a higher N-DBP FP of wastewater-impacted water bodies [
145]. Recently, a companion study systematically investigated the spatial variation in DOM characteristics on a basin scale using size-exclusion chromatography and fluorescence spectroscopy and identified 36 DBPs that had formed in chlorinated surface water [
172]. Based on the results of DOM characterization and DBP formation, Fang et al. [
172] reported a high diversity in DOM characteristics along the Yangtze River; the varied DOM characteristics corresponded with different formation patterns of DBPs upon chlorination. In addition, inspired by the perspective that DBP monitoring can provide insights into hydro-biogeochemical dynamics within a watershed, Leonard et al. [
173] combined historical drinking water quality archives, including the speciation and levels of DBPs, synoptic watershed sampling, and targeted spectroscopic characterization methods, to identify an export control point for DOM in a subalpine headwater stream.
In addition to the aforementioned general parameters, certain contamination indicators in source water can be jointly used with DBPs and DBP precursors to provide more in-depth information regarding pollution in drinking water sources. Primidone, carbamazepine, and NDMA FP have been shown to be possible indicators of wastewater discharge [
8]. Based on this finding, Aydin et al. [
19] made efforts to explore the source of contaminants in a watershed in Turkey by determining caffeine levels and measuring THM FP and NDMA FP. The presence of caffeine and THM FP suggests that the pollution is due to the discharge of municipal wastewater, whereas NDMA FP is mostly associated with domestic discharge, industrial discharge, and/or agriculture runoff. With the combined use of these indicators, Aydin et al. [
19] were able to identify specific contamination sources of the watershed.
5. Summary and environmental insights
DBP precursors in drinking water sources have a range of origins, including natural, domestic, industrial, and agricultural sources. These precursors are generally classified into inorganic halide ions, NOM, and anthropogenic contaminants. The presence of bromide and iodide can promote total DBP formation, especially brominated DBPs and iodinated DBPs. With respect to NOM, allochthonous NOM with a high aromatic content can act as the primary precursors of C-DBPs, whereas autochthonous NOM with low aromatic and high organic nitrogen contents contributes significantly to the formation of N-DBPs. Research related to DBP formation from anthropogenic contaminants mostly focuses on NAs and iodinated DBPs. There are two primary disinfectant types—that is, chloramines and ozone—that have a higher potential to react with the precursors to form NAs. Moreover, it has been found that conventional wastewater effluent can be a dominant source of chloramine-reactive NA precursors and ozone-reactive NA precursors. It is notable that iodinated DBPs are generally formed from the reactions of iodide, NOM, and chloramine, while iodinated X-ray contrast media significantly contribute to the formation of iodinated DBP upon chlorination.
DBP precursors originating from various sources (e.g., natural, domestic, industrial, and agricultural sources) exhibit different DBP formation behaviors. Moreover, based on the theoretical absence of DBPs in non-polluted natural waters, the detection of relatively high levels of DBPs in drinking water sources implies contamination by treated wastewater discharge (e.g., WWTP effluent, farm effluent, and industrial effluent). Thus, we argue that DBPs and DBP precursors in raw water can serve as alternative indicators to assess the pollution of drinking water sources. Here, environmental insights and future research needs are highlighted in three aspects.
(1) The concentration and speciation of DBPs originally present in source water and DBP precursors determined using DBP FP tests can provide insightful information on the pollution significance of drinking water sources and can be used to trace pollution origins. When used jointly with other common water quality parameters (e.g., DOC, DON, UV254, fluorescence, and molecular weight distribution) or certain individual contaminants (e.g., primidone, carbamazepine, caffeine, and sucralose), DBPs and DBP precursors can offer a more comprehensive picture of the impacts of various natural incidents and human-induced activities on environmental quality in watersheds that serve as drinking water sources.
(2) Effective and feasible risk management for drinking water sources requires the identification of harmful substances, pollution origins, and hazardous events posing potential risks to water quantity and quality. This study suggests that DBPs and DBP precursors can be used as potential contamination indicators to improve the understanding of policymakers, stakeholders, and researchers regarding the origins of major pollutants in drinking water sources. Desired management strategies can then be tailored accordingly to minimize pollution impacts.
(3) Academic studies are urged to offer more quantitative information and to establish a database regarding the use of DBP/DBP precursors as indicators and corresponding pollution information. For example, a quantitative correlation of different DOM sources with specific DBP FP (microgram of DBP per milligram of DOC) ranges or DBP FP ratios (e.g., the ratio of HAN FP to THM FP) may provide a benchmark for tracing the origins of DOM in drinking water sources. Moreover, the challenge of how to better integrate alternative indicators with traditional water quality parameters must be surmounted in future research.
Acknowledgments
This work was supported by the National Natural Science Foundation of China (52325001, 52170009, and 52091542), the National Key Research and Development Program of China (2021YFC3200700), the Program of Shanghai Academic Research Leader, China (21XD1424000), the International Cooperation Project of Shanghai Science and Technology Commission (20230714100), the Key-Area Research and Development Program of Guangdong Province (2020B1111350001), and Tongji University Youth 100 Program.
Compliance with ethics guidelines
Rong Xiao, Yang Deng, Zuxin Xu, and Wenhai Chu declare that they have no conflict of interest or financial conflicts to disclose.