Environ. Sci. Technol. 2002, 36, 3556-3563
Fluorescent Whitening Agents in Tokyo Bay and Adjacent Rivers: Their Application as Anthropogenic Molecular Markers in Coastal Environments YUKO HAYASHI, SATOSHI MANAGAKI, AND HIDESHIGE TAKADA* Laboratory of Organic Geochemistry, Faculty of Agriculture, Tokyo University of Agriculture and Technology, Fuchu, Tokyo 183-8509, Japan
Two kinds of stilbene-type fluorescent whitening agents (i.e., DSBP and DAS1), minor components of laundry detergents, were analyzed in surface waters of Tokyo Bay and adjacent rivers and in sewage effluents to examine their usefulness as molecular markers in the marine environment. Sensitive determination using HPLC (high performance liquid chromatography) with fluorescence detection with postcolumn UV radiator was employed. DSBP and DAS1 were found in Tokyo rivers at concentrations of a few µg/L and ∼1 µg/L, respectively. DSBP and DAS1 were widely distributed in Tokyo Bay waters at concentrations in the range of 0.019-0.264 µg/L and 0.021-0.127 µg/ L, respectively. Comparison of these concentrations with those in sewage effluents (DSBP: 8 µg/L and DAS1: 2.5 µg/L on average) yielded sewage dilutions in Tokyo Bay on the order of 102. FWAs-salinity diagram in the Tamagawa Estuary showed fairly conservative behaviors of the FWAs with ∼20% and ∼10% removal of DSBP and DAS1, respectively. This is thought to be caused by photodegradation. The persistent nature of FWAs and their widespread distribution in coastal environments demonstrates the utility of FWAs in tracing the behavior of water from rivers and sewage outfalls. The DSBP/DAS1 ratio showed a decreasing trend from sewage effluents, to rivers, to Tokyo Bay, indicating selective photodegradation of DSBP. The DSBP/DAS1 ratio is proposed as an index of the degree of photodegradation and residence time and freshness of water mass in coastal environments.
Introduction During the past few decades various anthropogenic molecular markers have been discovered (1). They are specific to certain pollution sources (e.g. domestic wastes, sewage sludge, streetrunoff) and can be used to identify sources of pollutants and to estimate the transport pathways and fate(s) of anthropogenic contaminants. Typical markers include coprostanol, which is derived from human feces (2, 3), linear alkylbenzenes (LABs), and trialkylamines (TAMs), which are contained in anionic surfactants (4, 5) and cationic surfactants (6, 7), respectively, and benzothiazoles from automobile tires (8, * Corresponding author phone: +81-42-367-5825; fax: +81-42360-8264; e-mail:
[email protected]. 3556
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 16, 2002
FIGURE 1. Structure of the two studied FWAs and internal standard. Full name: DSBP, 4,4′-bis(2-sulfostyryl)biphenyl; DAS1, 4,4′-bis[(4-anilino-6-morpholino-1,3,5-triazin-2-yl)amino]stilbene-2,2′-disulfonate; internal standard, 4, 4′-bis(5-ethyl-3-sulfobenzofur-2yl)biphenyl. 9). These markers have been used to detect inputs from pollution sources to coastal environments. For example, coprostanol (10) and LABs (11) were measured in sediments from Narragansett Bay, U.S.A., to determine the area affected by sewage particles discharged from surrounding sewage treatment plants. In a study of Chesapeake Bay, U.S.A., coprostanol measurements indicated that accumulation of sewage particles occurred in an area 2 km away from the sewage outfall in addition to their accumulation in immediate vicinity of the outfall (12). LABs have also been detected in deep sea sediments, such as the San Pedro Basin (869 m water depth (4)) and the Deep Water Dump Site 106 (∼2000 m water depth (13)), demonstrating the long-range transport of sewage particles discharged from submarine wastewater outfall systems and barges, respectively. Most anthropogenic markers used so far are hydrophobic and are useful for tracking sewage particles and particlereactive pollutants. On the other hand, hydrophilic markers are potentially useful for tracing the movement of water mass and water-soluble pollutants in aquatic systems. Because endocrine disrupting activities have been recently demonstrated for some water-soluble contaminants, such as 17βestradiols and nonylphenol, in sewage effluents (14, 15), hydrophilic molecular markers are becoming increasingly useful. However, only a small number of water-soluble markers have been proposed (1, 16-18). Furthermore, most water-soluble markers (e.g. linear alkylbenzenesulfonates (19), urobilin (20)) are not persistent, are easily degraded in freshwater and estuarine environments, and are difficult to exploit in marine environments. Recently, some stilbene-type fluorescent whitening agents (FWAs), minor components (∼0.15%) of laundry detergents, have been proposed as persistent water-soluble molecular markers (21). They include 4,4′-bis(2-sulfostyryl)biphenyl (DSBP) and 4,4′-bis[(4-anilino-6-morpholino-1,3,5-triazin2-yl)amino]stilbene-2,2′-disulfonate (DAS1) as shown in Figure 1. They are water soluble due to the ionic sulfonyl group, although a small but significant part of the FWAs is adsorbed to suspended particles in the water column. A laboratory adsorption experiment using river sediment yielded a distribution coefficient (Kd) of ∼102 (22). Approximately 90% of the FWAs were found in the operationally defined dissolved phase of secondary wastewater effluents 10.1021/es011352o CCC: $22.00
2002 American Chemical Society Published on Web 07/17/2002
FIGURE 2. Study area and sampling locations: (a) Tokyo Bay, rivers, and sewage treatment plants and (b) the Tamagawa Estuary. (23). Because they are poorly biodegradable (24), the FWAs could be useful to detect sewage input to coastal zones and to trace dilution and movement of sewage effluents in marine environments. Intensive studies have been conducted in Swiss rivers and lakes (21, 23-29). Their potential as molecular markers in aquatic environments has been proposed (1, 30, 31). However, the occurrence and distribution of FWAs in marine environments have not been reported. Here, we report for the first time the distribution of FWAs in a coastal zone (i.e., Tokyo Bay) together with those in adjacent rivers and wastewater effluents. The behavior of FWAs in an estuary was also studied to examine the stability of FWAs in estuaries where many land-derived pollutants are removed before they reach the marine environment.
Materials and Methods Site Description/Sampling. The study area and sampling locations are shown in Figure 2. Tokyo Bay is surrounded by one of the most urbanized areas (i.e., Tokyo Metropolitan area) in the world. The population in the catchment is approximately 26 million, and large amounts of untreated and treated wastewater are introduced to the bay through rivers and effluents from sewage treatment plants located along the coast. The wastes of approximately 80% of the population are processed by municipal sewage treatment plants (32), and the wastewater is discharged to rivers and the bay following primary and secondary treatment. Gray water (i.e., domestic wastewater excluding human feces) from the remaining 20% is discharged to rivers without any treatments; human wastes are treated prior to discharge. Tokyo Bay has an area of 1200 km2 with an average water depth of 15 m. The residence time of water in the bay is estimated to be 1-3 months (33). The Sumidagawa and Tamagawa Rivers are major rivers flowing into Tokyo Bay. They comprise ∼30% of the freshwater inflow to the bay.
The Sumidagawa River has a drainage basin of 610 km2 and a population of 5.6 million people. Its length is 50 km, and it has a freshwater inflow of approximately 30 m3/s. The Tamagawa River is 140 km long with a normal freshwater inflow of ca. 15 m3/s. The population and area of its drainage basin are approximately 3 million and 1240 km2, respectively. The Tamagawa estuary is a partially mixed estuary that extends about 12 km with no secondary tributaries. Based on the volume of the estuary and the freshwater inflow, the residence time of the freshwater is estimated to be approximately 2-4 days (34). The 24-h composite and grab wastewater samples (raw sewage, primary effluent, secondary effluent) were obtained from four full-scale mechanical-biological sewage treatment plants in Tokyo (STP1-STP4) during 1997 and 1998. Freshwater surveys were also conducted in the Tamagawa River (TR01 and TR02) and the Sumidagawa River (SR01-SR03) at 5 times (August and October 1997 and February, May and August 1998). The same sampling regime in 1997 and 1998 was conducted in the Tamagawa Estuary where water samples were collected at station TE4 (Figure 2b). Additional water samples were collected from six tributaries (a-f) of the Tamagawa River in May, 1998. A survey was conducted in the Tamagawa estuary in August, 1998, and water samples were collected from nine locations along the salinity gradient (Figure 2b). The physical characteristics on the estuary survey are listed in Table 1. Tokyo Bay water samples were collected at nine locations (F1-F7 and st.02, st.05, and st.06) during the cruise of R/V Seiyo-maru (Tokyo Fishery University) in June 1998 and at three locations (T6, T8, T14) in October 1998. The water samples were taken from surface layer using a stainless steel bucket. All samples were transferred to glass amber bottles immediately after sampling and were stored in the dark without preservative at 5 °C. The samples were analyzed within 1 week of sampling. To examine sorption/ degradation loss during storage, an aliquot of a water sample collected from Tokyo Bay was analyzed immediately after sampling, and another aliquot was analyzed after 1-week storage in the dark at 25 °C. No significant decrease ( 0.99) was confirmed for the concentration range of 0.4200 ng/mL for DAS1 and 0.05-20 ng/mL for DSBP. Analytical standards were kindly supplied by Dr. J. Kramer of CibaGeigy AG, Switzerland. Analytical precision was examined through replicate (4 times) analyses of 50 mL of Tokyo Bay water containing 0.031 µg/L of DSBP and 0.044 µg/L of DAS1. Relative standard deviations for DSBP and DAS1 concentra3558
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 36, NO. 16, 2002
tions were 3.8% and 1.8%, respectively. To examine FWA recovery, 1 ng (10 ng/mL × 100 µL) each of DSBP and DAS1 was spiked into 50 mL of the Tokyo Bay water sample, and the sample was analyzed for the FWAs. The recoveries of spiked standards were 92.0% for DSBP and 93.5% for DAS1. Furthermore, to monitor the performance of the method during routine analysis, a recovery surrogate (Figure 1c; 4,4′bis(5-ethyl-3-sulfobenzofur-yl)biphenyl) was spiked into every sample before the solid-phase extraction and its recovery was determined. The recovery of the surrogate was more than 90% for all of the samples. Therefore, no recovery correction was made. When 100 mL of distilled water was analyzed, 0.04 ng of DSBP and 0.5 ng of DAS1 was detected. Therefore, quantitation limits for DSBP and DAS1 were set at 0.4 ng/L and 5 ng/L, respectively, in the present study. A breakthrough for extraction of 200 mL of a Tokyo Bay water was examined using tandem cartridges, and no significant amounts of analytes (
