Environ. Sci. Technol. 2003, 37, 1713-1719
Effects of Three Pharmaceutical and Personal Care Products on Natural Freshwater Algal Assemblages BRITTAN A. WILSON* Department of Ecology and Evolutionary Biology, University of Kansas, Lawrence, Kansas 66045 VAL H. SMITH Department of Ecology and Evolutionary Biology, University of Kansas, Lawrence, Kansas 66045 FRANK DENOYELLES, JR. Kansas Biological Survey, University of Kansas, Lawrence, Kansas 66045 CYNTHIA K. LARIVE Department of Chemistry, University of Kansas, Lawrence, Kansas 66045
Treated wastewaters in the United States contain detectable quantities of surfactants, antibiotics, and other types of antimicrobial chemicals contained in pharmaceutical and personal-care products (PPCPs) that are released into stream ecosystems. The degradation characteristics of many of these chemicals are not yet known, nor are the chemical properties of their byproducts. They also are not currently mandated for removal under the U.S. Clean Water Act. Three representative PPCPs were individually tested in this study using a series of laboratory dilution bioassays: Ciprofloxacin (an antibiotic), Triclosan (an antimicrobial agent), and Tergitol NP 10 (a surfactant), to determine their effects on natural algal communities sampled both upstream and downstream of the Olathe, KS wastewater treatment plant (WWTP). There were no significant treatment effects on algal community growth rates during the exponential phase of growth, but significant differences were observed in the final biomass yields (p < 0.001). All three compounds caused marked shifts in the community structure of suspended and attached algae at both the upstream and downstream sites (p < 0.05). Increasing the concentrations of all three compounds over a 3 orders of magnitude range also caused a consistent decline in final algal genus richness in the bioassays. Our results suggest that these three PPCPs may potentially influence both the structure and the function of algal communities in stream ecosystems receiving WWTP effluents. These changes could result in shifts in both the nutrient processing capacity and the natural food web structure of these streams.
Introduction Many chemical ingredients in the products that we use daily are subsequently released into the environment. For example, * Corresponding author phone: (785) 864-4565; fax: (785) 8645321; e-mail:
[email protected]. 10.1021/es0259741 CCC: $25.00 Published on Web 03/14/2003
2003 American Chemical Society
approximately 70 000 human-produced compounds are currently in use, with more than 1000 new formulations being added each year (1). After use, many of these parent chemicals and their degradation products are then introduced into the environment through a variety of routes (2). Rivers and streams serve as primary conduits for chemical waste, despite the fact that the freshwater contained within these ecosystems is itself a valuable strategic resource (3-5). We focus here on the impacts of chemical compounds that enter stream ecosystems from municipal wastewater treatment plant (WWTP) effluents. These WWTP effluents frequently contain a wide variety of anthropogenic chemicals, including compounds originating from the use and disposal of pharmaceutical/human health and personal care products (PPCPs). WWTPs are not typically designed for the effective removal of these specialized organic chemicals, and as a result many of these compounds can be found in detectable concentrations in streamwaters worldwide (2, 6, 7). Although human health and personal care chemicals may interact with the resident biota in receiving stream ecosystems, their potential ecological effects have not yet been quantified. There is reason to expect that PPCPs may have significant impacts on natural biotic communities. For example, widely used antimicrobial agents such as those found in hand soaps and toothpastes are typically designed to kill or to inhibit the growth of a wide range of “undesirable” microbial species. Such broad-spectrum biological activity potentially could cause unintended impacts on sensitive, co-occurring nontarget organisms in the resident community. Our ability to predict these potential collateral effects on natural communities is currently very limited, however. Moreover, individual chemical compounds potentially can interact synergistically or antagonistically with other chemical agents or stressors that may be present in the affected environment (8). We report here an initial assessment of the ecological impacts of three compounds commonly found in PPCPs: Triclosan, an antimicrobial agent; Ciprofloxacin, a commonly prescribed antibiotic; and Tergitol NP 10, a surfactant. All three compounds are frequently used or consumed in U.S. households (2, 7, 9), and both Triclosan and Tergitol NP10 and similar compounds can also be found in commercially available products that are used in industrial applications (10-12). The potential effects of these three compounds on the algal communities of stream ecosystems were explored using a series of laboratory bioassays.
Materials and Methods Sites and Sampling Methods. The streamwater analyzed in this study was collected from Cedar Creek, KS, at sites located both upstream and downstream of treated wastewater effluent released from the Olathe, KS WWTP. This activated sludge facility treats 3 million gallons per day, and seasonally the wastewater contributes one-third to one-half of the total Cedar Creek streamflow. A sewage-impacted sample was taken 25 m downstream from the effluent outfall. A nonimpacted site was also chosen 0.5 mi upstream from the Olathe treatment plant; this latter reference condition site was chosen using county land-use maps to ensure that no known industrial or major agriculture sources were present in the upstream portion of the stream. Water column samples were taken from 1-1.5-m-deep pools at each of the two sites, using a plexiglass integrating tube sampler. Samples were taken from evenly spaced points across the longest pool transect, and two subsamples were immediately preserved onsite with Lugol’s preservative for VOL. 37, NO. 9, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1713
TABLE 1. Characteristics of the Three Pharmaceutical and Personal Care Product Compounds Tested in This Study product
chemical formula
Ciprofloxacin
1-cyclopropyl-6-fluoro-1, 4dehydro-4-oxo-7-(1-piperazinyl)-3-quinolinecarboxylic acid 5-chloro-2-(2,4-dichlorophenoxy)phenol
Triclosan Tergitol R (NP 10)
poly(oxy-1,2-ethanediyl,R(4-nonylphenyl)-Ω-hydroxy
description
86393-32-0
fatty acid biosynthesis inhibitor antiseptic (2); contained in many household products including toothpaste, acne creams/soaps, athlete’s foot ointments, and antiseptic-impregnated kitchen items strong water-soluble surfactant; disrupts cell membrane integrity and deionizes the cell (12); used as an industrial surfactant, and also contained in most spermicidal lubricants and gels (nonoxynol-9), hair dyes, and nail treatments; it is a polymer of ethylene oxide and nonylphenol
3380-34-5
later algal identification and counts using optical microscopy (13). The remaining whole water samples were then returned to the laboratory for immediate bioassay analyses (see below), and for spectrophotometric measurements of total nitrogen (14) and total phosphorus (15) concentrations in the streamwater. Measurements of water temperature, dissolved oxygen, specific conductivity, in vivo fluorescence, turbidity, pH, and flow rate were also made just below the surface and at every 0.5 m depth, using a YSI 6600 multiple-probe sonde and a YSI 610DM datalogger. The characteristics of the riparian zone and adjacent land use were qualitatively surveyed on each sampling date as well. The structure of algal communities is very sensitive to environmental contaminants (16), and algal assemblages have been successfully used to monitor the impacts of aquatic stressors (17, 18) and to assess the biotic integrity of stream ecosystems (19, 20). Algae respond rapidly and predictably to a wide range of pollutants and can act as an early warning system (21). In this study we used bioassays containing natural algal assemblages from the sewage-impacted and nonimpacted sites to evaluate the sensitivity of Cedar Creek biota to the three target compounds (Table 1). All assays utilized Triclosan obtained from the Rita Corporation (95% pure), Ciprofloxacin obtained from Serological Proteins Inc. (95% pure), and Tergitol R NP 10 obtained from Sigma-Aldrich (90% pure). The two stream sites were sampled for single concentration bioassays in November 2001 and January 2002 (experiment 1). Two additional collections from the upstream site were subsequently taken in April and May of 2002 for a second experiment using varying concentrations of the original test compounds. Bioassay Protocols, Experiment 1. Batch culture dilution assays of the resident suspended algal communities were initiated within 12 h of each stream sampling to assess the effects of the three PPCPs on the rate of algal biomass accumulation, total algal biomass production, and algal community structure. A water sample from both of the experimental (upstream and downstream) sites was first filtered through an 80-µm Nitex mesh to remove macrozooplankton and large particulate debris. A 1-L portion of Nitex filtrate from each site was then refiltered at low vacuum through a Whatman GF/F glass fiber filter to remove the suspended algae. For each site, the natural stream algae retained by the filter were then diluted to 1:20 with WC medium (22) containing the PPCPs compound of interest. The algae-inoculated WC medium from each of the two sampling sites was then separately transferred into a series of 60 sterile, 20-mL glass screw-capped test tubes, for a total of 120 tubes. In experiment 1, each of the two sets of 60 tubes (sewageimpacted and nonimpacted) utilized four treatment groupings (Table 2). The bioassay concentrations chosen for Triclosan and Ciprofloxacin in experiment 1 were based on 1714
9
CAS number
fluoroquinolone carboxylic acid; a Gyrase inhibitor antibiotic frequently present in hospital and domestic wastewater (2); ciprofloxacin hydrochloride was specifically used
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 9, 2003
26027-38-3
TABLE 2. Experimental Concentrations (µg/L) Used for Each Test Group Experiment 1a
Experiment 2a
Triclosan
Ciprofloxacin
Tergitol NP10
0.12
0.15
0.20
a
Triclosan
Ciprofloxacin
Tergitol NP10
0.012 0.12 1.2
0.015 0.15 1.5
0.005 0.05 0.50
Both experiments also had a control group in WC growth medium.
maximum streamwater concentrations observed in Europe (2, 6) and correspond to the average values that have been reported recently in the United States (7). Analytical methods for the quantitation of Tergitol NP10 are still not fully satisfactory, and the average concentration of this surfactant in receiving waters has not yet been established. A concentration of 0.2 ppm was used in experiment 1, a level greatly exceeding that of the other two compounds studied. However, Tergitol NP10 was tested at levels 1-2 orders of magnitude lower in the subsequent concentration gradient study (below). All test tubes in experiment 1 were placed in a Percival incubator to control environmental conditions. The tubes were incubated at 18 °C with a 12/12 light cycle at 500 µE m-2 s-1. The test tubes were kept aerated by leaving the caps slightly loose to allow air exchange. During days 1-13 following algal inoculation, fluorometric estimates of algal biomass were made by sacrificing one test tube from each of the four experimental treatments in the upstream (nonimpacted) and downstream (sewage-impacted) groups. Measurements of extracted chlorophyll a were then made for both attached and suspended algae in each of the 8 test tubes that were sacrificed on a given sampling date. The entire content of suspended algae in each test tube was captured by filtering the entire 15-mL test tube volume through a Whatman GF/C glass fiber filter, and the algae were frozen at -20 °C for later extraction with boiling 90% ethanol. The remaining chlorophyll a associated with attached algae growing directly on the test tube wall was then extracted by adding 15 mL of boiling 90% ethanol directly to each tube. The concentration of chlorophyll a in both sets of ethanol extracts was then measured using a Turner Quantec fluorometer (23). All fluorometric readings were converted to chlorophyll a concentrations using Turner chlorophyll a standards, and these data were then used to estimate the biomass of algae in each test tube on each date sampled. These time series data were then used to estimate the maximal rate of biomass accumulation in each tube (r, day-1), by performing linear regressions on biomass data taken during the exponential phase of growth. In addition to the above measurements, the last three test tubes from each treatment group were preserved with
FIGURE 1. Total accumulation of chlorophyll a by treatment: (A and B) November 2001; (C and D) January 2002. Key to experimental treatments: CSC ) nonimpacted stream (NI) control; CSTRI ) NI Triclosan; CSCIP ) NI Ciprofloxacin; CSTER ) NI Tergitol; SSC ) sewage-impacted stream (SI) control; SSTRI ) SI Triclosan; SSCIP ) SI Ciprofloxacin; SSTER ) SI Tergitol. Lugol’s iodine on the final sampling date. Algal cell counts were then performed using a Sedgewick-Rafter cell to identify the major algal genera that were presen, to estimate the final abundance of each genus, and to measure the average cellular biovolume of each genus counted. Genus-level identification has been found to be sufficient for detecting the effects of pollutants on biotic community structure, and in some cases has been considered to be more reliable than species-level studies (20, 24). For genera containing more than one species, a weighted average of each species’ biovolume was used to determine the total biovolume for the genus being analyzed. The number of algal units counted and measured varied with the absolute density of the algal sample being analyzed. One hundred microscopical fields at 430× were counted for each sample using an Olympus BH-2 phase-contrast microscope. In addition, the total algal biovolume attained on the last sampling date (day 13) was used as a measure of total biomass accumulation (final yield). All statistical analyses were performed using the SYSTAT statistical software for Windows. Bioassay Protocols, Experiment 2. Natural algal assemblages from the nonimpacted upstream site were used to evaluate the sensitivity of Cedar Creek biota to a range of concentrations of the three target compounds examined in experiment 1. All dilution assays were performed using the procedures described above. The natural stream algae were filtered and re-suspended, and this cell suspension was diluted 1:20 with WC medium containing the PPCP compounds at each concentration of interest. The algaeinoculated WC medium was then transferred into a series of 40 sterile, 20-mL glass screw-capped test tubes. The bioassays in experiment 2 comprised 10 different treatment groups (Table 2). These bioassay concentrations were chosen to include levels 1 order of magnitude above and below those used for Triclosan and Ciprofloxacin in experiment 1. However, because the initial concentration of Tergitol NP 10 that was used in experiment 1 was much higher than that of the other two compounds, the two lowest
concentrations for Tergitol in experiment 2 were chosen to allow an evaluation of its ecological effects at concentrations that were 1 and 2 orders of magnitude lower than those used previously.
Results and Discussion Experiment 1: November 2001 and January 2002. Experiment 1 consisted of analyses of differences in dissolved oxygen, pH, specific conductivity, temperature, turbidity, and suspended algal biomass (measured here as in vivo fluorescence) between the two stream sites (upstream, downstream) on both sampling dates. In particular, streamwater total nitrogen and total phosphorus concentrations were significantly higher at the downstream site, reflecting point source inputs of N and P from the WWTP plant. Flow rates between the two sites did not significantly differ (data provided as Supporting Information). Impact of Triclosan, Ciprofloxacin, and Tergitol NP 10 on Algal Biomass. Maximum exponential rates of chlorophyll a accrual in the algal bioassays (r, day-1; based upon fluorescence measurements during log-phase growth) were not found to differ significantly, either across chemical treatments or between the nonimpacted and impacted sites (Figure 1). However, strong treatment effects on final yield were found when the algal biovolume data from the last bioassay date were analyzed. When data from both stream sites were considered, final algal yields in the Tergitol NP 10 treatment were reduced to only 15% of control levels in November and 22% of controls in January (Figure 2). An analysis of variance (ANOVA) of the absolute biomass data from the last sampling date confirmed strong treatment effects on final biomass yields (p < 0.001) attributable to the effects of Tergitol NP 10, and also revealed that the biomass response differed significantly between the two sampling sites (p < 0.05). In contrast, the antibiotic Ciprofloxacin and the antimicrobial compound Triclosan did not significantly alter algal biomass yields in November, although the Ciprofloxacin treatment biomass increased to 120% of control values in January. VOL. 37, NO. 9, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1715
FIGURE 2. Comparison of total biovolume yields on the final date of the bioassay for both streams by treatment, November 2001 and January 2002. Key to experimental treatments: 1 ) control; 2 ) Triclosan; 3 ) Ciprofloxacin; and 4 ) Tergitol NP 10. Impact of Triclosan, Ciprofloxacin, and Tergitol NP 10 on Algal Community Structure. On the basis of microscopical measurements of average algal cell sizes and their respective geometric shapes, the population densities (cells/mL) of each genus detected in the laboratory bioassays were converted to biovolume prior to statistical and graphical analyses. The percent of total algal biovolume contributed by each algal genus on the final bioassay sampling date was then calculated for both sites. The effects of these three chemicals on individual genera were then analyzed for both sites using a series of one-way ANOVAs. The effects of both Triclosan and Ciprofloxacin on absolute algal biomass were statistically significant relative to controls at the upstream, nonimpacted site for the chroococcalian cyanobacteria and for the green alga Chlamydomonas, but not for other algal genera in November 2001. In January, the effects of Ciprofloxacin in absolute biomass were statistically significant for chroococcalian cyanobacteria and Chlamydomonas. However, the effects of Triclosan were significant for Chlamydomonas and Scenedesmus. In contrast, the effects of Tergitol NP 10 were highly significant at the upstream site for the biomass of all five algal genera present in the upstream bioassay in November. The effects for Tergitol were significant for genera upstream except for Navicula in January (significance was assigned for all P-values less than 0.05, see Supporting Information for individual P-values). Sites located downstream of the WWTP are continuously exposed to chemicals contained in the effluent plume, and algae in these communities thus might be expected to show evidence of adaptation or species selection due to this exposure. It was hypothesized that treatment effects on algal community structure would also differ between the sampling two sites. Ciprofloxacin was not found to have significant effects on any of the algal genera present in the downstream site bioassay in November, but the effects were statistically significant for Synedra and Chlamydomonas in January. However, upstream-downstream differences in communitylevel responses to the bioassay were found for both Triclosan and Tergitol NP 10. The presence of Triclosan resulted in a significant increase in the relative biomass of Melosira at the downstream site (both samplings), and Tergitol NP 10 caused significant changes in the relative biomass of Navicula, Synedra, Melosira, and Scenedesmus in November. In January, the effects were significant for these genera as well as chroococcalian cyanobacteria, Sphaerocystis, and Scenedesmus (significance assigned for all P-values less than 0.05, see Supporting Information for individual P-values). These results suggest that there may be an adaptive response of the algal community to prolonged exposure, as is commonly found in bacteria-antibiotic interactions (25, 26). A two-way MANOVA was used to determine whether the effects of these toxicants differed between the upstream and downstream sampling sites. In November, for Navicula, 1716
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 9, 2003
Synedra, and the two chroococcalean cyanobacteria, Tergitol NP 10 caused the only significant treatment effect. For Sphaerocystis and Chlamydomonas, the treatment of Tergitol NP 10 was significant, as was the effect of site, and the treatment*stream interaction term (significance assigned for all P-values less than 0.05, see Supporting Information). For Melosira, Anabaena, and Scenedesmus the effect of site was significant in all cases because these three taxa were not detected in samples taken upstream. There was thus no way to test whether significant differences between the two sites for these genera existed. The effect due to site differed in significance between the two samplings. Site effects were significant for Navicula, Scenedesmus, and Sphaerocystis for all treatments in the January sampling but not in November. The effects found for Tergitol NP 10 for treatment, stream, and treatment*stream interactions were significant for Navicula, chroococcalian cyanobacteria, Sphaerocystis, and Scenedesmus. Ciprofloxacin was significant for Synedra and Chlamydomonas. Triclosan was significant only for Navicula. For Melosira, Pediastrum, and Anabaena the effect of site was significant in all cases because these three taxa were not detected in samples taken from both sites, and thus were not present in the initial inoculum (significance assigned for all P-values less than 0.05, see Supporting Information). Experiment 2: April and May 2002. Analyses of water quality at the upstream sampling site showed similarities in dissolved oxygen, pH, suspended algal biomass (measured as in vivo fluorescence), total nitrogen, and total phosphorus between the two sampling dates. Water temperature, specific conductivity and turbidity decreased significantly by the second sampling (P < 0.05). No significant differences were found between the two samplings for flow rate, which was exceptionally low at all times (Supporting Information). Final yields of chlorophyll a were significantly lower than control values for Tergitol NP 10 at all treatment concentrations (Figure 3). In addition, all three chemicals caused significant shifts in algal composition. Triclosan and Ciprofloxacin both had significant effects on rare algal genera; in particular, these two compounds significantly decreased the absolute biovolume of Chlamydomonas at higher concentrations on both sampling dates. Ciprofloxacin also had a significant impact on common diatoms: biovolume of Synedra increased significantly at lower concentrations, with a corresponding decrease in Navicula. Tergitol NP 10 had significant effects on the final biovolume of both common and rare genera at all concentrations. In April, Triclosan had significant impacts on both common and rare algal genera at the two highest concentrations studied. However, there was a significant reduction of Chlamydomonas at concentrations of 0.15 µg/L and 1.5 µg/ L, and a significant reduction of Sphaerocystis at 1.5 µg/L (P < 0.05). Ciprofloxacin caused a significant increase in the
FIGURE 3. Comparison of total biovolume yields on the final date of the bioassay for both samplings by treatment in (A) April 2002 and (B) May 2002. Key to experimental treatments: 1 ) control; 2 ) 0.015 µg/L Triclosan; 3 ) 0.15 µg/L Triclosan; 4 ) 1.5 µg/L Triclosan; 5 ) 0.012 µg/L Ciprofloxacin; 6 ) 0.12 µg/L Ciprofloxacin; 7 ) 1.2 µg/L Ciprofloxacin; 8 ) 5 µg/L Tergitol; 9 ) 50 µg/L Tergitol; 10 ) 500 µg/l. Tergitol.
FIGURE 4. Effects of varying concentrations of Triclosan, Ciprofloxacin, and Tergitol NP 10 on algal genus richness in experiment 2. common diatom Synedra at concentrations of 0.012 µg/L and 0.12 µg/L, and a significant reduction of the common diatom Navicula at 0.12 µg/L. Reductions also occurred in the green alga Chlamydomonas at 0.12 µg/L and 1.2 µg/L, and for Sphaerocystis at 1.2 µg/L (P < 0.05). Tergitol NP 10 significantly reduced the biomass of Navicula, Synedra, chroococcalian cyanobacteria, Sphaerocystis, and Chlamydomonas (P < 0.05) at all experimental concentrations. In May, Triclosan caused a significant increase in Synedra and a significant reduction of the rare genus Chlamydomonas at 0.015 µg/L and 0.15 µg/L. There was also a significant reduction of chroococcalian cyanobacteria at the concentration of 0.15 µg/L (P < 0.05). Ciprofloxacin caused a significant increase in the common diatom Synedra at the concentration of 0.12 µg/L. Although there was no effect on Navicula, significant reductions occurred in the green alga Chlamydomonas and chroococcalian cyanobacteria at 0.12 µg/L and 1.2 µg/L (P < 0.05). Tergitol significantly reduced Synedra, chroococcalian cyanobacteria, and Chlamydomonas (P < 0.05) at all concentrations studied. There was also a significant
reduction in Sphaerocystis at 50 µg/L and 500 µg/L. Unfortunately, no data were available for Ciprofloxacin at the 0.012 µg/L concentration due to an accidental loss of the samples. The effects of these three toxicants on algal genus richness on the final date of the experiment 2 bioassays are presented in Figure 4. Figure 4 suggests that there was a graded response to each of the three compounds: increasing the environmental concentrations of Triclosan, Ciprofloxacin, and Tergitol NP 10 resulted in a consistent and dramatic decline in genus richness. Future experiments are needed to clarify the shapes of these response curves, and to determine if a threshold concentration exists for the effects of these chemicals on algal community structure. In addition, future experiments with Tergitol NP 10 at concentrations ,5 µg/L will be required to allow more direct comparisons of the ecological effects of these 3 compounds. Environmental Implications of PPCPs. Kolpin et al. (7) have stressed that we unfortunately know very little about the transport, occurrence, and fate of many of the synthetic chemicals that are released into the global environment by VOL. 37, NO. 9, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1717
human activities. These authors provided the first reconnaissance of 95 organic wastewater contaminants in fluvial ecosystems across the continental U.S., and detected these contaminants in widely varying concentrations in 80% of the 139 streams sampled. In this study, we focused upon two of the pharmaceutical and personal care product compounds (PPCPs) detected by Kolpin et al. (Triclosan and Ciprofloxacin), as well as the common surfactant Tergitol NP10, and we have performed one of the first evaluations of their potential environmental impacts at concentrations similar to those observed in receiving waters. Scientists at the Universities of Guelph and Toronto have recently reported significant effects on phytoplankton, zooplankton, duckweed, and sunfish by toxicant mixtures containing Ciprofloxacin (32), but their experiments were performed at concentrations far exceeding those presently found in the environment. The biotic integrity of stream ecosystems is reflected by the condition, abundance, and diversity of its resident biota (19, 20), and algae are particularly sensitive indicators of environmental change. Our algal bioassays revealed that the three PPCPs chemicals studied here can significantly modify algal community structure in vitro, suggesting that there is also strong potential for corresponding effects on the structure and function of natural stream ecosystems that receive WWTP effluents containing these compounds. We thus conclude that in addition to identifying the factors that determine the occurrence and concentrations of organic contaminants in our surface water resources (7), we also should examine much more closely the ecological effects of these contaminants. This study has demonstrated a consistent reduction of algal genus diversity as the concentration of all three PPCPs was experimentally increased. The nature of these relationships should be explored further to determine whether a threshold concentration exists at which no significant ecological effects of these chemicals can be observed. Such analyses will provide valuable guidance for the development of future policies regarding impact control and the possible future need for developing special treatment processes to remove these organic contaminants. However, we note that these studies will necessarily be very complex, because the parent compounds potentially can be transformed into other biologically active compounds via numerous physical and biological processes, including hydrolysis, photodegradation, and biodegradation, and seasonally important factors such as temperature, available light, and dissolved oxygen. The human-excreted residues of many PPCPs, including their breakdown products and conjugates, are transported to receiving waters by WWTP effluents (2). It is unknown how these compounds react when mixed at various concentrations in the receiving waters. Unfortunately, very little is known about the effects of these mixtures on the biotic communities of receiving ecosystems. The breakdown products and conjugates of many PPCPs may have chemical structures similar to those of the parent compounds, and thus may have effects on resident organisms that are sensitive to the parent compound’s mode of action (27). Both Triclosan and its transformation product, methyl-Triclosan, are emitted by wastewater treatment plants, and both have recently been detected in Swiss receiving waters by Lindstrom et al. (28). Direct and indirect effects of PPCPs on aquatic foodwebs are also possible, both for parent compounds and for their degradation products. In addition to the direct effects on algae reported in this study, a vertical cascade of contaminant effects upward from producers to higher trophic levels may also occur (29). Moreover, organic contaminants have been found to have significant direct effects on vertebrates (30, 31) even at low-level concentrations. In higher-level organisms exposed to multiple toxicants, additive effects may also occur, particularly with chemicals that are designed to cross 1718
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 9, 2003
cell membranes (10, 30, 31). The loss of algal taxa even at low toxicant concentrations reaffirms concerns expressed in the literature that subtle effects due to chronic, low level exposures to bioactive PPCPs could lead to cumulative, adverse impacts that might be incorrectly attributed to processes of natural change or ecological succession (2). Further studies on the effects of these and other PPCPs should include work on various concentrations of these compounds, as well as the ecological effects of these compounds when WWTP effluents containing these compounds are mixed into receiving waters.
Acknowledgments We thank Michael Thurman (U.S. Geological Survey) for his advice and kind support, and Norm Slade (University of Kansas) for his technical expertise and his assistance with the statistical analyses of these data. We hank Andrew Dzialowski for reviewing this manuscript, and we also thank Richard Wilson, Elizabeth Davis, and Irene Karel for their help in the field and in the laboratory. This project was supported in part by a Research Development Fund Grant from the University of Kansas, and NSF grant DEB-9816192 to Val H. Smith and M.A. Leibold.
Supporting Information Available Stream site parameters and individual P-values (pdf). This material is available free of charge via the Internet at http:// pubs.acs.org.
Literature Cited (1) Hansen, H.; DeRosa, C. T.; Pohl, H.; Fay, M.; Mumtaz, M. M. Environ. Health Perspect. 1998, 106 (suppl 6), 1271-1280. (2) Daughton, C. G.; Ternes, T. A. Environ. Health Perspect. 1999, 107 (suppl 6), 907-938. (3) Naiman, R. J.; Magnuson, J. J.; McNight, D. M.; Stanford, J. A.; Karr, J. R. Science 1995, 270, 584-586. (4) Postel, S.; Carpenter, S. R. Freshwater ecosystem services. In Nature’s Services: Societal Dependence upon Natural Ecosystems; Daily, G., Ed.; Island Press: Washington, DC, 1997; pp 195214. (5) Jackson, R. B.; Carpenter, S. R.; Dahm, C. N.; McNight, D. M.; Naiman, R. J.; Postel, S. L.; Running, S. W. Ecol. Appl. 2001, 11, 1027-1045. (6) Halling-Sorensen, B.; Nors Nielsen, S.; Lanzky, P. F.; Ingerslev, F.; Holten Lutzhoft, H. C.; Jorgensen, S. E. Chemosphere 1998, 36, 357-393. (7) Kolpin, D. W.; Furlong, E. T.; Meyer, M. T.; Thurman, E. M.; Zaugg, S. D.; Barber, L. B.; Buxton, H. T. Environ. Sci. Technol. 2002, 36, 1202-1211. (8) Folt, C. L.; Chen, C. Y.; Moore, M. V.; Burnaford, J. Limnol. Oceanogr. 1999, 44, 864-877. (9) Raloff, J. Sci. News 1998, 153, 187-198. (10) Magnan, R. L.; Moreno, D. S. J. Econ. Entomol. 2000, 94, 150156. (11) McMurray, L. M.; Oethinger, M.; Levy, S. B. Nature 1998, 394, 531-532. (12) Levy, C. W.; Roujeinikovai, A.; Sedelnikova, S.; Baker, P. J.; Stuitje, A. R.; Slabas, A. R.; Rice, D.; Rafferty, J. B. Nature 1999, 398, 383-384. (13) Vollenweider, R. A. Primary Production in Aquatic Environments; Blackwell Scientific Publications: Oxford, U.K., 1969. (14) Bachmann, R. W.; Canfield, D. E., Jr. Hydrobiologia 1996, 323, 1-8. (15) Prepas, E. E.; Rigler, F. H. Can. J. Fish. Aquat. Sci. 1982, 39, 822-829. (16) Goldsborough, L. G.; Robinson, G. G. C. Hydrobiologia 1986, 139, 177-192. (17) Hoagland, K. D.; Carder, J. P.; Spawn, R. L. Effects of Organic Toxic Substances. In. Algal Ecology; Stevenson, R. J., Bothwell, M. L., Lowe, R. L., Eds.; Academic Press: San Diego, CA, 1996. (18) Wilson, J. G.; Elkaim, B. The Toxicity of Freshwater: Estuarine Bioindicators. In Bioindicators and Environmental Management; Jeffery, D. W., Madden, B., Eds.; Academic Press Limite: San Diego, CA, 1991.
(19) Hill, B. H.; Herlihy, A. T.; Kaufmann, P. R.; Stevenson, R. J.; McCormick, F. H.; Johnson, C. B. J. N. Am. Benthol. Soc. 2000, 19, 50-67. (20) Hill, B. H.; Stevenson, R. J.; Pan, Y.; Herlihy, A. T.; Kaufman, P. R.; Johnson, C. B. J. N. Am. Benthol. Soc. 2001, 20, 299-310. (21) McCormick, P. V.; Cairns, J., Jr. J. Appl. Phycol. 1994, 6, 509526. (22) Smith, W. L.; Chanley, M. H. Culture of Marine Invertebrate Animals; Plenum Press: New York, 1972. (23) Nusch, E. A. Arch. Hydrobiol. Beih. Ergebn. Limnol. 1980, 14, 14-36. (24) Clements, W. H.; Carlisle, D. M.; Laxorchak, J. M.; Johnson, P. C. Ecol. Appl. 2000, 10, 626-638. (25) Koutsolioutsou, A.; Martins, E. A.; White, D. G.; Levy, S. B.; Demple, B. Antimicrob. Agents Chemother. 2001, 45, 38-43. (26) Sulavik, M. C.; Houseweart, C.; Cramer, C.; Jiwani, N.; Murgolo, N.; Greene, J.; DiDomenico, B.; Shaw, K. J.; Miller, G. H.; Hare, R.; Shimer, G. Agents Chemother. 2001, 45, 1126-1136.
(27) Henshcel, K. P.; Wenzel, A.; Diedrich, M.; Fleidner, A. Regul. Toxicol. Pharmacol. 1997, 25, 220-225. (28) Lindstrom, A.; Buerge, I. J.; Poiger, T.; Bergqvist, P.-A.; Mu ¨ ller, M. D.; Buser, H.-R. Environ. Sci. Technol. 2002, 36, 2322-2329. (29) Breitburg, D. L.; Sanders, J. G.; Gilmour, C. C.; Hatfield, C. A.; Osman, R. W.; Riedel, G. F.; Seitzinger, S. P.; Selner, K. G. Limnol. Oceanogr. 1999, 44, 837-863. (30) Sohoni, P.; Tyler, C. R.; Hurd, K.; Caunter, J.; Hetheridge, M.; Williams, T.; Woods, C.; Evans, M.; Toy, R.; Gargas, M.; Sumpter, J. P. Environ. Sci. Technol. 2001, 35, 2917-2925. (31) Harris, C. A.; Santos, E. M.; Janbakhsh, A.; Pottinger, T. G.; Tyler, C. R.; Sumpter, J. P. Environ. Sci. Technol. 2001, 35, 2909-2916. (32) Renner, R. Environ. Sci. Technol. 2002, 36, 268A-269A.
Received for review July 16, 2002. Revised manuscript received January 30, 2003. Accepted February 5, 2003. ES0259741
VOL. 37, NO. 9, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
1719
