Abstract In order to measure organic pollutants and heavy metals in coastal waters. previous studies have directly assayed for these compounds in tissues of bioaccumulators such as Mytilus galloprovincialis. In contrast, this study measured the activity of the detoxification enzyme glutathione S-transferase (GST) in the digestive gland of M. galloprovincialis as a bioindicator of the state of organic pollution at three sites in the San Francisco and Monterey Bays with varying pollutant levels. Glutathione S¬ transferase activity was measured through the conjugation of 1-chloro-2,4-dinitrobenzene (CDNB) to reduced glutathione (GSH) to form CDNB-SH, at a wavelength of 340 nm. The activity data were collected in milliabsorbance units/min and were normalized by dividing by total tissue wet weight. Activity was highest in M. galloprovincialis from the most polluted site, Moss Landing (0.0573 + 0.0226 mAbs/min/g tissue), intermediate in specimens from the Palo Alto site (0.0327 + 0.0140 mAbs/min/g tissue), and lowest in samples from the least polluted site, Monterey Marina (0.0228 + 0.0068 mAbs/min/g tissue, ANÖVA p = 3.8 x 10°). Organisms from Moss Landing also have greater digestive gland masses per total tissue mass and have higher protein concentrations than organisms from less polluted sites. Mytilus galloprovincialis was transplanted from polluted water at Moss Landing into pristine seawater and GST activity was measured over 30 days. There was a significant decrease in GST activity with reduced exposure to organic pollutants. Introduction Anthropogenic marine pollution is a hazard for marine invertebrates inhabiting coastal waters, especially when they live in close proximity to areas of agricultural run¬ off, or inhabit bays where such pollution may collect. In recent decades, the application of pesticides and other organic chemicals to farmlands in California’s agricultural valleys has increasingly affected the coastal oceanic and bay environment. Toxins such as DDT and its breakdown products, polyaromatic hydrocarbons (PAH’s), and polychlorinated biphenyls (PCB's) have been found in the Monterey and San Francisco Bays (SWRCB 1996). Although concentrations of these pollutants in seawater can be measured directly. many of these lipophilic compounds are not very soluble in water, so data analysis can be difficult. For example, contamination can be an issue when concentrations are low, and the methods involved in direct water sampling can be complicated and time-consuming. An alternative method is to examine animal biomarkers. Biomarkers are defined as organisms that tolerate and accumulate into tissues toxins from the environment, There are many characteristics that define good biomarkers: 1) They must amplify as well as accumulate lipophilic compounds. Organic pollutants are often hydrophobic and they enter the food web through phytoplankton, which store energy in lipid droplets. The lipid droplets draw lipophilic pesticides from the surrounding water. Thus, any organism, including marine filter feeders, that gets most of its energy from phytoplankton quickly accumulates pesticides in its tissues if these compounds are present in the water (J. Watanabe, pers. comm.). 2) They are sessile rather than motile, ensuring that they are exposed to contaminants in a limited area. 3) They are transplantable for the purpose of comparison studies. 4) They are naturally abundant and occur over a wide geographical range as well as over a wide range of pollutant concentrations. In the past, water quality monitoring programs have used assay techniques on the tissues of biomarker organisms to determine directly the amount of specific pollutants at a monitoring site. For example, the State of California Mussel Watch program and the San Francisco Estuary Institute use tissues of native or transplanted bivalves to assess the abundance and distribution of several heavy metals and organic toxins along the California coast (SWRCB 1996; SFEI 1999). Another way to use biomarkers to determine relative contaminant levels in the marine environment is to examine an organism’s biochemical response to the toxicity of the contaminant. By observing the activity of enzymes involved in detoxification, a more integrated measure of toxin exposure can be obtained than by assaying directly for individual chemicals. Specific chemical tests may not find all of the toxins present in seawater or biomarker tissue. Thus, by studying a detoxification pathway in a biomarker organism, one can evaluate the effect of all pollutants of a particular type (e.g. organics) on an organism. There are a variety of detoxification pathways that biomarker organisms may use to metabolize the organic compounds common in agricultural run-off. These include the mixed-function oxidase system, which includes the cytochrome p450 group (Livingstone et al. 1989); flavin-containing monooxygenase (Schlenk and Buhrer 1989). multixenobiotic transport proteins (Cornwall et al. 1995), and glutathione S-transferase (Fitzpatrick et al. 1995). The activity of glutathione S-transferase (GST) is an ideal marker for the presence of organic pollutants. Glutathione S-transferase detoxifies xenobiotic compounds by catalyzing the conjugation reaction between reduced glutathione (GSH) and any electrophile. Consequently, GST will link any compound with an electrophilic center to GSH, and the resulting product, a glutathione S-conjugate, will be less reactive and more polar than the original substrate. The glutathione S-conjugate can then be compartmentalized and exported from the cell (Karam 1998). Many organic toxins such as DDT and its derivatives are electrophilic and act as substrates for GST. Thus, examination of GST activity provides an integrated assessment of the exposure of organisms to organic compounds in the environment. Other contaminant types that are currently measured through water sampling or in biomarker tissues may be detected by increased activity from other detoxification pathways. For example, the presence of heavy metals, which do not induce a GST response (Regoli and Principato 1995), may instead increase levels of metallothionein (Bolognesi et al. 1999) An ideal organism for study of GST activity is the bay mussel, Mytilus galloprovincialis. In addition to being a filter-feeding accumulator, M. galloprovincialis occurs over a wide geographical range and can tolerate high concentrations of organic pollutants. The mussel’s digestive gland is responsible for most of the organism’s detoxification of pollutants. Thus, the highest level of GST activity in Mytilus is found in the digestive gland (Regoli 1998; Kaaya et al. 1999). This study compares GST activity in M. galloprovincialis from three sites varying in pollutant levels. The sites studied were chosen based on water quality data from the California Mussel Watch Program (SWRCB 1996). Moss Landing, at the mouth of the Salinas River, is subject to toxin-rich agricultural run-off from the entire Salinas Valley. Mussel Watch has found high levels of many anthropogenic pollutants such as DDT and chlordane in the water at Moss Landing Harbor. Monterey Marina does not receive or collect contaminated run-off. It is also likely to experience more ocean flushing with clean water than Moss Landing Harbor. Bixbee Park, near Palo Alto in South San Francisco Bay is not as polluted as Moss Landing (SWRCB 1995; SFEI 1999). Figure 1 shows the locations of these sites, and Figure 2 provides data on representative pollutant concentrations at each site. Given these toxin concentrations, I hypothesized that M. galloprovincialis from Moss Landing would show more GST activity than mussels from Palo Alto, and that GST activities would be lowest in Monterey Marina. I also expected that mussels transplanted from Moss Landing into less contaminated water in the laboratory would show a decrease in GST activity with time. Materials and methods Collection of specimens Mytilus galloprovincialis were collected from floating docks in three different locations. chosen for their relative concentrations of organic toxins: Monterey Harbor and Moss Landing Harbor in Monterey Bay, and Bixbee Park in the South San Francisco Bay, Mussels were between 4 cm and 6 cm in length. During transport, individuals from each blotted with a kimwipe, and the tissue (not including gonad or byssal threads) was weighed. Preparation of samples Each digestive gland was homogenized in 4 volumes O.05M potassium phosphate buffer, pH 7.4 (Ig tissue 2 1mL), for 15-20 seconds, using an IKA Ultra-Turrax model T tissue homogenizer. One mL of the resulting homogenate was spun in a refrigerated microcentrifuge for 20 min at 4° C and 10,000 g. The supernatant was divided into two parts, one for measurement of protein concentration, which was kept on ice for up to one week, and one for the GST activity assay, which was used immediately, Measurement of GST activity The homogenate sample, to which 0.2mM dithiothreitol and 0.25 ug/mL Pefabloc (a protease inhibitor, Boehringer Mannheim) were added, was kept on ice. The conversion of 1-chloro-2,4-dinitrobenzene (CDNB) and reduced glutathione (GSH) to form oxidized glutathione was measured at 340 nm on a Shimadzu UV-160 1 spectrophotometer, following the method of Lima and Storey (1993). The spectrophotometer was connected to a PC and data were collected using the Kinetics software provided by the manufacturer. To begin, 3 mL cuvettes were filled with 2.0 mL of assay buffer (0.05 M potassium phosphate (pH 7.4), 2mM GSH and 2mM CDNB). A 60-second time course was run at 20 °C to determine the background activity in the absence of GST. To initiate the assay, 50 uL of the homogenate was added to the cuvette and activity was measured for another 60 seconds. GST activity was determined after subtracting the background activity from that of the homogenate. Activity was expressed in milliabsorbance units/minute/g total tissue and was normalized in three ways: with respect to total protein concentration; the whole digestive gland; and total tissue mass of the organism (see results below). Total protein concentration Total protein in the homogenate sample was assayed using the Coomassie Plus Protein Assay (Pierce, Rockford, IL). The homogenates were diluted 20x with distilled water to adjust protein concentration to within the working range of the system. The standard used was bovine serum albumin. Measurement of GST activity: Time Course To test how GST activity levels change as mussels experience a decrease in concentration of organic toxins, fifty individuals were transplanted from Moss Landing Harbor into the flow-through tanks at Hopkins Marine Station. These mussels were acclimated to the local water as described above. A group of individuals was removed from the tanks and GST activity was measured at t = 0 days, 4, 10, 15, and 30 days. Results Estimation of total tissue wet weight The correlation between tissue wet weight and shell length is shown in Figure 3a. The r’ value (0.71) was low, so I included other morphological parameters to create a size index more representative of the volume of a mussel shell, and thus to characterize more accurately the tissue mass. The size index was the product of the three dimensions of each shell, length * width * depth. Using the same group of individuals as before, the relationship between measured total tissue wet weight and size index (Figure 3b) shows a strong correlation (r* = 0.96), thus the equation for the regression line can be used to estimate tissue mass from shell dimensions in cm: Mass (g) = 0.10791*(1*w*d) - 0.32598. This estimation of mass was used to normalize GST activity to total tissue wet weight (see below). Measurement of GST activity: Location Figure 4 shows GST activity normalized to total protein concentration. Activity from mussels at Moss Landing was 0.359 + 0.0923 mAbs/min/mg protein (average + standard deviation; n = 14 from each site); that from Monterey specimens was 0.398 + 0.0848 mAbs/min/mg protein; and that of Palo Alto specimens was the highest, 0.488 + 0.0153 mAbs/min/mg protein. According to single factor ANOVA, there was a statistically significant effect of site (p = 0.0147), and a Tukey-Kramer test indicated the Moss Landing and Palo Alto groups were statistically different (p « 0.05). The data normalized in this way do not support the hypothesis that organisms from Moss Landing, the most polluted site, have higher levels of GST activity. This normalization was based on the convention of expressing enzyme activity in units per mg protein (Regoli and Principato, 1995), which requires the assumption that total protein concentrations in M. galloprovincialis digestive glands vary randomly among individuals regardless of location. Since the normalized GST activity data do not support my hypothesis, 1 decided to examine the assumptions underlying the use of total protein concentration as the normalization factor. If, for example, protein concentration were not independent of site or pollutant concentration, then normalization by total protein concentration would obscure any effect of site on GST activity. To test this possibility, I correlated protein concentration per gram digestive gland with mass of digestive gland for all three sites. The protein concentration is higher in M. galloprovincialis digestive glands from Moss Landing than in those from Monterey and Palo Alto (Figure 5), contradicting the assumption that protein concentration varies independently of location. Because protein concentration appears to be correlated with site and pollution level, it is inappropriate to normalize GST activity with this factor. This figure also suggests that the mass of the digestive gland varies with location. To confirm this observation, I compared the mass of M. galloprovincialis digestive glands to total tissue wet weights. These data are shown in Figure 6, and indicate that for their total tissue wet weight, mussels from Moss Landing have larger digestive glands than mussels from Monterey or Palo Alto. Thus, the data in Figures 5 and 6 illustrate both that the digestive gland comprises a greater percentage of total tissue, and that the total protein concentration in the digestive gland is higher in M. galloprovincialis from a polluted site. To analyze the GST activity in a more meaningful and appropriate way, activity data in mAbs/min/gram digestive gland (from each aliquot tested in the spectrophotometric assay) was multiplied by the mass of the digestive gland of each individual (Figure 7). This manipulation gives activity for each organism—that is, the total amount of GST detoxifying power each individual mussel has in its digestive gland. Glutathione S-transferase activity at Moss Landing (0.170 + 0.101 mAbs/min) was statistically different (by Tukey-Kramer, p « 0.05) from activity levels at Palo Alto (0.072 + 0.034 mAbs/min) and Monterey (0.063 + 0.021 mAbs/min) (ANOVA p = 6.66 x 103 However, Figure 7 does not take into account the effect differing body size on GST activity. To correct for this, the data were normalized to total tissue wet weight, as shown in figure 8. Total tissue wet weight was extrapolated in each case from valve dimensions using the equation noted above. The activity at Moss Landing (0.0573 + 0.0226 mAbs/min/g tissue) was significantly higher than the activities at both Palo Alto (0.0327 + 0.0140 mAbs/min/g tissue) and Monterey (0.0228 + 0.0068 mAbs/min/g tissue) (ANÖVA p = 3.8 x 10“, Tukey-Kramer p 50.05). This analysis, like that shown in Figure 7, supports the hypothesis that GST activity is higher in organisms from a polluted site. Measurement of GST activity: Time Course Figure 9 shows the GST activity data from the 30-day transplantation time-course, normalized to total tissue mass. At t = 0 days, GST activity is 0.0439 + 0.0152 mAbs/min/g tissue and at t = 30 days, GST activity is 0.0315 + 0.0155 mAbs/min/g tissue. There is a significant decrease in GST activity over time (two-sample t-test p = 0.0283), supporting the hypothesis that mussels that experience a decrease in pollutant concentration will show a decrease in GST activity. Discussion The influx of anthropogenic pollutants from industry, agriculture, and other human activities into coastal water has a direct effect on marine life. In this study. I chose to examine an aspect of that effect—one of many biochemical pathways that metabolize those chemicals—as an analytical tool to determine the level of pollution in coastal marine habitats. Currently, testing for both organic chemicals and heavy metals is conducted in two ways: direct water sampling, and measurement of pollutant concentration in bioaccumulator tissue. Most organic pesticides are lipophilic and sparingly soluble in water, thus the direct sampling method sometimes used by regional monitoring programs (SFEI, 1999) is complicated and costly due to the low concentrations of pollutants in the water. Lipophilic compounds accumulate in filter-feeding organisms, so higher concentrations may be found in the tissues of filter-feeders like mussels and clams, and an alternative testing method for organic compounds uses the tissues of these filter-feeders. This method of testing provides specific information on concentrations of individual pollutants accumulated by the organism, but because pollutants may accumulate at different rates (Bjork and Gilek 1997), the relative concentrations of accumulated toxins in tissues may not represent the actual concentrations of those tissues in the surrounding environment. An alternative to these direct methods is to study enzymatic detoxification pathways to gauge organic stress in marine bioaccumulators. There are a variety of enzymes involved in detoxification. In addition to GST, some of these enzymatic pathways, such as cytochrome p450 and multixenobiotic transfer proteins have been considered as biomarkers to study toxic stress on an organism. Most studies on cytochrome p450 have been performed on fish or other vertebrates, although Livingstone et al. (1989) have assayed for cytochrome p450 activity using a fluorescence assay based on ethoxy-o-deethylase. The multixenobiotic transport system has been studied in M. californianus. Cornwall et al. (1995) determined that moderately hydrophobic pesticides such as dacthal and chlorbenside are substrates for the multixenobiotic transport protein, but extremely hydrophobic pesticides such as DDT and its derivatives are not. Thus, studying GST-- which does detoxify these hydrophobic pesticides-in conjunction with multixenobiotic transport protein activity, would provide a more complete illustration of the concentration of most types of toxic pesticides in coastal waters. Furthermore, examining a variety of detoxification pathways in several aquatic organisms linked in the food web would show the course of toxins from phytoplankton into higher trophic levels. The results from three different sites show a significant correlation between GST activity in the digestive gland of M. galloprovincialis and organic pollutant levels. The results from the time-course transplantation experiment indicate that GST activity is dependent on the ambient concentration of organic pollutants. This study uncovered unexpected correlations between the pollutant level of the three sites and both mass of the digestive gland and total protein concentration in the gland. That mussels from Moss Landing, the most polluted site, have larger digestive glands per total tissue wet weight and more total protein per gram digestive gland suggests that these two factors are themselves responsive to pollutant levels. The digestive gland is the primary detoxifying organ in M. galloprovincialis, and it may contain several detoxification pathways that are induced by high pollutant concentrations. Thus, the total protein concentration in the digestive gland would increase with more pollutants in the water, and the digestive gland may itself respond by increasing its size to increase its detoxification power. Thus, a very simple estimate of the pollution in a habitat may be made by examining the relative mass of the digestive gland compared to those of individuals from pristine sites. There are many directions to take in future studies on GST activity in M. galloprovincialis as an indicator of marine pollution. Though this paper suggests a strong correlation between GST activity and marine pollution, more studies would provide evidence for whether organic pollutants are the cause of elevated detoxification enzyme activity. Including other enzyme pathways, as well as expanding the time course experiment as mentioned above would supply more conclusive data for the dependence of enzymatic detoxification on pollutants in the water. Future studies: possible confounding factors There are many factors other than pollutant concentration that may vary among sites, but which this study did not consider. What are the effects of seasonal changes, including climate and run-off, on the mussels in each site? Do GST activity levels vary with the reproductive cycle of each individual? How does food availability vary among the sites? In future studies, I would perform variations of the time course with more controls. In addition to moving M. galloprovincialis from a polluted site into a clean site, 1 would transplant individuals from a clean site to a site higher in pollutant concentration. 1 would also control for seasonal and shorter-term variations by taking parallel measurements of individuals that had not been transplanted. Dosing organisms from a clean site with known concentrations of organic chemicals in the laboratory would show the organism’s detoxification response to only one pollutant at a time. Future studies: other enzyme systems Unlike the other detoxification enzymes mentioned, metallothionein is not induced by organic compounds, but instead is induced by heavy metal pollutants. For a complete look at pollution in the coastal marine environment, metallothionein must be included along with enzymes that catalyze the breakdown of organic pollutants. For example, M. galloprovincialis from Monterey Marina, the location with the lowest concentration of organic pollutants, has the lowest level of GST activity, indicating that these mussels experience relatively low levels of organic pollutants. However, it is thought that this population of mussels is experiencing toxic stress due to copper and tributyltin pollution from anti-fouling paints and other byproducts of heavy boat use in the marina. Mussels from the Monterey Marina exhibit shell morphologies (large shell width to length ratio) that suggest they are not healthy (Mark Stephenson, pers. comm.). and assays of other enzyme systems, which detoxify compounds other than organic pollutants, may be useful in determining the source of this stress. Conclusions GST activity is higher in the digestive glands of M. galloprovincialis from water with higher concentrations of anthropogenic organic pollutants. GST activity does decrease when organisms are transplanted from a site with high concentrations of organic pollutants to a less contaminated site. From these two conclusions, I believe that activity levels of GST and other detoxification enzymes in marine organisms are useful indicators of the general level of marine pollution. In combination with direct sampling methods, this biochemical assay system can provide a comprehensive measure of the concentrations of organic pollutants in certain environments, as well as the ecological implications of extended exposure to these pollutants in the biomarker species, M. galloprovincialis. Acknowledgements 1 would like to thank George Somero, Peter Fields, and Nancy Eufemia for their guidance and ideas. Thank you to Mark Stephenson at Moss Landing Marine Labs. Thank you Caren Braby, Kevin Brown, Gina Kang, Chris Patton, Eric Sanford, the Somero Laboratory, and the Spring 2000 students at HMS. Literature Cited Bayne, B.L. 1978. Mussel watching. Nature 275:87-88. Bjork, M. and M. Gilek. 1997. Bioaccumulation kinetics of PCB 31, 49 and 153 in the blue mussel, Mytilus edulis L as a function of algal food concentration. Aquat. Toxicol. 38:101-123. Bolognesi C., E. Landini, P. Roggieri, R. Fabbri, and A. Viarengo. 1999. Genotoxicity biomarkers in the assessment of heavy metal effects in mussels: experimental studies. Environ. Mol. Mutagenesis 33:287-292. Cornwall, R., B.H. Toomey, S. Bard, C. Bacon, W. Jarman, and D. Epel. 1995. Characterization of multixenobiotic/multidrug transport in the gills of the mussel Mytilus californianus and identification of environmental substrates. Aquat. Toxicol. 31:277-296. Fitzpatrick, P.J., D. Sheehan, and D.R. Livingstone. 1995. Studies on isoenzymes of glutathione S-transferase in the digestive gland of Mytilus galloprovincialis with exposure to pollution. Mar. Environ. Res. 39:241-244 Grelle, C. and M. Descamps. 1998. Heavy metal accumulation by Eisenia fetida and its effects on glutathione-S-transferase activity. Pedobiologia 42:289-297. Habig, W. H. and W. B. Jakoby. 1981. Assays for differentiation of Glutathione S- Transferases. Meth. Enzymol. 77:398-405. Habig, W. H. and W. B. Jakoby. 1981. Glutathione S-Transferases (Rat and Human). 77.218-231. Meth. Enzymol. 11.2 Hermes-Lima, M. and K. B. Storey. 1993. Antioxidant defenses in the tolerance of freezing and anoxia in garter snakes. Am. J. Physiol. 265:R646-R652. Kaaya, A., S. Najimi, D. Ribera, J. F. Narbonne, and A. Moukrim. 1999. Characterization of glutathione S-transferases (GST) activities in Perna perna and Mytilus galloprovincialis used as a biomarker of pollution in the Agadir Marine bay (south of Morocco). Bull. Environ. Contam. Toxicol. 62:623-629. Karam, D. 1998. Glutathione S-transferase: an enzyme for chemical defense in plants. http:/www.colostate.edu/Depts/Entomology/courses/en570/papers 1998/karam.htm Livingstone, D. 1989. Cytochrome P-450 and oxidative metabolism in invertebrates. Biochem. Soc. Trans. 18:15-19. Livingstone, D. 1991. Organic xenobiotic metabolism in invertebrates. Adv. Comp. Environ. Physiol. 7:45-185 Livingstone, D., M. A. Kirchin, and A. Wiseman. 1989. Cytochrome P-450 and oxidative metabolism in molluscs. Xenobiotica 19:1041-1062. Pederson, K. 2000. Characterizing watersheds in mussel hybrid zone: central California Coast. Unpublished work. California State University Monterey Bay. Geographical Information Systems Class, Spring 2000. Regoli, F. 1998. Trace metals and antioxidant enzymes in gills and digestive gland of the mediterranean mussel Mytilus galloprovincialis. Arch. Environ. Contam. Toxicol. 34:48¬ 63. Regoli, F. and G. Principato. 1995. Glutathione, glutathione-dependent and antioxidant enzymes in mussel, Mytilus galloprovincialis, exposed to metals under field and laboratory conditions: implications for the use of biochemical biomarkers. Aquat. Toxicol. 31:143-164. SFEI. 1999. 1997 Annual Report: San Francisco Estuary Regional Monitoring Program for Trace Substances. San Francisco Estuary Institute, Richmond, CA. SWRCB (State Water Resources Control Board). 1996. State Mussel Watch Program. 1993-1995 Data Report. http:/ww.swrcb. ca. gov/programs/smw/index.html Schlenk, D. and D. Buhler. 1989. Xenobiotic biotransformation in the pacific oyster (Crassostrea gigas). Comp. Biochem. Physiol. 94C:469-475. Figure Legend Figure 1. Map of study sites. (Pederson 2000) Figure 2a. Anthropogenic organic pollutants in Mytilus tissue. Locations are PA: Palo Alto, ML: Moss Landing, and MM: Monterey Marina. The y-axis shows concentrations of accumulated organic chemicals in ppb, or ug/L, from O to 2000 (SWRCB 1996; SFEI 1999) Figure 2b. Anthropogenic organic pollutants in Mytlius tissue. Same locations and data as Figure 2a. The scale on the y-axis is increased by a factor of 10 (SWRCB 1996; SFEI 1999) Figure 3a. Total tissue wet weight (g) vs. shell length(mm). Figure 3b. Total tissue wet weight (g) vs. size index (cm’) (see text). Figure 4. GST activity normalized to total protein concentration. Each vertical bar is the average of the activities of 14 individuals from each site. Sites are MM: Monterey Marina, ML: Moss Landing, and PA: Palo Alto. Error bars are standard deviations from the mean. Shared letters (A, B, AB) indicate groups whose means are not significantly different (Tukey-Kramer p 2 0.05). Figure 5. Total protein concentration (mg/mL) vs. mass digestive gland (g). Figure 6. Mass digestive gland (g) vs. total tissue wet weight (g). Figure 7. GST activity normalized to whole digestive gland. Each vertical bar is the average of the activities of 14 individuals from each site. Sites are as above. Error bars are standard deviations from the mean. Shared letters (A, B) indicate groups whose means are not significantly different (Tukey-Kramer p 2 0.05). Figure 8. Activity normalized to total tissue mass. Each vertical bar is the average of the activities of 14 individuals from each site. Sites are as above. Error bars are standard deviations from the mean. Shared letters (A, B) indicate groups whose means are not significantly different (Tukey-Kramer p2 0.05). Figure 9. GST Activity: Time Course. GST activity is normalized to total tissue wet weight. Figure 1: Study Sites (Pederson 2000) Concentration (ppb, dry weight) o8888888. EDEE OZO 9 8 0 0 a Concentration (ppb, dry weight) o88888888 EDEE 93 O 8 8 Total Tissue Wet Weight (no gonad) (g) — R O 8 Total Tissue Wet Weight (no gonad) (g) O O c Figure 4 GST Activity Normalized to Total Protein Concentration MM ML PA AB B A Site Total Protein Concentration (mg/mL) NOONO 2 PE 9: + 5 85 9 + % 8 +2 O Mass digestive gland (g) oaoo 8 3 5 0 X5 8 0.4 0.3 0.2 0.1 0.0 - Figure 7 GST Activity Normalized to Whole Digestive Gland PA MM ML B Site 0.10 0.08 0.06 0.04 0.02 0.00 + Figure 8 Activity Normalized to Total Tissue Mass . PA ML MM Site O 8 8 GST Activity (mAbs/min/g) N0 S 990 90 0 009 00°0 9 o°% * 9 •