JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 13 September 2006
Published in J Environ Qual 35:1884-1893 (2006)
DOI: 10.2134/jeq2005.0122
© 2006 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Dubé, M. G.
Right arrow Articles by Wassenaar, L. I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Dubé, M. G.
Right arrow Articles by Wassenaar, L. I.
Agricola
Right arrow Articles by Dubé, M. G.
Right arrow Articles by Wassenaar, L. I.
Related Collections
Right arrow Industrial Wastes
Right arrow Other Environmental Contamination
Right arrow Water Pollution
Right arrow Isotopes

TECHNICAL REPORTS

Organic Compounds in the Environment

Contrasting Pathways of Assimilation

Stable Isotope Assessment of Fish Exposure to Pulp Mill Effluents

Monique G. Dubéa,*, Glenn A. Benoyb and Leonard I. Wassenaarc

a Univ. of Saskatchewan, Toxicology Centre, 44 Campus Dr., Saskatoon, SK S7N 5B3, Canada
b National Water Research Institute, Environment Canada, Potato Research Centre, 850 Lincoln Rd., Fredericton, NB E3B 4Z7, Canada
c National Water Research Institute, Environment Canada, 11 Innovation Blvd., Saskatoon, SK S7N 3H5, Canada

* Corresponding author (monique.dube{at}usask.ca)

Received for publication April 10, 2005.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Fish are commonly used for monitoring the quality of waters receiving pulp mill effluents (PMEs). Isotopic assays of fish tissues have the potential to provide empirical evidence to link an effluent source to exposure. We show in a 45-d factorial laboratory experiment that different exposure pathways lead to isotopic signatures in fish tissue. Rainbow trout (Oncorhynchus mykiss) were exposed to 10% PME in three ways; direct exposure through addition of PME to aquaria, indirect exposure through invertebrate food consumption (Chironomus tentans cultured in 10% PME), and a combination of both exposure pathways. Of the four stable isotopes measured ({delta}13C, {delta}15N, {delta}34S, {delta}37Cl), {delta}13C, {delta}34S, and {delta}37Cl showed significant differences in exposed animal tissues. {delta}37Cl of fish muscle tissue showed consistent differences across trophic levels and revealed contrasting pathways of PME exposure. Contrasting {delta} 37Cl values in C. tentans due to the presence or absence of 10% PME did not translate into {delta}37Cl differences in fish. Rather, {delta}37Cl ratios of fish muscle tissue were specifically related to 10% PME exposure in the aquaria (waterborne exposure pathway). Feasible distributions of 37Cl source contributions for observed mixture ratios confirmed that PME accounted for observed differences in {delta}37Cl among treatments. Direct uptake of chloride ions across gill surfaces is the most likely pathway for assimilation of PME into fish tissues. Considering the variability of PMEs and receiving environments, use of a multi-isotope approach is recommended for tracing exposure of fish. Use of 37Cl should also be considered in light of its alternative assimilation pathway.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
ORGANISMS IN RECEIVING WATERS may be exposed to a host of anthropogenic stressors due to the impacts of urbanization, agricultural conversion, and discharge of industrial and municipal effluents. The ability to link the effects of exposure to specific stressors is crucial to distinguish between the cumulative, and sometimes confounding impacts of multiple stressors, and to direct mitigation and restoration efforts. This requirement has been recognized in North America with particular reference to the discharge of pulp mill effluents (PMEs) (Walker et al., 2002; Hewitt et al., 2003; Dubè, 2004). Impacts of PMEs on aquatic systems can range from impairment of biochemical processes that may affect rates of primary and secondary production (Dubè et al., 1997; Chambers et al., 2000; Culp et al., 2000) to alteration of fish reproductive behavior that may affect community structure (Owens, 1991; Sepúlveda et al., 2003; Dubè et al., 2005a).

Adult fish are routinely used as sublethal indicators of exposure levels to effluents in aquatic environments, as they are relatively long-lived organisms that have the potential to integrate and reflect receiving water environmental quality (Gibbons and Munkittrick, 1994; Munkittrick et al., 2002). Although biological responses of fish exposed to PMEs can be measured, the confounding effects of multiple sources, fish mobility, and seasonally variable dilution rates in rivers and streams makes it difficult to quantify exposure and link biological effects to specific stressors (Dubè et al., 2002; Dubè, 2004). Since PMEs contain terrestrially derived and processed organic matter, chemical additions such as sodium sulfate, and various by-products of chlorine-based bleaching compounds, it is feasible that PMEs may have distinct isotope compositions compared to receiving environments, and therefore may be traced (i.e., isotope forensics) in aquatic consumers such as benthic macroinvertebrates and fish (Dubé, 2004).

The efficacy of linking exposure to effects using stable isotopes is based on a key assumption: that pulp mill processes (anthropogenic compounds, process chemistry) impose a discernable suite of different elemental isotopic patterns on the natural environment, and that these signals may be measured in various tissues of exposed organisms as a proxy for exposure. Isotopic ratios of C (13C/12C) and S (34S/32S) in aquatic consumers are considered to be indicative of their food source where biological fractionation (or discrimination) associated with the ingestion of prey is minimal (~0 to 1{per thousand}; Fry and Sherr, 1984; Fry, 1991). Variation in {delta}13C values are a function of the relative contribution of terrestrial or exogenous C (i.e., wood pulp) to aquatic food webs (Rounick and Winterbourn, 1986; France, 1995) and variation in {delta}34S values can indicate loadings of sodium sulfate used in the digestion of pulp (Fry, 1989; Wassenaar and Culp, 1996). In contrast, N (15N/14N) isotopic values are positively correlated with trophic position (~3 to 3.5{per thousand}; Vander Zanden and Rasmussen, 1999; Lake et al., 2001). Variation in {delta}15N values also can be used as an indicator of municipal sewage effluent enriched in ammonia and nitrate (Van Dover et al., 1992; Lake et al., 2001) and as a surrogate for the bioaccumulation potential of contaminants (Cabana and Rasmussen, 1994). Chlorine isotopes (37Cl/35Cl) have historically been used as a tracer of chlorinated contaminants (Beneteau et al., 1999; Stewart and Spivak, 2004), and have the potential to serve as a relatively specific tracer of PMEs in aquatic biota. Van Warmerdam et al. (1995) showed that a combination of stable carbon and chlorine isotopes could be used to identify different manufacturers of the same chlorinated solvents and specific solvent types. This enabled Dubé et al. (2005b) to hypothesize that altered {delta}37Cl values in longnose sucker (Catostomus catostomus) downstream of pulp and paper mills in Alberta, Canada were possibly due to the use of chlorine-based bleaching chemicals. Chlorinated organic compounds also may be generated naturally via enzymatic pathways, but these have very negative {delta}37Cl values distinct from both anthropogenic compounds and natural sources (Winterton, 2000; Redd et al., 2002).

If the PME isotopic "signals" are wholly or partially distinct from the unpolluted environment, then stable isotopes may provide an independent means of identifying exposure in sentinel biota. The objectives of this study were 2-fold: (1) to determine which stable isotopes ({delta}13C, {delta}15N, {delta}34S, {delta}37Cl) could most effectively discriminate fish populations according to the presence or absence of 10% PME under controlled exposure and feeding regimes, and (2) to identify the pathways by which PME affects isotopic ratios in fish. Rainbow trout (Oncorhynchus mykiss) were subjected to a two-factor manipulation involving exposure to PME through their environment (aquaria with and without 10% PME) and through their food (prey cultured in aquaria with and without 10% PME). Probabilistic mixing models were used to interpret multi-isotopic patterns (IsoSource; Phillips and Gregg, 2003).


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Fish Exposure Experiment Laboratory Procedures
To test for the effects of PME exposure to fish among trophic pathways, a factorial experiment with two factors was conducted (Fig. 1 ). Hatchery-reared juvenile rainbow trout (RBT) were exposed to a combination of 10% PME and control conditions for a period of 45 d. The experimental design was as follows: (1) RBT held in 100% control water aquaria and fed midge larvae (Chironomus tentans) cultured in 100% control water (CC); (2) RBT held in 100% control water aquaria and fed C. tentans cultured in 10% PME (90% control water) (CE); (3) RBT held in 10% PME (90% control water) aquaria and fed C. tentans cultured in 100% control water (EC), and (4) RBT held in 10% PME (90% control water) aquaria and fed C. tentans cultured in 10% PME (90% control water) (EE) (Fig. 1). There were 16 RBT assigned to each treatment combination for a total of 64 fish.


Figure 1
View larger version (21K):
[in this window]
[in a new window]
  		Fig. 1. Experimental design schematic for the establishment of pulp mill effluent (PME) treatments. Rainbow trout were exposed to 10% PME through the water (treatments EC and EE) and through food ingestion (treatments CE and EE). Chironomus tentans was cultured in 10% PME to alter the composition of food resources. For fish treatment designations of CC, CE, EC, and EE, C = control, E = 10% effluent; first letter refers to water exposure, and second letter to food exposure.

 
The 14-wk-old RBT were collected from a provincial government fish hatchery (Fort Qu'Appelle Hatchery, SK) in early May 2001. Upon arrival at the temperature-controlled lab facility at the National Hydrology Research Centre (Saskatoon, SK), the fish were acclimated for a period of 5 d (120 h) in 75-L tanks by gradually decreasing the proportion of hatchery water and increasing the concentration of the control water, by increasing water temperatures from 9-10°C to 14-16°C, and by altering the photoperiod from 8 h of light and 16 h of dark to 12 h of light and 12 h of dark. At the onset of the experiment, all fish were consuming commercial trout pellet food at 4% body weight, and appeared to be healthy.

Eight experimental fish tanks were set up into four treatment groups, with two tanks per group. Dechlorinated tap water from the Saskatoon municipal water supply was used as the reference water. The PME was obtained from a bleached kraft pulp mill in Prince Albert, SK, that uses 100% chlorine dioxide during the Papricycle bleaching process and produces pulp from hardwood (white and black poplar) and softwood (jack pine, white and black spruce) tree species (Golder Associates, 2000). Before PME discharge to receiving waters, it undergoes secondary treatment by aerated stabilization basins. PME was collected once a week throughout the study. Each week, 300 L of effluent were pumped from the outfall of the secondary treatment ponds and transported to the lab facility in a 380-L high-density polyethylene container where it was stored in a temperature-controlled room at 14 to 16°C. Following the acclimation period, fish were individually removed from the 75-L tanks, weighed (± 0.001 g), measured for standard length (± 1 mm), and randomly allocated into one of the eight 38-L experimental tanks until each tank contained eight fish. All experimental fish were of the same age and size class and weighed approximately 4 to 5 g each.

Treatment solutions were statically renewed and feces and excess food were suctioned out daily. Temperature, pH, and dissolved oxygen (DO) levels were monitored and recorded daily; ammonia (NH3) levels weekly. Throughout the 45-d experiment, these levels were held constant across all treatment levels: temperature = 14.5 ± 0.06°C, pH = 8.8 ± 0.02, DO = 91.3 ± 0.36%, NH3 = 0.31 ± 0.005 mg l–1. Mortalities were recorded as they occurred. Fish were fed four times a week. On three occasions per week they were fed cultured C. tentans (see below: C. tentans culturing methodology) and on the fourth they were fed commercial trout pellet food. During the course of the experiment, the fish were gradually fed an increasing amount of C. tentans, from 4% to 10% mean body weight. Feeding amounts were adjusted for each tank where mortalities occurred.

C. tentans Culturing Methodology
C. tentans used in the experiment as fish food were cultured under controlled lab conditions using a two-stage process involving stock tanks and C. tentans treatment tanks. Temperature, pH, DO (%) and NH3 levels were held constant across all tanks: temperature = 20.1 ± 0.07°C, pH = 8.7 ± 0.0, DO = 85.7 ± 0.9%, NH3 ± 0.1 mg l–1. Stock C. tentans cultures were grown in four 75-L aquaria with a 1-cm layer of washed silica sand (250 to 425 µm) substrate (Unimin Corporation, New Connan, CT, USA) and fed TetraMin slurry (USEPA, 1993). Egg sacs for the stock tanks were purchased from Environmental Consulting & Testing (Superior, WI). Cultures from these tanks were grown to adult stage to produce egg sacs which were then used to inoculate another six 75-L C. tentans treatment tanks: three control water (C) and three 10% PME (90% control water) (E) tanks. The control water and PME was the same as that used in the fish tanks.

One set of C. tentans treatment tanks (C and E) was then inoculated each week with 10 egg sacs of the same age (20 total) from the stock tanks to produce second and third instar C. tentans (approximately 2 wk post-hatch). This procedure was cycled through all sets of C. tentans treatment tanks throughout the experiment to maintain a constant fresh food supply for the fish. Based on fish feeding rates (e.g., 4% body weight per fish per feeding day required approximately 1.4 g wet weight of chironomids per fish per feeding day) the number of chironomid egg sacs required to produce the required weight of chironomid larvae were back-calculated. Based on our previous knowledge culturing estimates of larval survivorship, hatching success, and growth rates were known (Hruska and Dubè, 2005). Sediment core samples were taken from each culturing tank throughout the experiment to estimate larval densities on a per tank basis. Every second day, C. tentans were fed 10 to 30 mL of TetraMin slurry per tank; tanks containing greater numbers of individuals received more food. C. tentans were harvested three times a week and fed to the experimental fish. C. tentans were separated and sorted from substrate, weighed for each tank, and placed into the corresponding fish tank. Fish consumed all food within 2 min.

Morphometric and Physiological Analyses
After a period of 7 wk, the experiment was terminated. Fish were anaesthetized, euthanized via spinal severance, counted, weighed (± 0.001 g), measured for standard length (± 1 mm), and then dissected to remove the liver and other visceral organs. Livers were weighed (± 0.001 g) to enable calculation of a liver somatic index for each fish. Tissues and eviscerated bodies were stored in a –40°C freezer for subsequent multiple stable isotope analysis.

C. tentans were collected from the C. tentans treatment tanks throughout the experiment. Samples were frozen (–40°C), freeze dried, and stored in a dessicator for subsequent stable isotope analyses.

Stable Isotope Analyses
Stable isotope analyses were conducted on freeze-dried PME, control water dissolved inorganic carbon (DIC) and chloride ions (Cl–), C. tentans and commercial trout pellet food sources, and RBT dorsal muscle and liver tissues. Lipids were extracted from biotic samples before 13C, 15N and 34S analyses (Sotiropoulos et al., 2004).

Analysis of stable C ({delta}13C) and N ({delta}15N) isotope ratios was performed by elemental analyzer combustion to CO2 and N2 and measured by continuous-flow isotope ratio mass spectrometry using a GV Instruments Isoprime and Eurovector EA at the National Water Research Institute, Environment Canada (Saskatoon, SK). Sample weights for C and N isotopes ranged from 0.4 mg for food and tissue samples to 1 to 4 mg for PME samples, respectively. Analysis of ({delta}34S) isotopic ratios was performed by continuous-flow isotope ratio mass spectrometry at Isotope Science Laboratory, University of Calgary (Calgary, AB). Sample weights ranged from 1.5 to 2.5 mg (non-lipid extracted), depending on the S content. For Cl ({delta}37Cl) assays of tissues and PME, between 0.3 to 0.9 g (non-lipid extracted) samples were combusted using the ASTM Parr Bomb procedure D808-00 (ASTM, 2000). All Cl in the samples was quantitatively converted to NaCl by this high-pressure oxidation-combustion technique. The resulting NaCl was dissolved in distilled water, filtered, and a 100% AgCl precipitate was obtained. The pure AgCl was converted to MeCl (methyl chloride) for measurement of {delta}37Cl values at the Environmental Isotope Laboratory, University of Waterloo (Waterloo, ON).

All stable isotopic values are expressed using delta notation ({delta}) normalized to the ratio of the sample for a primary standard (Vienna Pee Dee Belemnite for 13C; air for 15N; Canyon Diablo Meteorite for 34S; and Standard Mean Ocean Chloride for 37Cl) in per mille ({per thousand}). Repeatability of internal homogenized working standards was better than ± 0.13, ± 0.15, ± 0.3, and ± 0.1 per mille for {delta}13C, {delta}15N, {delta}34S, and {delta}37Cl, respectively.

Statistical Analyses
Chi-square analysis ({chi}2) was used to test for equality in fish survivorship rates between treatments. Survivorship was calculated as a percentage between the numbers of fish alive at the end of the experiment compared to the start. We did not expect fish morphometric endpoints to be affected by the pathway of exposure for this length of experiment and thus conducted one-way analyses of variance (ANOVA) to test for significant differences in the end weight, end length, liver somatic index (LSI = liver weight/body weight x 100), and condition factor (CFI = body weight/body length3 x 100) for each of the four treatments (CC, CE, EC, EE). Before parametric analysis, Levene's test was used to verify the assumption of homogeneity of variance. Data not meeting this assumption were log10 transformed. Least significant differences (LSD) post-hoc tests were used to compare means. Data that failed to meet parametric assumptions after transformation were analyzed using the Kruskal-Wallis (H) test for significant differences in the nontransformed data. Two-way ANOVAs were conducted for these endpoints first with tank and treatment as factors to ensure there were no tank effects within a treatment. As neither tank effects, nor interactions between tank and treatment occurred, fish were used as the level of replication for our analyses (Zar, 1999). Two-way ANOVAs were also used on fish muscle and liver tissue data to test for significant differences as a function of either water or food exposure to PME (and any interaction therein) for isotopic responses in {delta}13C, {delta}15N, {delta}34S, and {delta}37Cl. Student's t tests (two-tailed) were used to compare isotopic ratios of C. tentans cultured in the C. tentans treatment tanks (C and E treatments). Data failing Levene's test were log10 transformed. Data that still failed to meet parametric test assumptions after transformation were analyzed using Mann-Whitney tests. Statistical significance for all tests was set at 0.05. All statistical analyses were performed using SPSS 11.0 (Chicago, IL).

IsoSource (Phillips and Gregg, 2003) was used to generate distributions of feasible solutions for the source contributions of significant isotopic differences observed in RBT muscle tissues. When the number of isotopic sources exceeds the number of isotopes analyzed by greater than 1 (e.g., 3 or more sources, 1 isotope), unique algebraic solutions for calculating source proportions are inestimable. IsoSource determines all feasible solutions for the contribution of each isotopic source by identifying those combinations of isotopic sources that sum to the observed mixture isotopic ratios. By considering all combinations of each source (0 to 100%) in small increments (e.g., 1%), frequency distributions are constructed.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Not all 64 fish survived until the conclusion of the exposure experiment (Fig. 2 ) and a significant difference in survivorship was obtained across the treatment levels ({chi}2(3) = 1.47). All 16 RBT survived in the CC treatment and 15 RBT survived in the EC treatment. RBT fed C. tentans cultured in 10% PME had lower survivorship; 11 RBT survived in the CE treatment and 9 RBT survived in EE treatment. Over the course of the experiment, when RBT from all treatments were combined and compared from start to finish, fish length increased by ~14% and fish weight increased by ~61% (data not shown). On average, fish measured 70 mm in length and 6.5 g in weight. At the conclusion of the experiment, no significant differences were observed among CC, CE, EC and EE treatment levels for fish weight (ANOVA; F(3, 47) = 0.297, P = 0.827), fish length (ANOVA; F(3, 47) = 0.490, P = 0.691), LSI (ANOVA; F(3, 47) = 2.268, P = 0.093), or CFI (ANOVA; F(3, 47) = 1.418, P = 0.249).


Figure 2
View larger version (15K):
[in this window]
[in a new window]
  		Fig. 2. Survivorship of rainbow trout at the conclusion of the 10% pulp mill effluent exposure food web experiment. Sample sizes are indicated above the survivorship plot. For fish treatment designations of CC, CE, EC, and EE, C = control, E = 10% effluent; first letter refers to water exposure, and second letter to food exposure.

 
Stable isotope values of RBT muscle tissue at the end of the experiment revealed significant differences in {delta}13C and {delta}37Cl, but not {delta}15N or {delta}34S (Fig. 3 ). The {delta}13C was slightly enriched in the EC treatment (ANOVA; F(3, 46) = 2.906, P = 0.045). The {delta}37Cl values were positive for CC and CE treatments and negative for EC and EE treatments (ANOVA; F(3, 33) = 46.748, P < 0.0001). No significant differences were observed for stable isotopes of RBT liver tissues and no significant interaction terms were detected for any of the statistical models (Table 1).


Figure 3
View larger version (24K):
[in this window]
[in a new window]
  		Fig. 3. Stable isotope values of rainbow trout (RBT) muscle and liver tissues at the completion of the experiment averaged over replicates. Sample sizes for each of the treatment bars are as follows: CC (16), CE (11), EC (15), and EE (9). Stable isotope values for Chironomus tentans are based samples of whole organisms collected from the C. tentans culturing tanks. Error bars are standard errors. For fish treatment designations of CC, CE, EC, and EE, C = control, E = 10% effluent; first letter refers to water exposure, and second letter to food exposure.

 

View this table:
[in this window]
[in a new window]
  		Table 1. Results of two-way ANOVAs for rainbow trout (RBT) muscle and liver tissues. The degrees of freedom (df) for each of the water exposure, food exposure and interaction terms is 1. Significant terms are in italic.

 
Stable isotopes of C. tentans from the two sets of C. tentans treatment tanks revealed significant differences in {delta}34S and {delta}37Cl, but not {delta}13C or {delta}15N. The {delta}34S values were lower in control water (C) tanks than 10% PME (E) tanks (t(9) = –2.789, P = 0.021), which is consistent with higher values of {delta}34S in PME (13.3 {per thousand}) than in all other potentially contributing sources (trout pellets = 6.2 {per thousand}, TetraMin slurry = 7.9, control water = 0.1 {per thousand}). The {delta}37Cl values were more negative in E tanks than C tanks (t(7) = 6.199, P = 0.001), which is consistent with the lowest (most negative) values of {delta}37Cl in PME (–1.69 {per thousand}) than in all other potentially contributing sources (trout pellets = 0.24 {per thousand}, TetraMin slurry = –0.03 {per thousand}, control water = 0.61 {per thousand}).

Figure 4a shows the {delta}37Cl signatures across all treatments (CC, CE, EC and EE) and for all sources which could potentially contribute to {delta}37Cl signatures in RBT muscle tissue and C. tentans. It is apparent that {delta}37Cl values of RBT muscle tissues (black box) were unrelated to ratios of {delta}37Cl in C. tentans (open triangle) (Fig. 4a). Chlorine isotopic ratios differed according to the presence or absence of PME (open circle) in the aquaria (i.e., EC and EE aquaria more negative in {delta}37Cl than CC and CE aquaria). If {delta}37Cl signatures in RBT muscle tissue were related to the culturing environment of C. tentans, it would have been reflected as a similar pattern as shown for the CE and EE treatments (Fig. 4a).


Figure 4
View larger version (16K):
[in this window]
[in a new window]
  		Fig. 4. a) Chlorine isotope values of rainbow trout (RBT) muscle tissue and all contributing isotopic sources. Error bars are standard errors. b) Probability distributions for feasible solutions of isotope source proportions that contribute to observed {delta}37Cl values in rainbow trout muscle tissues (IsoSource model output; Phillips and Gregg 2003). Histograms for each source contribution have been replaced by curves to facilitate interpretation. PME = 10% pulp mill effluent. For fish treatment designations of CC, CE, EC, and EE, C = control, E = 10% effluent; first letter refers to water exposure, and second letter to food exposure.

 
Probability distributions of contributing isotopic sources for observed RBT muscle tissue {delta}37Cl values are shown in Fig. 4b (PME = –1.69 {per thousand}, trout pellets = 0.24 {per thousand}, TetraMin slurry = –0.03 {per thousand}, control water = 0.61 {per thousand}). For the CC and CE treatments (3 sources: control water, trout pellets and C. tentans), the overall shapes of the distributions were similar for all sources and C. tentans contributed the least of the three sources to the RBT muscle tissue {delta}37Cl as evidenced by the C. tentans ranges (~0 to 30%). Of the isotopic sources for the EC treatment (4 sources: control water, trout pellets, C. tentans, PME), the contribution of 10% PME was highest, with a distribution that ranged from ~30 to 75%. All four isotopic source frequency distributions for the EE treatment were similar.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
In waters affected by multiple effluent discharges, it is difficult to attribute changes in aquatic biota to a specific source in the absence of a reliable tracer. Identification of the causal source is necessary to shape policies intended to encourage mitigation and restorative efforts. Stable isotopes of 13C, 15N, and 34S have been used to assess how effluent discharges affect food web structure (Wayland and Hobson, 2001; Wassenaar and Culp, 1996) and to trace biotic exposure (Dubé et al., 2005b). However, due to variability of different aquatic environments as well as in the composition of PME as a complex mixture, no single stable isotope has shown consistent potential to serve as a tracer of exposure. Wassenaar and Culp (1996) measured 13C, 15N, and 34S in the Thompson River food web exposed to both PME and a municipal sewage effluent. They reported that although PME and municipal sewage were isotopically distinct in 34S from the river, this isotope was not a reliable tracer of exposure due to hydrological mixing patterns between two rivers with different signatures. Analyses of 15N were confounded by the presence of an unmeasured terrestrial source of N. However, 13C was a good tracer for following the fate of PME-derived carbon into the riverine food chain. Wayland and Hobson (2001) reported that 15N and 34S (but not 13C) were somewhat effective for tracing sewage and pulp-mill effluent in two river systems, their usefulness dependent on the extent of effluent dilution in the river coupled with the magnitude of the difference between the isotope signatures in the effluent and in the river. Dubé et al. (2005b) measured 13C, 15N, 34S, and 37Cl in the Athabasca River downstream of two pulp mills and a sewage treatment discharge in an effort to trace exposure of biota to the different effluent streams. Results showed that 15N, 34S and 37Cl were more effective at classifying fish tissue among sites than isotopes of 13C although the usefulness of the specific isotopes depended on the discharge (e.g., 15N was most effective for delineating sewage inputs).

Theoretically, 13C, 34S, and 37Cl all have potential to characterize different constituents of a complex PME mixture as mills produce effluent from pulping tree fiber (source of exogenous C), cooking the fiber with sulfate-based chemicals, and often bleaching the pulp with chlorine-based chemicals. The isotope 15N is more commonly used to trace exposure to sewage effluents (Wayland and Hobson, 2001). Dubé et al. (University of Saskatchewan, Toxicology Centre, unpublished data) conducted stable isotope analyses on 11 different treated pulp mill effluents across Canada. The variability in isotopic composition of the different effluents was high with 13C ranging from –25.5 to –14.7{per thousand}, 15N ranging from –4.0 to 2.6{per thousand}, 34S ranging from –3.7 to 16.9{per thousand}, and 37Cl ranging from –1.7 to 0.8 {per thousand}. This suggested that each effluent had the potential to have a unique isotopic signature and the tracer selected to measure exposure of aquatic biota would likely depend on the effluent being assessed. Based on evidence from field studies, differences in the isotopic composition of receiving waters, and the complexity and composition of PME mixtures, development of tracers for exposure should proceed using a multi-isotope approach. Furthermore, as stated by Wayland and Hobson (2001) and Gannes et al. (1997), use of controlled exposure studies should be considered to better delineate the usefulness of stable isotopes as tracers of effluent exposure in food webs.

The objective of our experiment was to determine if fish exposed to an environmentally relevant concentration of PME (i.e., 10%) under controlled exposure conditions would result in multi-isotopic signatures distinct from the control fish. Of the four stable isotopes used for this experiment (13C, 15N, 34S and 37Cl), only {delta}37Cl yielded consistent differences across species according to the PME treatments in both C. tentans and RBT muscle tissue over the time frame of the study. Additional significant differences in RBT {delta}13C ratios and C. tentans {delta}34S ratios were observed after effluent exposure with enrichment of 13C measured in RBT muscle in the EC treatment (–16.9{per thousand}) compared to controls (–17.1{per thousand}). However, this change was marginally significant (p = 0.045) and likely not due to the PME with a 13C signature of –22.1{per thousand}. Enrichment of {delta}34S by 1.22 {per thousand} was measured in C. tentans tissues after exposure to effluent with a {delta}34S signature of 13.3 {per thousand}. Wayland and Hobson (2001) also reported a consistent pattern of {delta}34S enrichment in food chain components from upstream to downstream for two pulp mills studied. In our laboratory studies, 15N signatures did not differ between control and effluent-exposed fish and 13C and {delta}34S did not show consistent changes across trophic levels. It might be argued that our exposure duration (45 d) was insufficient for fish growth and thus for isotopic uptake into fish tissues. However, fish in all of our treatments increased in body weight by ~61% over the experimental period suggesting that this length of exposure was sufficient. The 15N signatures in our effluent measured –2.8{per thousand}, which was similar to that measured in PME (–4.4{per thousand}) by Kukkonen et al. (1996). Wayland and Hobson (2001) measured 15N (0.4 {per thousand}) in the same PME as used in our study; they also did not report changes in 15N signatures in sediments or aquatic biota measured downstream of the PME discharge.

Ingestion of C. tentans cultured in 10% PME did not translate into altered {delta}37Cl values in RBT muscle tissue. However, {delta}37Cl values in RBT seem to be a function of the presence or absence of 10% PME in the aquaria, indicating that the pathway of {delta}37Cl accumulation into muscle tissue was not through dietary assimilation of prey digested in the gut, as for C, N and S (Peterson and Fry, 1987; Gannes et al., 1998). This finding was a direct result of the factorial design used in this experiment to test for differences in modes of uptake among isotopes. Significant differences in {delta}37Cl between only the CC and EE treatments yielded the same conclusion; PME alone is responsible for more negative chlorine isotopic ratios in RBT muscle tissue. But it was the combination of PME treatments that revealed how the uptake pathway of 37Cl differs from 13C, 15N and 34S. The most plausible physiological mechanism by which 37Cl is incorporated into fish tissues is through ionic regulation of chloride uptake (Cl–) across gill surfaces (Perry, 1997). Hence, the adage "you are what you eat, plus a few per mille" does not necessarily apply as a global model for all stable isotopes in organismal biology. Similarly, uptake of 37Cl in C. tentans was likely due to ion transport across anal papillae into the hemocoel (Jarial, 1995). Recognition of this alternative pathway of assimilation of 37Cl isotopes into biota will enable enhanced understanding of how this isotope might be used as a tracer of exposure in sentinel species (e.g., Dubé et al., 2005b).

The intention of this experiment was to test for isotopic uptake in RBT tissues under sublethal PME concentrations and controlled exposure conditions. In our study, exposure to PME did not affect fish length, weight, condition factor, or liversomatic index. Neither the presence of 10% PME in the experimental fish aquaria or C. tentans cultured in 10% PME and fed to the fish, translated into effects on fish morphometry or physiological condition. However, RBT survivorship, an acute indicator, unexpectedly differed among treatments; fish exposed to 10% PME in their aquaria and fed C. tentans cultured in 10% PME had a survivorship rate just over 50% (9 of 16), while the survivorship rate for fish held in control water and fed only C. tentans cultured in control water was 100% (16 of 16). Intermediate rates of survivorship were ~69% for fish held in control water and fed C. tentans cultured in 10% PME and ~94% for fish held in 10% PME and fed C. tentans cultured in only control water. These results suggest that consumption of food cultured in 10% PME could have been a dominant pathway affecting survivorship of juvenile RBT although waterborne exposure to effluent also had a negative, albeit apparently lesser, influence. Pulp mill effluents have been shown to alter sublethal indicators in wild fish from Canadian (Munkittrick et al., 1991, 1992; McMaster et al., 1995; Dubé, 2004; Dubé et al., 2005a), Scandinavian (Andersson et al., 1988; Sandstrom et al., 1988), New Zealand (Van Den Heuvel et al., 2004), and United States (Thomas and Hall, 2004; Haley and Hall, 2000) waters. Effects observed include low fish biomass, reduced recruitment, decreased gonad size, and increased liver size. Extensive efforts have been invested to document these responses and to understand physiological and reproductive processes affected by effluent exposure (Kovacs et al., 1997). Few studies, however, have investigated how different pathways of PME exposure affect fish responses. Although low statistical power may account for the observed survivorship results, further work is required to investigate how different pathways of PME exposure affect fish survivorship.

At low concentrations of PME, often benthic invertebrate abundance and fish growth increase (nutrient enrichment response). Other components of aquatic food webs also experience greater rates of production in the presence of low levels of PME, including periphyton (Dubé and Culp, 1996; Dubé et al., 1997). Fixed quantities of food (i.e., C. tentans and trout pellets) were added to each aquarium according to fish biomass. Further, on a daily basis, all aquaria were drawn to 10% volume, cleaned and replenished with new control water and, in EC and EE aquaria, 10% PME. This effectively prevented establishment of periphyton that may have supported supplementary food sources for either RBT directly, or C. tentans indirectly. Therefore, although this laboratory experiment effectively isolated the mechanism underlying 37Cl patterns in fish tissue, it did so at the expense of simulating variation in trophic status as a function of PME subsidization. Such a relevant factor ought to be included in follow-up laboratory studies as a covariate of PME concentrations, since these types of concomitant interactions may affect processes that determine stable isotope ratios in fish and other consumers that are presumed to function solely as tracers of effluent exposure.

An important goal of multi-isotopic assays is to determine the relative contributions of isotopic sources to a particular mixture. Based on mass balance equations of linear mixing models, this approach has been limited to analyses where the number of isotopic sources exceeds the number of isotopes by only 1 (Phillips, 2001). In this experiment, of the four isotopes measured, only 37Cl had significant differences for all fish treatment levels. But the number of sources that contributed to the observed {delta}37Cl values in RBT muscle tissues was either 3 (CC and CE treatments) or 4 (EC and EE treatments), which prevented the determination of unique algebraic solutions. An alternative approach to quantifying the proportional contributions of specific sources is to construct distributions of feasible solutions based on combinations of sources that sum to mixture isotopic ratios (Phillips and Gregg, 2003). The overlaid distributions for the EC treatment clearly revealed that compared to all other sources, 10% PME in the control water was the primary driver of {delta}37Cl in RBT muscle tissue. Further, the peaked distributions in the CC (0 to ~35%) and CE (0 to ~25%) treatments indicates that incorporation of 37Cl into RBT muscle tissues was least likely via ingestion of C. tentans. In combination with plots of observed isotopic ratios, inclusion of distributions of feasible solutions offers an informative approximation of source contributions.

Based on the field literature to date, the complex composition of PMEs, and the range of isotopic signatures across different PMEs, it is clear that a multi-isotope approach provides a stronger weight of evidence to trace biotic exposure than use of single isotopes. This conclusion was also reached by Wassenaar and Culp (1996) and Wayland and Hobson (2001). With continuous improvements in mass spectrometric technologies, additional stable isotopes such as 37Cl may be routinely incorporated in multi-isotopic assays (Wassenaar and Koehler, 2004). While multi-isotopic assays represent an opportunity for tracing sources of pollutants and effluents in receiving waters and associated aquatic biota (Wassenaar and Culp, 1996; Wayland and Hobson, 2001; Dubé et al., 2005b), the interpretation of multi-isotopic patterns in the environment are challenged by our understanding of how biogeochemical, ecological, and physiological factors interact to yield observed isotopic patterns (Griffiths, 1998). Carefully designed laboratory experiments represent only one approach for testing putative mechanisms that may account for empirical field patterns (Soimasuo et al., 1995). Although laboratory studies do not replicate field conditions and the complex ecological processes therein, they can provide controlled and replicated exposure to effluents to specifically assess tracer uptake into tissues. Our laboratory studies showed that 37Cl is a tracer of effluent exposure and assimilation from PME is through biochemical processes (i.e., ion regulation via gills in fish and possibly anal papillae in C. tentans) and not by ingestion through food web dietary interactions. This result will assist with interpretation of field-based data. Further work is required to determine the length of exposure required for incorporation of 37Cl signatures into fish tissue and if signatures depurate in the presence of clean water. Further work is also required to determine if effects of PME on fish are correlated with uptake of isotopic signatures of 37Cl into fish tissues. This is an important point. The 37Cl results from this study suggest that PME signatures can be picked up through the water and not necessarily through the food source. However, survivorship data show that acute lethality might be correlated with consumption of prey cultured in 10% PME. These results might suggest that 37Cl has potential as an early indicator of PME effects. However, depending on the fish response indicators measured, it might also lead to Type II statistical errors where exposure is confirmed, yet a conclusion of no effect is made.

This food web experiment tested for the effects of one type of PME at one concentration of exposure. The bleach kraft mill that participated in this study uses 100% chlorine dioxide as a bleaching chemical. Pulp mills across North America use similar industrial processes, suggesting that application of stable isotope analysis of 37Cl, as a tracer of PME, could potentially have widespread application. Many receiving waters, especially rivers, are affected by discharges from several sources. These effluents are frequently affected by variable flow regimes and dilution. Because {delta}37Cl values appear to be unaffected by food web processes, they may function as a conservative tracer of exposure. When combined with other chemical and biological tracers, such as 13C and chlorophenols, the potential for "forensic fingerprinting" with isotopes exists (Kukkonen et al., 1996). Van Warmerdam et al. (1995) and Jendrzejewski et al. (2001) found that 13C and 37Cl clearly discriminated among chlorinated organic solvents. Although removal of elemental chlorine in the bleaching process at kraft mills around the globe significantly reduced the discharge of highly toxic and highly substituted chlorinated organic compounds (e.g., dioxins and furans), the substitution of chlorine with chlorine dioxide in bleaching still produces chlorine-based compounds resulting in a high applicability for 37Cl as a PME tracer. However, it remains to be seen if {delta}37Cl values from PME are able to discriminate among specific pulp mills and specific milling processes as effectively as among organic solvents.


   ACKNOWLEDGMENTS
 
Funding for this research was provided by the Prairie, Northern (S. Blenkinsopp), and Ontario (N. Ali) Regions of Environment Canada, and the National Water Research Institute of Environment Canada (Environmental Effects Monitoring Program, K. Hedley). We thank A. Martin, M. Benjamin, J. Inkster, and V. Tumber for assistance conducting the exposure experiments and G. Koehler for assistance with isotope measurements. We acknowledge the cooperation of the bleached kraft pulp mill in Prince Albert, SK, the University of Calgary (S. Taylor), and the University of Waterloo (R. Drimmie).


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Dubé, M. G.
Right arrow Articles by Wassenaar, L. I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Dubé, M. G.
Right arrow Articles by Wassenaar, L. I.
Agricola
Right arrow Articles by Dubé, M. G.
Right arrow Articles by Wassenaar, L. I.
Related Collections
Right arrow Industrial Wastes
Right arrow Other Environmental Contamination
Right arrow Water Pollution
Right arrow Isotopes

HOME HELP