A mass balance model of contaminant fate-transport was used to assess the fate of four metals: As, Cd, Cu and Zn, in the Bay of Quinte for hydrologic conditions and loadings in 2000. Results were compared with previous model results of 1988. The model was based on the QWASI (Quantitative Water Air Sediment Interaction) approach and the fugacity/aquivalence concept. The Bay was divided into five geographic segments based on hydrodynamics and chemical loadings. The model identified tributaries and Lake Ontario as the major sources of metal loadings to the Upper and Lower Bays, respectively. Metal concentrations in water decreased by 40–75% in the tributaries between May 1988 and 2000, which resulted in decreased metal concentrations that were, in 2000, all below the Provincial Water Quality Objectives. Measured sediment concentrations exceeded the Lowest Effect Levels (LEL) for all metals at many sites in the Upper and Lower Bays. Using 2000 metal loadings from tributaries and Lake Ontario, the model predicted that sediment concentrations will meet or come within 20% of LELs for all metals in segment 1, As in segments 2 and 3, and Zn in segment 3 within 40 years. Although the model predicted that sediment concentrations would decline to the LEL for all segments within 26 (Zn) to 54 (Cu) years, evidence suggests that benthos are now not impaired by ambient sediment concentrations (excluding “hot spots”). Reducing sediment concentrations faster is expected to be very difficult because their loadings originate from diffuse sources in the watersheds of tributaries and Lake Ontario.

Introduction

Forty-three locations in the Great Lakes that had experienced environmental degradation were identified as Areas of Concern (AOCs) by the International Joint Commission (IJC) in the early 1980s. Remedial Action Plans (RAPs) were developed for each AOC with the aim of restoring impaired beneficial uses. The Bay of Quinte is an AOC located on the northeastern end of Lake Ontario that extends 100 km in length and has an 18000 km2 watershed. The Bay is home to about a quarter of a million residents and is used as a source of water for drinking, agriculture, cooling and industrial processes, and for fisheries and recreational purposes such as swimming and sailing.

Monitoring in the Bay of Quinte in the 1980s raised concerns about elevated concentrations of some metals and persistent organic pollutants (POPs) that exceeded water and sediment quality guidelines (Bay of Quinte RAP Coordinating Committee, 1990). Elevated concentrations of metals were attributed to discharges from historical mining, industrial activities and sewage treatment plant (STP) discharges. The Bay was designated as an AOC in 1983 due to excessive concentrations of nutrients (specifically phosphorus), toxic contaminants resulting in fish consumption advisories, impaired benthic communities, bacterial contamination, and loss of habitats such as wetlands. As part of the RAP's Stage 2 efforts, Diamond and co-workers developed a contaminant fate and transport model reflecting conditions in the Bay in 1988 (Diamond et al., 1994, 1996). The goal of that modeling effort was to improve the understanding of contaminant sources, fate, and persistence in this system for 17 chemicals (4 metals and 13 hydrophobic organic compounds or HOCs).

Since the first modeling effort, ecological changes in the Bay have been marked by the invasion of zebra mussels in 1995 that changed the food web structure, water clarity and, presumably, rates at which contaminants exchange between the sediment and water (Klerks et al., 1996, 1997). The second change since 1990 was the reduction of some chemical loadings due to pollution prevention and remedial actions. In this study we asked whether the remedial actions taken to reduce metal loadings have changed the concentrations of metals in water and sediment of the Bay between 1988 and 2000. We also asked if and when the metal concentrations would decline in sediments below the Lowest Effect Level (LEL) set by the Ontario Ministry of the Environment (OMOE). To accomplish this, we updated the model to reflect hydrology and loading conditions in the Bay for 2000 (Hodge and Diamond, 2002). Most of the model framework and assumptions were kept the same as those of Diamond et al. (1996) in order to compare conditions in 1988 and 2000. This comparison is based on average model predictions for the years 1988 and 2000 and do not reflect long-term dynamic changes in contaminant movement during or beyond this time period. The study included 24 chemicals—4 metals and 20 organic compounds. In this paper, we describe findings for the four metals (As, Cd, Cu, Zn). Model results for the organic compounds are described in the companion paper (Gandhi et al., in press).

Methods

Model framework

Fate and transport model

The contaminant fate and transport models used in this and the previous study were based on the QWASI (Quantitative Water Air Sediment Interaction; Mackay et al., 1983) approach using the fugacity/aquivalence concept of Mackay and Diamond (1989) and Diamond et al. (1992). A mass balance equation was constructed for each compartment (e.g. water, sediments) in each segment of the Bay. In this study, sediment was divided into two vertical layers (surficial and lower sediments), compared to the first model that had one sediment layer. Underlying the lower sediment is buried sediment, which can receive contaminants due to burial. This, as well as water export to Lake Ontario, are the two ultimate removal processes from the system. Air was considered as a compartment with infinite volume and specified concentrations. Other details of the model are provided by Diamond et al. (1994, 1996).

The Bay was divided into five segments (basins) based on hydrodynamics, as discussed by Diamond et al. (1994). The segments are: segment-1 (Upper Bay West), -2 (Upper Bay East), -3 (Hay Bay), -4 (Middle Bay), and -5 (Lower Bay) as shown in Figure 1. The water columns of segments 4 and 5 were divided into two layers, the epilimnion and hypolimnion, to account for thermal stratification. An estuarine-like inflow of Lake Ontario water enters the hypolimnion of segments 4 and 5, and then returns to Lake Ontario via the epilimnetic layers of these two segments. The model assumed that the system was at steady state, i.e. no changes over time, and that the year 2000 contaminant concentrations in water, sediment and biota were supported by the year 2000 loadings. This simplifying assumption was necessary due to a lack of data that prevented from running a time dependent version of the model that would account for the delayed response of sediment concentrations to loading reductions.

Model parameterization

The LogKd values for metals (As: 4, Cd: 5.5, Cu: 4.3, Zn: 5) were taken from Diamond et al. (1994) and were assumed to be similar for all the segments. Dimensions and properties of each segment and particle transport parameters (e.g. sedimentation and resuspension rates) were taken from Diamond et al. (1994) and are summarized in Table SI– 1. The sources of chemical loadings were categorized into atmospheric input via particle deposition, discharges from tributaries and treated effluent from STPs, industrial effluents, direct runoff into the segments, and backflow from Lake Ontario

Atmospheric loadings were assumed to consist of total deposition of particulate metals. Meteorological data and atmospheric concentrations of the metals in the particulate phase were taken from measurements at nearby Point Petre (Table SI– 2) by the Integrated Atmospheric Deposition Network (IADN).

The loadings from all other sources were the product of geometric means of daily measured volumetric flow rates and metal concentrations in each of the above sources. Flow rates of four tributaries and Wilton Creek were taken from Minns and Moore (2004). Metal concentrations in the tributaries, which represented unfiltered total concentrations, were from the Provincial Water Quality Monitoring Network (PWQMN) and Drinking Water Surveillance Program (DWSP) of the OMOE. Data were taken from water monitoring stations closest to the mouths of the tributaries.

Flow rates of STPs and industrial sources were from data collected as part of the Municipal Industrial Strategy for Abatement (MISA) of OMOE. Treated effluent from six STPs entered the Bay as follows: Trenton and CFB Trenton STPs entered segment 1, Napanee, Desoronto, and Belleville STPs entered segment 2, and Picton STP entered segment 4. Metal concentrations in all STP discharges were extrapolated from measured values at Napanee and Desoronto STPs (Todd Harvey, Greater Napanee Utilities, Kingston, ON, unpublished data). Since As and Cd concentrations in the Napanee STP were below the detection limits of 0.001 and 0.01 mg l−1, respectively, we assigned half of the detection limit for As (e.g. 0.0005 mg l−1) and a value of 0.001 mg l−1 for Cd since its detection limit was relatively high. These values were assessed to be reasonable by means of model calibration. Metal loadings from storm water outfalls were not explicitly considered due to a lack of data. Three industrial sources entered segments 1 and 4 (Sonoco and Norampac discharged to the Trent River but were treated as entering segment 1, Essroc Cement entered segment 4). Metal concentrations in discharges from these industrial sources were not available.

Runoff and hydrologic flows from Lake Ontario and backflows from segment-to-segment were taken from Diamond et al. (1994) due to a lack of current measurements. All input loading parameter values for the model are summarized in Table 1.

Results and Discussion

Model evaluation

The model code was verified by comparing results from the updated and previous model for As (Table SI– 3). We selected As due to the relative abundance of measured data and concentrations that were well above detection limits. The comparison showed that both models predicted water and sediment concentrations within 15% (Figure 2). The discrepancy between model results was likely due to the inclusion of the second layer of sediment in the updated model as well as slightly different parameterization of current loading data. The model was next evaluated by comparing measured and modeled water and sediment concentrations (Figure 3) where the measured water concentrations were from DWSP for 2000 (Mojgan Sharifi, OMOE, Toronto, ON, unpublished data) and measured sediment concentrations were from several studies (Environment Canada, 2000, 2001, Hans Biberhofer, Burlington, ON, unpublished data; Milani and Grapentine, 2004; GPEC, 2001; OMOE, Mary Thorburn, Etobicoke, ON, 2000 unpublished data). The measured water concentrations were surface water samples that did not include measurements of the hypolimnion of Middle and Lower Bay (segments 4 and 5). The similarities between measured and modeled concentrations were compared with a 2-tailed t-test.

Modeled As water concentrations in the water of the Bay were within the standard error of measured concentrations (Figure 3). Modeled sediment concentrations underestimated measured values but the difference was not statistically significant. The exception to this difference was the modeled sediment concentration in segment 2 that was significantly lower than measured concentrations (t = 9.805 > tα/2, p < 0.1, df = 54). Modeled concentrations of Cd in the water were comparable to the measured values with the exception of segment 2, for which the model overestimated water concentrations which were less than the detection limit of 50 ng l−1 (DWSP). Cd concentrations in the sediment of the Bay were systematically underestimated by about a half, except for segment 3, for which the concentration was overestimated. The underestimation of Cd sediment concentrations was only significant for segments 1 (t = 3.74 > tα/2, p < 0.1, df = 31) and 2 (t = 3.96 > tα/2, p < 0.1, df = 54) which were the main water bodies that received higher loadings in the past. The modeled Cu concentrations in the water were comparable to the measured values and were within the standard errors of the measured data in segments 1 and 2. Modeled concentrations in sediment were within a factor of 1.4 of measured values and were not significantly different than the measurements except for segment 2 (t = 4.50 > tα/2, p < 0.1, df = 54) and 4 (t = 4.41 > tα/2, p < 0.1, df = 3). Finally, modeled concentrations of Zn in water were within the range of standard errors of measurements and were not significantly different than the measured values in all segments. The modeled concentrations in sediment were generally underestimated; however the difference was only significant in segment 2 (t = 3.23 > tα/2, p < 0.1, df = 54) and segment 4 (t = 6.34 > tα/2, p < 0.1, df = 3).

We attributed the consistent underestimation of metal concentrations in sediment, particularly in segment 2, to the slow response of sediment to present lower metal loadings, i.e. the sediment concentrations have not yet fallen in line with current loadings because of the ∼40 year response time of sediment versus 10 days in the water column (Diamond et al., 1994). Segment 2 receives discharges from the Moira River which has a history of higher metal loadings originating from the Deloro mine site (Crowder et al., 1989; Diamond, 1995; Diamond et al., 1996). Other explanations for the model's underestimation of sediment concentrations were also possible. For example, sediment is usually sampled in zones where fine grain material with higher concentrations accumulates relative to coarse grain material with lower concentrations that are outside of these accumulation zones. In contrast, the model estimated a “global” sediment concentration for fine and coarse grain sediment. As well, the model did not consider macrophytes that may sequester metals.

Comparison between 1988 and 2000 loadings

Metal loadings were compared between May 1988 and May 2000. Further comparison would have been useful but was constrained by data availability. During this time, analytical improvements in metal measurement have lowered detection limits and improved data accuracy, which also made a strict comparison between 1988 and 2000 difficult.

According to the data, metal concentrations decreased by 40–87% in the tributaries between May 1988 and 2000: As concentrations decreased by ∼50% in the Moira River, Cu concentrations decreased by an average of ∼78% in the Trent, 90% in the Moira, and ∼94% in Salmon Rivers, and Zn concentrations decreased by ∼25% in Moira to 60% in the Salmon River (Table 2). Concurrent with decreases in metal concentrations in the tributaries, water discharge was 1.7, 2.0 and 2.7 times higher in May 2000 relative to 1988 for Trent, Moira, and Salmon Rivers, respectively. Although this difference between the years can be attributed to natural variability, water flows in year 2000 were particularly high despite an overall trend to lower flows in the recent years. The combination of lower metal concentrations and higher river discharge rates, translated into slightly higher (∼6%) loadings for Zn, comparable for As and significantly less (∼66%) for Cu from these tributaries for May 2000 vs. May 1988 (Table 2). These decreased loadings varied from only 4% for Zinc in Trent River to 2.5, 5 and 6.2 times for Copper in Trent, Moira, and Salmon Rivers, respectively.

Sediment core profiles can be used to extend this temporal comparison beyond that of the tributary data. The sediment core profile from the early 1970s of Mudroch and Capobianco (1980) clearly showed that As concentrations increased and then decreased over time in segment 2 (Figure SI– 1) which can be related to changes in mining and mineral processing activity along the Moira River at Deloro (Diamond, 1995). Unfortunately the picture was less clear when the 2000 sediment profile from segment 2 was analyzed (Environment Canada, Hans Biberhofer, Burlington, ON, unpublished data). The 2000 sediment profile did not illustrate a clear decrease in recent sediment concentrations that were expected from the earlier core profile. The lack of a clear trend in the 2000 core, despite decreased As loadings from the Moira River, suggests either increased loadings from other sources or, more likely, post-depositional movement of As. This interpretation, however, was based on only two cores which may not be representative of the segment. Similarly to As, Cd concentrations in the sediment of Bay have decreased from 1988 to 2004 (Table SI– 4).

Sources and fate of metals

The model estimated total loadings of As, Cd, Cu and Zn in 2000 of about 16,000, 490, 20,000 and 30,000 kg y−1 to the Bay of Quinte, respectively. The net external sources (not including Lake Ontario) of these metals contributed about 4800, 400, 7750, and 29,800 kg y−1 of As, Cd, Cu and Zn, respectively, to the Bay. Estimated percentage contribution of each source type to the Bay is summarized in Figure 4. The fate of each metal in each segment of the Bay is shown in Figure SI– 2. The model identified tributaries as the major source of metal loadings to the two segments of the Upper Bay. Lake Ontario was the main source of metals to the Middle and Lower Bays by virtue of the large water volume that enters and then exits. Lake Ontario was estimated to contribute 70–80% of As, Cu and Zn to the entire Bay except for Cd, which originated mainly from tributaries. Tributary loadings were estimated at ∼500 and 30,000 kg y−1 for Cd and Zn, respectively. Considering only the tributaries, the Moira River contributed 50% of As (Figure 4), whereas the Trent River contributed 60% and 80% of Cu and Zn, respectively. For Cd, tributaries contributed about 65% of total loadings; the Moira River contributed 30% of just the tributaries loadings. Metal loadings from STPs, industry and atmospheric deposition were negligible.

From 1–10% of all metals were estimated to be lost to sediment burial (Figure SI– 3) and of this, the proportion was highest in the Upper Bay segments 1 and 2 (Figure SI– 2). Conversely, 90–99% of metal loadings were estimated to be exported to Lake Ontario (Figure SI– 3). Among the metals, Zn burial was the greatest. Segment 3 (Hay Bay) had the highest burial rate because it is relatively isolated with less advective flow than other segments of the Bay. Although burial rates were low, metals in the Upper Bay experienced significant sediment-water exchange due to particle deposition and resuspension; diffusive sediment-water exchange and

mixing between two layers of water column was minimal compared to advective export to Lake Ontario (Figure SI– 2).

Residence times

Chemical residence time in the system is defined as the time necessary for 67% of chemical to leave the system if all loadings cease and is calculated as the ratio of the chemical mass in a compartment divided by total removal rate. The removal pathways for metals in the sediment were burial to the deeper sediment layer or resuspension into the water. Diamond et al. (1994) reported that metal residence time in the sediment of the Bay of Quinte was ∼40 years depending on the compounds and the segment. Similarly, the updated model estimated metal residence times of 40–44 years in the upper sediment layer of the top 3 cm (Figure SI-4). In comparison, the metal residence time in water was estimated to vary from 5 to 8 days for Zn and As, respectively, as it is controlled by the large advective flow rates (Figure SI-4).

Comparison with provincial quality guidelines

The issue that prompted this study was the status of contaminants in water and sediment with regard to provincial guidelines and objectives. It is gratifying that all water concentrations measured in the Bay around the year 2000 were below the Provincial Water Quality Objectives (PWQO) for metals.

In response to decreased metal concentrations and hence loadings from tributaries, the average measured water concentrations of Cu in segment 2 decreased from 2000 to 800 ng L−1 from 1988 to 2000 (Diamond et al., 1994; DWSP of OMOE, unpublished data). Over the same time period, Lake Ontario water concentrations remained constant at ∼1.0 μg l−1 (Rossmann and Barres, 1988; DWSP of OMOE, Mojgan Sharifi, Toronto, ON, unpublished data). Measured Zn concentrations in the water of segment 2 did not change significantly between 1988 and 2000.

Data from the 2001 sediment survey of the OMOE showed that metal concentrations in sediment in Upper Bay exceeded the Lowest Effect Level (LEL) for As of 6000 ng g−1 in 9 out of 14 sites, and for Cd, Cu and Zn of 600, 16,000 and 120,000 ng g−1, respectively, in 11 out of 14 sites (Thorburn, 2004). The LEL for metals were established by the OMOE for the protection of benthic organisms. We estimated that reaching the LEL in the Upper Bay would take from 26 years for Zn to over 50 years for Cu. Achieving this reduction would require reducing loadings to segment 2 mainly from Moira River for As and Cd and Trent River for Cu and Zn. The model predicted at the year 2000 metal loadings from tributaries and Lake Ontario, sediment concentrations will meet or come within 20% of LELs for all metals in segment 1, As in segments 2 and 3, and Zn in segment 3 within 40 years. Modeled sediment concentrations were estimated to exceed the LEL for As and Cu in the Lower Bay. Achieving sediment concentrations below the LEL of As and Cu in segments 4 and 5, and Zn in segment 5 will be very challenging if not impossible because loadings originate from Lake Ontario. However, it appears that waiting decades for the sediment concentrations to decrease will not be necessary as recent toxicity tests indicated that Bay of Quinte benthos, at least in locations that are not local “hot spots,” were no longer impaired by the prevailing chemical concentrations in the sediments (Draft report on the Proceedings and Findings, 2009 Bay of Quinte Science and Monitoring Forum, Peter Hodson, Kingston, ON).

Conclusions

The Bay of Quinte on the north shore of Lake Ontario was established as an Area of Concern because of several impairments, which include exceeding the LEL of As, Cd, Cu and Zn in the sediments of the upper part of the Bay. A mass balance previously developed for the Bay of Quinte for the base year 1988, was updated to reflect conditions in 2000 in order to estimate if and when these levels would decrease below the LEL and if loading reductions have translated into lower metal levels in water and sediment in comparison with 1988.

Water concentrations in 2000 in the Upper Bay reflect loadings from the Bay's tributaries in contrast to the Lower Bay that is influenced by backflows from Lake Ontario. The water concentrations of As, Cu, and Zn decreased by 60%, 75%, and 25%, respectively, from 1988 to 2000 in the Moira River which discharges into the Upper Bay. For the same period, Cu and Zn concentrations decreased by 75% and 60% in Salmon River water, respectively. However, because of higher tributary water discharges in May 2000 relative to May 1988 (due to natural variability but data availability constrained us to this comparison), loadings of As and Zn were approximately the same and were lower for Cu between 1988 and 2000. Sediment concentrations were typically overestimated by the model which, we suggest was due to the slow response time of sediment concentrations to these decreased loadings. Whereas the water and sediment of the Upper Bay are strongly influenced by loadings from tributaries, Lake Ontario is the major source of the metals to the Lower Bay and metal concentrations in the lake have not changed significantly. The fate of metals in the Bay was mainly export to Lake Ontario, with only 1–10% of total loading being retained (buried) in the sediment.

The model estimated a residence time of metals in the sediment of ∼40–45 years and as such, current As, Cu and Zn concentrations in the Upper Bay sediment are probably still responding to decreased loadings from the Moira and other rivers over the past decades, as noted above. The model predicted that steady-state sediment levels will meet the LELs for all metals within 26 years (Zn) to >50 years (Cu). Meeting the LEL for the Lower Bay will be very difficult because the concentrations are supported by loadings from Lake Ontario to the Lower Bay. However, since benthos no longer appear to be toxicologically impaired by sediment metal concentrations, the time necessary to reach the LELs does not appear to be necessary for the restoration of benthic community impairment.

Acknowledgements

This project was funded by the OMOE through Lower Trent Region Conservation Authority on behalf of the Bay of Quinte Remedial Action Plan (BQRAP). We thank Barry Jones (BQRAP), Mojgan Sharifi, Aaron Todd, Mary Thorburn, Duncan Boyd, Emily Awad, Shenaz Sunderani, Andreas Radman and Carolyn O’Neill (OMOE), Tom Dann, Pierrette Blanchard, Marilyn Dunnet, Hans Biberhofer, Bill Booty and John Struger (EC), Scott Millard, Marten Koops, Ken Minns and Mike Whittle (DFO), Fred Luckey (USEPA), Jim Moore, Jim Hoyle, John Casselman, Bruce Morrison and Tom Stewart (MNR), and Paul Johanson of the Lower Trent Conservation Authority. Data on atmospheric loads were from the Integrated Atmospheric Deposition Network (IADN). Finally, we thank Catherine Abreu for edits to the previous version of this document.

Supplementary Material

Supplementary tables and figures (labelled SI) for this article are available at http://www.aehms.org/Journal_14_1_Gandhi_Appendix.html

References

Bay of Quinte RAP Coordinating Committee
.
1990
.
Bay of Quinte RAP. Stage 1: Environmental setting and problem definition
Crowder, A., Dushenko, W. T., Greig, J. and Poland, J. S.
1989
.
Metal contamination in sediments and biota of the Bay of Quinte, Lake Ontario, Canada
.
Hydrobiologia
,
188
:
337
343
.
Diamond, M. L.
1995
.
Application of a mass-balance model to assess in-place Arsenic pollution
.
Environ. Sci. Technol.
,
29
:
29
42
.
Diamond, M. L., Mackay, D. and Welbourn, P. M.
1992
.
Models of multimedia partitioning of multispecies chemicals - the fugacity aquivalence approach
.
Chemosphere
,
25
:
1907
1921
.
Diamond, M. L., Poulton, D. J., Mackay, D. and Stride, F. A.
1994
.
Development of a Mass-Balance Model of the Fate of 17 Chemicals in the Bay of Quinte
.
J. Great Lakes Res.
,
20
:
643
666
.
Diamond, M. L., Mackay, D., Poulton, D. J. and Stride, F. A.
1996
.
Assessing chemical behavior and developing remedial actions using a mass balance model of chemical fate in the Bay of Quinte
.
Wat. Res.
,
30
:
405
421
.
GPEC
.
2001
.
Comprehensive environmental assessment of Myers Pier Park, Belleville, Ontario
,
Ottawa, , Canada
:
GPEC International Ltd.
.
Hodge, E. M. and Diamond, M.
2002
.
Updating the existing model of contaminant fate in Bay of Quinte: Benefits, Options and Costs
,
Toronto, Ontario
:
University of Toronto
.
Klerks, P. L., Fraleigh, P. C. and Lawniczak, J. E.
1996
.
Effects of zebra mussels (Dreissena polymorpha) on seston levels and sediment deposition in western Lake Erie
.
Can. J. Fish. Aquat. Sci.
,
53
(
10
):
2284
2291
.
Klerks, P. L., Fraleigh, P. C. and Lawniczak, J. E.
1997
.
Effects of the exotic zebra mussel (Dreissena polymorpha) on metal cycling in Lake Erie
.
Can. J. Fish. Aquat. Sci.
,
54
(
7
):
1630
1638
.
Mackay, D. and Diamond, M.
1989
.
Application of the QWASI (Quantitative Water Air Sediment Interaction) fugacity model to the dynamics of organic and Inorganic chemicals in lakes
.
Chemosphere
,
18
:
1343
1365
.
Mackay, D., Paterson, S. and Joy, M.
1983
.
A quantitative water, air, sediment interaction (QWASI) fugacity model for describing the fate of chemicals in rivers
.
Chemosphere
,
12
(
9–10
):
1193
1208
.
Milani, D. and Grapentine, L. C.
2004
.
The application of BEAST sediment quality guideline to sediments in the Bay of Quinte, An area Of Concern
,
Burlington, , Canada
:
National Water Research Institution
.
Minns, C. K. and Moore, J. E.
2004
.
Modeling Phosphorus Management in the Bay of Quinte, Lake Ontario in the Past, 1972 to 2001, and in the Future
,
Burlington, , Canada
:
Fisheries and Oceans Canada
.
Mudroch, A. and Capobianco, J. A.
1980
.
Impact of past mining activities on aquatic sediments in Moira River basin, Ontario
.
J. Great Lakes Res.
,
6
:
121
128
.
Poulton, D. J.
1992
.
Heavy metals and toxic contaminants in effluents, water, and sediments of the Bay of Quinte, Lake Ontario
.
J. Great Lakes Res.
,
18
:
390
404
.
Rossmann, R. and Barres, J.
1988
.
Trace-Element Concentrations in near-Surface Waters of the Great-Lakes and Methods of Collection, Storage, and Analysis
.
J. Great Lakes Res.
,
14
:
188
204
.
Thorburn, M.
2004
.
2000 Sediment Survey. Ontario
,
Toronto, , Canada
:
Ministry of the Environment
.