J. Great Lakes Res. 28(3):437–450 Internat. Assoc. Great Lakes Res., 2002
Surficial Sediment Contamination in Lakes Erie and Ontario: A Comparative Analysis Christopher H. Marvin1,*, Murray N. Charlton1, Eric J. Reiner2, Terry Kolic2, Karen MacPherson2, Gary A. Stern3, Eric Braekevelt3, J.F. Estenik4, Lina Thiessen,1 and Scott Painter1 1Environment
Canada 867 Lakeshore Road, PO Box 5050 Burlington, Ontario, Canada L7R 4A6 2Ontario
Ministry of the Environment 125 Resources Road Toronto, Ontario, Canada M9P 3V6
3Freshwater
Institute, Department of Fisheries and Oceans 501 University Crescent Winnipeg, Manitoba, Canada R3T 2N6
4Ohio
Environmental Protection Agency Lazarus Government Center, P.O. Box 1049 Columbus, Ohio, USA 43216-1049 ABSTRACT. Sediment surveys were conducted in Lakes Erie and Ontario to characterize spatial trends in contamination, to assist in elucidation of possible sources of contamination, and for identification of areas where contamination exceeded Canadian sediment quality guidelines for protection of aquatic biota. Sediment levels of metals including nickel, lead, zinc, chromium, and copper were compared to pre-colonial concentrations, and sediment enrichment factors, defined as the ratio of surficial concentrations to background concentrations determined from benthos cores, were calculated. Sediments in Lake Ontario exhibited elevated contamination compared to Lake Erie. The average enrichment factor for Lake Ontario (2.6) was comparable to the western basin in Lake Erie but greater than those for the central (1.3) and eastern (1.0) basins. There was a gradient toward decreasing sediment contamination from the western basin to the eastern basin of Lake Erie, and from the southern to the northern area of the central basin. Sediment contamination in Lake Ontario was similarly distributed across the three major depositional basins. The spatial distribution of metals was similar to those of other contaminants including mercury, polychlorinated biphenyls (PCBs), and polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDDs/PCDFs). Lake-wide averages of sediment mercury, PCBs and PCDDs/PCDFs in Lake Erie were 0.185 µg/g, 96.5 ng/g, and 18.8 pg/g TEQs, respectively. Lake-wide averages of sediment mercury, PCBs and PCDDs/PCDFs in Lake Ontario were 0.586 µg/g, 100 ng/g, and 101 pg/g TEQs, respectively. INDEX WORDS: Lake Erie, Lake Ontario, polychlorinated biphenyls, polychlorinated dibenzo-p-dioxins, mercury, metals, sediment.
INTRODUCTION The presence of persistent pollutants can adversely impact Great Lakes wildlife, biodiversity, and aquatic ecosystems. Environment Canada, to-
*Corresponding
gether with collaborating agencies, conducts Great Lakes sediment surveys to measure the occurrence and spatial distribution of toxic substances, to further an understanding of the role human activities play in releasing these compounds to the environment, and to provide information for devising effective strategies to mitigate deleterious health effects.
author. E-mail: *
[email protected]
437
438
Marvin et al.
In 1997, a Lake Erie sediment survey was conducted for assessment of contemporary sediment contamination relative to sediment quality, as well for comparison with contaminant concentrations determined from a 1971 survey (Painter et al. 2001). The observed spatial trends in sediment contamination in Lake Erie were similar for a number of compound classes including total polychlorinated biphenyls (PCBs), polychlorinated dibenzo-pdioxins and dibenzofurans (PCDDs/PCDFs), total mercury, and metals. There was a decreasing trend in contamination from the western basin to the eastern basin, and from the southern to the northern portion of the central basin. In addition, concentrations of contaminants in Lake Erie sediments were found to have decreased significantly over the period 1971 to 1997. In 1998, a survey was conducted to assess the spatial distribution of sediment contamination in Lake Ontario. The Lake Ontario and Lake Erie surveys were conducted using common sampling and analytical methodologies, thereby offering an opportunity to assess trends in sediment contamination across the entire range of the lower Great Lakes. In this paper, a brief overview of trends in sediment contamination in the lower Great Lakes in areas ranging from the mouth of the Detroit River to the head of the St. Lawrence River is presented for selected contaminants including PCBs, mercury, and PCDDs/PCDFs. In addition, averaged trends in sediment contamination are presented by selected metals, expressed as “enrichment ratios,” which are defined as the ratio of surficial sediment concentrations to background concentrations determined from the bottom sections of benthos core samples. The data are also presented within the context of the Canadian Sediment Quality Guidelines (CCME 1999), which can be used as screening tools in the assessment of potential risk, and in the determination of the relative priority of sediment quality concerns. METHODS Sample Collection Surficial sediment samples were collected aboard the CCGS Limnos using a mini box core sampling procedure in 1997 and 1998 from sixty-three stations in Lake Erie and seventy locations in Lake Ontario. All samples collected from both surveys consisted of fine-grained sediments classified as glacio-lacustrine clay, sand, silt, or mud. The top 3 cm of the sediment was sub-sampled for analyses of
persistent organic pollutants, metals, particle size, and nutrients. At three index stations in Lake Erie and one index station in Lake Ontario, mini box core and benthos gravity core samples were collected to obtain both surficial and depth-integrated sediment samples. The top 3 cm were sampled from the mini box cores for characterization of surficial sediments. Background concentrations were determined from benthos cores sub-sampled in 1-cm intervals from 0–1 cm, 9–10 cm, 14–15 cm, 19–20 cm, 29–30 cm, and every 10 cm thereafter (i.e., 39–40 cm, 49–50 cm) to the bottom of the core. A box core for study of the accumulation of PCDDs/ PCDFs was sub-sampled in 1-cm increments from the surface to 16 cm, and every 2 cm thereafter to the bottom of the core. Samples for organic contaminant analyses were collected in pre-washed glass jars. Samples for other characterizations were collected in either high-density polypropylene or Teflon jars. All samples were frozen immediately for transport to the laboratory. Analyses Dry sediment samples (5 to 15 g) were extracted in dichloromethane using an Accelerated Solvent Extractor (ASE, Dionex Inc.). Solvent extracts for Lake Erie samples were subjected to an opencolumn fully-deactivated Florisil (National Laboratory for Environmental Testing 1997) cleanup procedure resulting in two fractions for analysis. Fractions were treated with mercury to remove sulphur. Solvent extracts for Lake Ontario samples were separated into three fractions on deactivated Florisil (1.2% v/w water). These fractions were treated with activated copper powder to remove elemental sulphur. Analyses for Lake Erie samples were carried out on a Hewlett-Packard Model 5890 gas chromatograph (GC) with dual columns (30 m 0.25 mm i.d. 0.25 µm stationary phase DB-5 and 30 m 0.25 mm i.d. 0.25 µm stationary phase HP-50) and dual 63Ni electron capture detectors (ECD); analyses for Lake Ontario samples were carried out using a Varian 3600 GC with a 60 m 0.25 mm i.d. 0.25 µm stationary phase DB-5 column with ECD. H2 was used as the carrier gas at 1 mL/min. Total PCB measurements for Lake Erie samples were based on the sum total of congeners quantitated using a 121-congener external standard from the National Laboratory for Environmental Testing, Environment Canada. Total PCB measurements for Lake Ontario samples were based on the sum total of 103 congeners quantitated using external stan-
Surficial Sediment Contamination in Lakes Erie and Ontario dards from Ultra Scientific. Procedural method blanks and standard reference materials (SRMs) were processed with each set of 10 samples. Surrogate standards applied to sediment prior to extraction included PCB #30, PCB #204, octachloronaphthalene, 1,3,5-tribromobenzene, 1,2,4,5tetrabromobenzene and δ-BHC. Recovery criteria for surrogates were 70%; measurements of SRMs were required to be within 30% of certified values. Procedural blanks and surrogate recoveries were also evaluated against criteria established by the National Laboratory for Environmental Testing resulting from replicate analyses of SRMs over a 2year period. Field replicates were typically within 30%. Analysis of polychlorinated dibenzo-p-dioxins and furans (PCDDs/PCDFs) in Lake Erie samples was carried out in accordance with the USEPA Method 1613B protocol for dioxins and furans (USEPA 1994) by Axys Analytical Services Ltd. (Sidney, BC). Analyses of Lake Ontario sediment for PCDDs/PCDFs were carried out using Ontario Ministry of the Environment (OME) standard methods (OME 2000). Prior to extraction, samples were spiked with 13 C-labeled surrogate standards. Toluene sediment extracts were subjected to a sequential cleanup including an acid/base wash, alumina column, carbon/celite column, and a second alumina column procedure (USEPA 1994), or a sequential cleanup including a modified silica column, alumina column, and an Amoco PX21— activated silica column procedure (OME 2000). Analysis by high-resolution gas chromatography— high-resolution mass spectrometry was carried out in selected ion monitoring (SIM) mode using a VG Ultima or Autospec high-resolution mass spectrometer equipped with a Hewlett-Packard 5890 gas chromatograph. Chromatographic separation was carried out using a 0.25 mm i.d. 60 m DB-5 column with a 0.25 µm stationary phase thickness. Procedural blanks and precision and recovery samples were processed with each sample batch and all data for individual samples were checked against surrogate recoveries. Toxic equivalents (TEQs) were calculated using the International Toxicity Equivalency Factor (ITEF) method (Van den Berg et al. 1998); congeners present at levels lower than the method detection limits were given zero values for TEQ calculations. Trace metals and mercury analyses were performed by Seprotech Laboratories (Ottawa, ON). Trace metal concentrations were determined by a hot aqua-regia extraction with measurement by
439
ICP-AES (McLaren 1981). Total mercury was determined by digestion with hot nitric acid and hydrochloric acid followed with measurement by cold vapor atomic absorption spectrometer (USEPA 1981). RESULTS AND DISCUSSION Sediment Distribution For the purposes of discussing trends in sediment contamination as they related to sediment distribution and sediment processes, the convention for sediment characterization described by Thomas et al. (1976) for Lake Erie, and by Thomas et al. (1972) for Lake Ontario was adopted. Sediments in both lakes were classified either as non-depositional, consisting of bedrock, glacial tills and glacio-lacustrine clays, or depositional which were comprised of fine-grained material including silts and clays that accumulate in deep water areas. Lake Erie was divided into three basins; the Pelee Lorraine sill separated the western basin from the central basin, and the central basin was separated from the eastern basin by the Long Point Erie sill (Fig. 1). Trends in some contaminant distributions in Lake Erie warranted further classification of the central basin in terms of the northern area and the southern area. Lake Ontario was also divided into three major depositional areas; the Niagara basin was separated from the Mississauga basin by the Whitby-Olcott sill, and the Scotch Bonnet sill separated the Mississauga basin from the Rochester basin. The outflow of the Lake into the St. Lawrence River, the Kingston Basin, experiences insignificant sediment transport from the main lake due to the presence of a major topographical barrier, the Duck-Galloo sill (Thomas et al. 1972). Spatial Distribution of Metals Sediment metals data for Lakes Ontario and Erie were assessed for spatial patterns to identify potential source areas, for comparison with Canadian Sediment Quality Guideline values, and to calculate the degree of surficial sediment enrichment as a measure of the impacts of anthropogenic activities by comparison with historical background levels. The Canadian Sediment Quality Guidelines (CCME 1999) are designed as screening tools for assessment of potential risk and determination of the relative priority of sediment quality concerns. The threshold effect level (TEL) represents the concentration below which adverse biological ef-
440
Marvin et al.
FIG. 1. Distribution of major non-depositional (near shore) and depositional areas in Lakes Erie and Ontario. fects are expected to occur rarely, while the probable effect level (PEL) defines the level above which adverse effects are expected to occur frequently (CCME 1999). These guidelines were selected over other Canadian and US guidelines because of their applicability to a larger suite of analytes, and because they are considered to represent a conservative approach to evaluating sediment quality (Rheaume et al. 2000). A comparison of Canadian, U.S., and Ontario sediment quality guidelines and their application can be found in Rheaume et al. (2000). Sediment enrichment factors were calculated as the average ratio of surficial sediment (top 3 cm) concentration to background concentration for five metals including chromium, lead, nickel, copper, and zinc. Based on 210Pb data, the top 3 cm of sediment was estimated to represent the period 1993 to 1997 in Lake Erie and 1990 to 1998 in Lake Ontario. Historical background metal concentrations were operationally defined as the average of concentrations from the 40 to 50 cm interval to bottom of benthos cores and are listed in Table 1. This in-
terval in both lakes pre-dates modern industrial activity; a depth of 34 cm in Lake Ontario was estimated to represent 1880. Table 1 also shows the surficial sediment 75 th percentile values for each metal, and the percentage of stations that exceeded the sediment quality guidelines. Background concentrations for metals in both Lakes Erie and Ontario shown in Table 1 were generally similar to those reported by Mudroch et al. (1988). These background levels are generally representative of sediment concentrations prior to the advent of anthropogenic influences. However, Pirrone et al. (1998) estimated that atmospheric emissions of mercury in North America, resulting from gold and silver production, peaked in 1879. As a result, the background mercury values shown in Table 1 may not entirely represent pre-colonial concentrations. The spatial distribution of sediment enrichment factors for the five metals in Lakes Erie and Ontario is shown in Figure 2. Enrichment factors were clearly elevated in Lake Ontario compared to Lake Erie. The average enrichment ratios for the Niagara, Mississauga, Rochester, and Kingston basins
Dioxin (pg/g TEQs)
Arsenic Cadmium Chromium Copper Iron Lead Manganese Mercury Nickel Nitrogen Phosphorus Zinc Aluminum PCBs (ng/g)
Background Concentrations (total metals) µg/g ex. Iron and Aluminum (%) West Center East Erie Erie Erie Ontario 5.8 7.1 6.8
