Prepared by: Michael A. Zarull, John H. Hartig, and Lisa Maynard
Sediment Priority Action Committee
Great Lakes Water Quality Board
III. CONTAMINATED SEDIMENT IN THE GREAT LAKES
In the late 1960s through the late 1970s, a series of comprehensive surveys of the geochemical composition of the surficial sediment in each of the Great Lakes were conducted. These surveys examined sediment samples from the top three centimeters, collected from a one or a ten square kilometer grid, to determine the spatial pattern of pelagic sediment. These data led Allan (1986) to conclude that there are two basic distribution patterns for trace metals in the pelagic zones of the Great Lakes. In addition, temporal changes in sediment quality were documented from sediment cores at selected stations (Zarull and Mudroch 1993).
The first grouping has its highest concentrations in the upper lakes, particularly Lake Superior and Georgian Bay, which is thought to be due to the bedrock composition of the Canadian Shield. This pattern of high concentration in the upper lakes occurs with most heavy metals associated with natural mineralization (e.g., chromium and nickel). These higher sediment concentrations may also result from the very low sedimentation rates and consequently low dilution of the upper lakes. Higher concentrations of chromium in some parts of Lakes Erie and Ontario have been attributed to the plating industries located in the lower lakes and connecting channels drainage basins (Thomas and Mudroch 1979; Allan 1986).
The other distribution pattern found in the open waters of the Great Lakes is associated with metals and organics originating from urban effluents. The greatest concentrations of these substances are found in the lower lakes, particularly in the vicinity of the western basin of Lake Erie, Lake St. Clair, and the Detroit River, along with the depositional basins of Lake Ontario (in particular, the Niagara basin). This pattern also holds for the distribution of lead, zinc, cadmium, and PCBs (Thomas and Mudroch 1979).
Analysis of contaminant concentrations from dated sediment cores indicates that the more recent concentrations of metals such as lead, copper, zinc, and mercury are considerably greater than their pre-industrial levels by up to a factor of ten. In general, the results showed that the loadings of inorganic contaminants had increased significantly since the 1900s and that organic contaminants began to accumulate in the sediment around the 1940s. The increase in these loadings to the Great Lakes sediment is ascribed to inputs from industry, agriculture, and municipalities along the shoreline, and to transport via tributaries. Atmospheric deposition has also contributed considerably to the sediment loadings of several contaminants (Kemp and Thomas 1976; Nriagu et al.1979; Thomas and Mudroch 1979; Durham and Oliver 1983; Nriagu 1986; Robbins et al. 1990).
The chronology of Lake Ontario sediment contamination by mirex and its subsequent redistribution illustrates the large-scale spatial and temporal changes that can be expected for a persistent organic contaminant. An investigation by Thomas and Frank (1987) indicated two sources of mirex to the lake. The Niagara River was the major source, which had resulted from loss during the manufacturing process; and the second source was from the Oswego River, which came from a spill to the river in the mid 1950s. Mirex from the Niagara River entered the lake and moved to the northwest, settling in the deep basin. A larger portion of the contaminated sediment was transported by a major circulation process and carried along the south shore. The size of the contaminated area of the surficial sediment continued to increase, even though production was discontinued in 1976. Changes in the distribution of mirex in the eastern basin are thought to have resulted from the transfer of sediment-bound mirex, since there was no additional source input in this area. The expanded distribution and increased concentrations that subsequently were observed between 1968 and 1977 could only be due to intermittent remobilization processes of Oswego River material. This phenomenon led to increased open lake contamination and far field contamination of the St. Lawrence River (Thomas and Frank 1987).
Another example of large-scale spatial and temporal changes in sediment contamination is the Saginaw River/Saginaw Bay, Lake Huron. Saginaw River is the major tributary to the Bay. During the 1960s, 1970s, and early 1980s, between 27 and 54 tonnes of PCBs were released from a General Motors Plant in Bay City, Michigan and found in and on the land adjacent to the Saginaw River (International Joint Commission 1987b). During 1986, a once-in-500 year flood occurred. This flood occurred in September 1986 and resulted from a rainfall of up to 30 cm over 36 hours in some areas of the watershed, followed by another 8-18 cm during the remaining 19 days of the month. This once-in-500 year flood resulted in considerable movement of PCB and other contaminated sediment throughout the watershed and Bay (Michigan Department of Natural Resources 1988).
The examples of mirex in Lake Ontario and PCBs in the Saginaw River/Bay demonstrate that local nearshore contamination is unstable and remobilization by physical, chemical, or biological processes will result in the transfer of an apparently local problem into lakewide contamination. Therefore, the time for positive action is when contaminated sediment is localized, since once the sediment disperses to the open lake, the resolution of the problem becomes very much more complex (Reynoldson et al. 1988).
Sediment contaminated with metals, persistent toxic organics, nutrients, and oxygen demanding substances can be found in many areas throughout the Great Lakes. However, the highest levels of sediment-associated contaminants and some of the worst manifestations of their resultant problems are found in the urban-industrial harbors, embayments, and river mouths. These are the areas that are likely the most significant, from an ecological point of view. These nearshore areas represent the spawning and nursery sites for most fish species, the nesting and feeding areas for most of the aquatic avian fauna, the areas of highest biological productivity, the areas of greatest human contact, and the primary places of direct human contact with the sediment.
All Areas of Concern contain some sediment with elevated levels of nutrients, metals, or persistent organic contaminants. Sediment data were gathered on different occasions over a number of years by a variety of investigators and were used not only to describe the extent of contamination, but also as the basis for "listing" a sediment problem in an Area of Concern. In these assessments, bulk chemical analyses were performed and the results were compared to dredging guidelines (International Joint Commission 1982). Early estimates of the potential costs of sediment cleanup, based on data such as these, provided a bleak economic picture for the Areas of Concern and the Great Lakes. Estimates by Leger (1989) for nine Areas of Concern - Southern Green Bay/Fox River, Milwaukee Harbor, Waukegan Harbor, Grand Calumet River/Indiana Harbor, Saginaw River/Bay, Clinton River, Rouge River, Black River, and Ashtabula River/Harbor - ranged from around $185 million to $604 million. In the Canadian Areas of Concern and the Ontario portion of the interconnecting channels, Wardlaw et al. (1995) estimated that the total volume of "highly" contaminated sediment was about 172,000 m3. If it is assumed that all of the material will be dredged and placed in an existing confined disposal facility and we employ the cost estimate of $25/yd3 used by Leger (1989), the cost of cleanup can be estimated to be around $4 million ($25 x 157,276.8 yd3). The term "highly" contaminated means having contaminant levels over Ontario's Severe Effects Level (Persaud et al. 1992). These preliminary cost estimates are highly sobering and show that contaminated sediment is a substantial challenge. However, these cost estimates have been so significant that benefits tend to be ignored, and the perception that prevails is one of cleanup activities being cost prohibitive.
Sediment cleanup in the Areas of Concern has been shrouded by the discussion of high costs. Also contributing to the perception that cleanup actions are not feasible is the lack of attention given to the potential to renew ecological well being. It is important to remember that there have been significant refinements to assessment approaches since dredging guidelines were derived. More recent approaches, while not specifically developed to quantify the contribution of sediment contaminants to beneficial use impairments, do have some ecological linkages. For example, Ontario's Provincial Sediment Quality Guidelines are biologically-based and literature derived chemical-by-chemical criteria (Persaud et al. 1992), and the U.S. EPA's chemically-based criteria are based on risk analysis (U.S. Environmental Protection Agency 1992).
A clear understanding of the ecological benefits to be accrued through sediment cleanup, and some level of quantification of those benefits, are critical for the development of a complete sediment management strategy. Documenting the sediment problem in this context will help stipulate how much needs to be cleaned up, why, and what improvements can be expected in the beneficial use impairments over time. This understanding can provide adequate justification for an active cleanup program, and also represents a principle consideration in the adoption of non- intervention alternative strategies. In developing this understanding, it is important not only to know the existing degree of ecological impairment associated with sediment contaminants, but also the circumstances under which those relationships and impacts might change (i.e., contaminants become more available or more detrimental).