Executive Summary

Scope of the Study

The project included five complementary monitoring activities; near shore, off shore, watershed, bathing beach, and wet weather event. The near shore testing was conducted at 23 major outfalls to the lake, including the mouths of the Clinton River and Spillway, urban storm drains, smaller rivers and creeks and retention basin discharge points. Many samples were collected further from the outfalls than previous years due to lower lake levels. Off shore sampling was conducted at 11 sites, approximately one-quarter mile from shore. Seven of the off shore locations correspond to major near shore sample locations, two were adjacent to public beaches and two were municipal drinking water intake sites. Water chemistry and sediment bacteriological samples were collected during the summer and fall seasons at near and off shore sites. Near shore aqueous bacteriological samples and water quality meter readings were collected weekly from late May through September. Off shore water bacteriology sampling and water quality meter readings were collected during the summer and fall seasons. Sediment chemistry samples were collected at 11 near shore locations. Lake Sediment samples were collected at 20 locations for mercury analysis. Aqueous samples for trace mercury analyses were collected at 10 sites on the lake and 6 sites in the watershed.

In the watershed, concurrent sediment and water samples were collected at 20 locations for bacteriological examination during the summer and fall. Sediment chemistry samples were collected at 5 locations. Water chemistry samples were collected at 5 locations during wet and dry conditions.

Bathing beach water and sediment sampling was conducted at fifteen sites on the public beaches along Lake St. Clair (Blossom Heath, Memorial Park, HCMA Metropark, and New Baltimore). Samples were collected monthly from July through September 2002 and analyzed for aqueous and sediment bacteriology.

Event sampling was conducted in the watershed in response to rain events exceeding one half inch in a 24-hour period. Water samples were collected for bacteriological analysis at 20 strategic locations between April and September. Sample locations were selected based upon three criteria: proximity to known sewer overflows, locations of frequently high bacteria counts and at the most downstream sample site of each major sub-watershed drainage area of the Clinton River.

Results and Observations

Water chemistry data (nitrate, TKN, ammonia, ortho-phosphorus, total phosphorus, BOD, TOC, chloride and aluminum) was collected at all near shore and off shore sites during the summer and fall. This year, as in previous years, the near shore values for most parameters were higher than the corresponding off shore values. Three of the 20 near shore sites had the highest average concentrations for the nine aqueous chemistry parameters tested. The Irwin Branch Relief Drain (n24) had the highest concentration of TKN, ortho and total phosphorous, and TOC. The Clinton River (n23) had the highest concentration of chloride, BOD, and nitrate. The Milk River (n1) had the highest concentration of ammonia and aluminum. Only two sites had average aqueous chemistry values above critical limits. These included the Clinton River (n23) with an average nitrate level of 0.40 mg/L (critical value= 0.30 mg/L), and the Irwin Branch Relief Drain (n24) with an average total-phosphorus level of 0.075 mg/L (critical value= 0.05 mg/L). The average nitrate level at the Clinton River site was more than twice the nitrate level at the next highest site.

Among the off shore sample sites, Metropolitan Beach (o6) had the highest average concentration of nitrate, TOC, and chloride, with the nitrate level exceeding the critical value. The average total-phosphorous level at the Clinton River Spillway (o8) exceeded the critical value, and was over four times the concentration found at any other off shore site.

In the watershed samples perhaps the most notable finding was that in six of the nine aqueous chemistry parameters measured (i.e. chloride, nitrate, TKN, ortho-phosphorous, total phosphorous, and TOC) the dry weather samples had higher average concentrations than the wet weather samples. In two of these six parameters, chloride and nitrate, the difference was statistically significant. In all watershed samples collected, both wet and dry, the nitrate and total-phosphorous concentrations were above the critical value.

Sediment samples were collected in triplicate at eleven near shore and five watershed sample sites. The samples were analyzed for total Kjeldahl nitrogen, ammonia, total phosphorus, TOC, COD, arsenic, cadmium, chromium, copper, lead, zinc, mercury, nickel, total petroleum hydrocarbons, oil and grease, PCBs, PNAs and pesticides. The sites were sampled in triplicate in an effort to evaluate the variation in pollutant concentrations at each site. Previous year's sediment data appeared to suggest that pollutants were not as homogeneously distributed in sediment as in the water. This observation made it difficult to address the question of whether the sediment quality was improving or worsening.

The triplicate data was analyzed in several ways. Results were evaluated by calculating the relative percent difference, a method for assessing duplicate samples for quality control purposes. Standard deviations were also calculated, to compare individual results to their respective means. A wide variation in concentrations was observed for many substances.

It appears that both the substance itself and the sample location influence the distribution of a given substance in sediments. Certain substances appear to be more evenly distributed than others. For example, mercury is homogeneously distributed in the sediments of all tested watershed sites and a majority of the near shore sample sites. On the other hand, the concentrations of oil and grease, TPH and ammonia were found to have a wide range at most sample sites. The distribution of substances also varied greatly between sample locations. Some locations were found to have a narrow range of concentrations for a large majority of the parameters in the triplicate samples, while others had a wide range.

Several factors may influence the distribution of substances in sediments, likely including the chemical nature of the substance and the make-up of the sediment. Further sampling should provide a better understanding of this issue and permit one to identify trends in pollutant concentrations over time.

Results of metals analysis were compared to Ontario Ministry of Environment and United States Environmental Protection Agency sediment metal pollution classification guidelines. All of the metals exceeded a guideline at least once at a near shore or watershed sample location. Five of the eleven near shore sites sampled exceeded at least one of these guidelines. The Milk River, Clinton River and Liberty Drain exceeded the greatest number of guidelines. All five watershed sites sampled exceeded at least three of these guidelines.

Results of PCB and PNA analysis were also compared to Ontario Ministry of Environment and United States Environmental Protection Agency sediment classification guidelines. PCB's were detected at five of the 16 monitored sites in the lake and watershed. PNAs were detected at five of the 16 monitored sites in the lake and watershed. All results exceeded the guidelines.

Samples were collected for sediment mercury analysis at eleven near shore sites, five watershed sites, and at 20 sites in the U.S. portion of Lake St. Clair where mercury was detected in 2001. Mercury was detected at four near shore sites, three of which exceeded the Ontario Ministry of Environment critical value. All watershed sites tested were below the critical value. Mercury was also detected at six of the 20 sites in the lake. Four results exceeded the critical value.

Aqueous samples for trace mercury analysis were collected at 11 sites in the lake and watershed. Six of the seven near shore sites and all of the watershed sites had concentrations exceeding the wildlife protection value. The location with the highest aqueous mercury concentration was the Liberty Drain (n2). The Milk River at Alger (w58), which had the highest aqueous mercury concentration last year, had the lowest concentration among the watershed samples this year.

The geometric means of the E. coli concentrations at each near shore site over the entire sampling season were far below Body Contact Standards. The weekly geometric mean for all near shore sites did not exceed the 30 day Total Body Contact Standard during 2002. In fact, only three individual results exceeded the daily Total Body Contact Standard during 2002.

Average near shore E. coli concentrations for each sampling date and amount of precipitation within 72 hours prior to each sampling date were plotted as a line graph together over time. Correlation analysis of average near shore E. coli levels for each sampling date and total rainfall 72 hours prior to sampling revealed a correlation coefficient of 0.43. The correlation coefficient of 0.43 was below the threshold value of 0.48 to establish a statistically significant correlation for a set of eighteen sample dates. The prevalence of dry weather prior to sampling during 2002 undoubtedly influenced the limited correlation. Correlation analysis of sewer overflows and E. coli levels were not possible because only one sewer overflow occurred within 72 hours of a sampling event.

The Clinton River Watershed was sampled during periods of rain at 20 strategic locations. The results were assessed in terms of their correlation to the sum of the rainfall in the 72 hours prior to sampling. No statistically significant correlation was found between rainfall and bacterial event samples.

Sediment and aqueous E. coli samples were collected concurrently at beach, near shore, off shore and watershed sample sites. A statistically significant correlation was established between aqueous and sediment E. coli levels in the watershed samples, but not in the beach, near shore, or off shore samples.

The average dissolved oxygen (DO) concentrations at the off shore sites was found to be higher than the near shore sites. This difference was not statistically significant. The following sites had average DO levels below the desired value of 7mg/L: Irwin Relief Drain, Stephens Relief Drain, Milk River and Liberty Drain.

The average turbidity exceeded the critical value of 25 NTU at 4 near shore sites and 1 off shore site. A significant correlation between near shore E. coli levels and turbidity was found.

In this report, the aqueous and sediment results for near shore, off shore and watershed sample sites are summarized by site. In these site summaries the individual results have been transformed into z scores. The z score indicates how far and in what direction the result deviates from the average of all near shore, off shore or watershed sites, expressed in units of the distribution's standard deviation. This is a useful transformation to compare the relative standing of an individual result or site with the others.

The Lake St. Clair Assessment Project began in 1998 and now includes five years of comprehensive data on water and sediment quality from Lake St. Clair and the Clinton River Watershed. With five years of data, it is now statistically possible to begin evaluating trends and associations in water and sediment quality over time. This report summarizes historical observations in water and sediment quality from 1998 to 2002.

Summary

The primary objective of this project was to augment the existing surface water quality database for future reference and comparison. This data set extends the benchmark of water quality, which began in 1998. It represents the western portion of Lake St. Clair bordering Macomb County, including near shore, off shore, watershed and beach sample locations during the spring, summer, and fall seasons. It encompasses a wide range of parameters, including inorganic, organic and microbiological measures, for both water and sediment. Spatial and temporal trends are apparent for many parameters measured. Sediment metals, PCB and PNA data were assessed in terms of pollutant criteria developed by the United States Environmental Protection Agency and the Ontario Ministry of the Environment, giving an indication of the extent and spatial distribution of contaminated sediments in the lake and watershed.

This data set represents a working database containing water quality information for Lake St. Clair, with multiple potential utility. The database will be useful as a tool to address specific questions related to water quality in Lake St. Clair including the potential identification of streams and drains contributing specific types of pollution to the lake, pollutant dynamics, and the relationships of pollutant levels to environmental factors. Such factors may include known point and non-point source pollution inputs, land use and other landscape factors, and climatological conditions. These types of analyses may elucidate relationships and information useful to the goal of water quality improvement in the lake and surrounding watershed. This database can provide information useful to scientists, environmental protection workers, planners, natural resource and recreation managers, as well as municipalities and townships.

Future Projects and Applications

An immediate goal should be to make this database accessible to persons that may find it useful, including scientists, environmental protection workers, planners, natural resource and recreation managers, as well as municipalities and townships. Additionally, analysis of the entire available set of existing surface water bacteria data in conjunction with spatially distributed sources of rainfall data would likely yield more conclusive information pertaining to the correlation of surface water bacteria with rainfall and CSO/RTB discharge data. A predictive model of surface water bacteria based on rainfall and other environmental data would be a useful product of this effort.

Future monitoring data will build upon the existing database, permitting evaluation of trends in water quality over time as well as space. It should reflect the previous year's results and fill data gaps, permitting more conclusive evaluation of the data. This will be useful in understanding pollutant dynamics in the lake and in the appraisal of efforts to improve water quality.

TO VIEW THE 2002 SITE SUMMARIES click here

Back To Top