264 



From Table V it is clear that channel catfish from site 1710 (Scanlon Dam) have significantly 

 more mcrcur>' contamination than fish from the other sites. Mercury in walleye from site 

 1610 (Fond du Lac Reservoir) are also high but there is a high degree of uncertainty 

 associated with those results because of the small range of fish sizes and few samples 

 collected. 



Figure 6 shows the site specific relationship between mercury concentrations in fish and 

 length for various species. In general, one finds that individual data points from the smaller 

 confined river stretches (sites downstream of 1740) exhibit less variabilities about the 

 modeled curves than those for sites above the last barrier within the study area (1740 and 

 above). 



For the downstream sites, the fish mercury data are adequate to descnbe the mercury 

 exposure within those areas. However, some upstream sites (e.g. 1760 and 1762; channel 

 catfish), clearly show the infiuence of fish that have been subject to different mercury 

 exposures. From this data one may hypothesize the existence of one or more mercury 

 source(s) somewhere within or above that stretch of the river. 



Consumption Advisory Levels. Of the 214 fish analyzed, 15 fell within the MDH advisory 

 category 1. 180 were in category 2, and 19 were in category 3. These categories are defined 

 in the footnotes in Table VI. 



As a final application of the regression constants in Table V and equation (1) we may 

 calculate the fish lengths that correspond to the MDH mercury consumption advisory levels. 

 Table VI summarizes the results of these calculations for each species and site listed in Table 

 V. These calculations were extended to 20% above/below the maximum and minimum fish 

 lengths sampled. 



Mercury in Water. A basic statistical summary of mercury in water, plankton (>80u), and 

 sediments are given in Table VII while the raw data for each sample are listed in Appendix C. 



Mercury concentrations in water can be characterized as mostly between the detection limit of 

 2 ng/L and 4 ng/L. The variabilities at some sites could be a result of varying amounts of 

 particulates in the water sample that have been shown to contain significant amounts of 

 mercury (Glass et al., 1990). The sites with samples higher than average mercury in water 

 concentrations are 1651 (Forbay Lake), 1780 (St. Louis River at Hwy 31), 1820 (St. Louis 

 River at Gowan), 2710 (Stoney Brook at Hwy 31), and 2850 (Whiteface River at Hwy 226). 



13 



