Evidence of Contaminant Loading to Lake Ontario from the Niagara River

The hypothesis that the concentrations of pollutants are the same throughout the length of the Niagara River, against the alternative that they are higher in the lower river, was examined using three sets of data: (1) Ontario Ministry of Environment's (MOE) results for organic substances in raw drinking water (1978–84); (2) MOE transect data for metals in water (1981); (3) Environment Canada data on organic substances in large-volume water samples and suspended sediments (1981). The designs of the studies dictated different statistical procedures for each set, with nonparametric procedures based on ranking being used for sets 1 and 3, and normal theory methods for set 2. Further, both univariate and multivariate procedures were used for set 3. Despite the considerable number of nondetections in these data, it was still possible to make quantitative comparative statements, with the degree of uncertainty providing the quantitative component. In each study, a number of substances were shown to be higher in the lower river than in the upper river, and by examining all substances of set 3 simultaneously, locations in the lower river were shown to be similar to each other as were locations in the upper river.