ture or rainfall in the period proceeding collection, can affect chemical concentrations (Phillips, 

 1980). Such natural factors can make trends in chemical concentrations difficult to detect. How- 

 ever, it would be difficult to find a natural factor with year-to-year changes that would cause 

 detection of a monotonic trend in chemical concentrations. The trends, therefore, are most likely 

 due to human actions. 



SITES WITH "HIGH" AND INCREASING CONCENTRATIONS 



It is comforting and expected that where trends exist they are, by far, mostly decreases. Increasing 

 trends are uncommon and most of those identified on Table 5 may only be chance sequences of con- 

 centrations. However, where such sequences include a concentration in the "high" range, the site 

 deserves attention. The following listing of such sites and chemicals includes trends in parenthesis that 

 were found with only a 90% level of confidence and sites where the "high" concentration was found in 

 only one year since 1990. It should also be noted that the highest concentrations were usually not those 

 measured in the most recent year. Nonetheless, these are trends and, unless shown otherwise, should be 

 interpreted as indicating that ongoing human activity is increasing chemical contamination. 



There are many more decreases than increases in chemical concentrations between 1986 and 1993. The 

 chemicals for which most of these trends exist are: chlorinated hydrocarbons, whose use has been 

 banned; tributyltin, whose use as an antifoulant on recreational boats has been banned; cadmium and 

 arsenic, whose uses have been severely curtailed; and copper, whose use has not deceased but whose 



8 



