2.5. Statistical Analysis 



Acute C. dubia toxicity test results were analyzed by Fisher's Exact Test (USEPA 1989). 

 Data collected from the C. dubia Survival / Reproduction Assays were analyzed by Fisher's 

 Exact Test to obtain an indication of lethal toxicity, and reproductive data were analyzed 

 according to methods for statistical analysis of C. dubia reproduction data (USEPA 19S9). 

 Normality of the data was tested by Shapiro-Wilk's test of normality, followed by Bartlett's test 

 for homogeneity of variance (USEPA 1989). Once normality of the data was established, and 

 variances of the means were found to be homogenous, the effect of the porewater solutions on 

 reproduction was determined by Dunnett's Procedure (USEPA 1989). Mussel filtering assay 

 results were expressed as the percent decline in filtering rates from the control value. The 

 treatment results (i.e.. filtering rates measured in porewaters) were divided by the control results 

 (i.e., filtering rate measured in filtered creek water), and then 1 was subtracted from the quotient. 

 A negative value indicated inhibition of filtering rates by the porewaters (chronic toxicity), a 

 positive value indicated stimulation of filtering by the mussels, and indicated no response 

 (Sparks et al. 1992). Mussel filtering rates were also analyzed by a one-way ANOVA to 

 determine if significant differences between treatments and controls could be detected. 



3. Results 



3.1. Year 1 



3.1.1. Ceriodaphnia dubia Acute Toxicity Assay 



Because of high mortality in the control solution ( 12 %), this test was considered invalid and 

 no statistical analyses were performed (Table 3). Ceriodaphnia dubia were fed the morning of 



is 



