Figure 1. Location of Zooplankton sampling stations. EN - 

 Discharge, NB - Station 5 (Niantic Bay). 



Differences between weekly means of the (Ln + 1 transformed) 

 zooplankton abundances in EN and NB samples and in day and night samples 

 for each station, were calculated. These differences were examined 

 using paired t-tests ( <^ = 0.05), to test the null hypothesis that the 

 mean difference was zero. The data used in these paired t-tests were 

 examined for normal distribution using the Kolmogorov-Smirnov test in 

 the Statistical Analysis System programs (SAS Institute, Inc. 1982). 

 Only taxa meeting the assumption of normality were used in subsequent 

 analyses . 



Harmonic regressions were used to model the log-transformed 

 zooplankton abundance data from 1976 to 1983. The harmonic component 

 used in these models described a seasonal abundance with a period of one 

 year. The annual mean residual (deviation of the observed data from the 

 model prediction) provided an indication of the yearly deviation from 

 the model. Regressions having R^ values approaching or greater than 

 0.50 were used . 



