dominant, species abundance related to abiotic 

 factors, reproductive/recruitment cycles, and long- 

 term climatic conditions. Removal of this varia- 

 tion was performed to improve the sensitivity of 

 statistical analyses used to compare data collected 

 during the pre-operational and operational peri- 

 ods. Regression analysis techniques were applied 

 to log-transformed (l,N(X+l)) quarterly abun- 

 dances (no. /core for total community and popu- 

 lation abundances) and quarterly mean numbers 

 of species (no. /core) collected from March 1979 

 through .June 1987. 



The following factors were used as explanatory 

 variables: 



Precipitation 



Daily precipitation records compiled by the IJ.S 

 Weather Bureau at the Ciroton Filtration Plant 

 were obtained from June 1976 through .Tune 1987. 

 Values to the nearest 0.01 inch were used as our 

 "rain" data for the regression model. 



Water and Air Temperature 



Ambient water temperatures (at the intake struc- 

 tures) and air temperatures (at the 33-foot level 

 of the Millstone mcteorolo^cal tower) were ob- 

 tained from the Northeast Utilities Environmental 

 Data Acquisition Network (EDAN). Daily av- 

 erages, based on observations made at 15-minute 

 intervals, were calculated for the period .Tunc 1976 

 to .Fune 1987. 



Wind Speed and Direction 



Wind speed and direction (at the 33-foot level 

 of the Millstone meteorological tower) were ex- 

 tracted from the EDAN database at 15-minute 

 intervals from .Tune 1976 to .June 1987. Ihese 

 values were used to calculate a Wind Index which 

 was wind speed weighted according to wind di- 

 rection. A NOAA navigational chart of the sam- 

 pling area was used to calculate site-specific, wind 

 directional weighting coefficients. The directional 

 weight ranged from 0, when wind could not in- 

 fluence the station, to 1, when the wind could 



result in waves directly affecting the area. The 

 Wind Index was then computed by multiplying 

 the directional weight by the wind speed. Because 

 the effect of wind was assumed to be cumulative, 

 daily averages were derived using only Wind Index 

 values greater than (that is, when the wind was 

 from a direction which could produce wind ef- 

 fects). 



Sedimentary Parameters 



Sedimentary parameters including mean grain 

 size and silt/clay content were obtained as part of 

 the monitoring studies and the quarterly values 

 included as explanatory variables in the regression 

 analyses. 



Seasonal Reproduction-Recruitment 

 Component 



Infaunal organisms in the Millstone area exhibit 

 annual peaks in abundance, which often reflect 

 the seasonal nature of reproduction and recruit- 

 ment cycles or periods of favorable climatic con- 

 ditions. Spectral analyses of quarterly data showed 

 that annual cycles in community abundance and 

 numbers of species were present. To account for 

 this periodicity, harmonic terms having a period 

 of 1 year were included as explanatory variables 

 in the regression models. 



Climatic Extremes (Deviations) 



Additional variables were created to represent 

 periods of extreme climatic conditions which have 

 occurred during the sampling period. High or low 

 deviations (i.e., extremes) were derived for each 

 abiotic factor as the difTerence between the quar- 

 terly mean or daily value and the 1 1 year mean 

 for that quarter. Deviations based on quarterly 

 means reflected the effects of longer term extremes 

 (e.g., an unusually cold winter), while those based 

 on daily values were intended to remove the ef- 

 fects of shorter-term episodic events (e.g., storms). 

 Daily deviations were averaged and summed (for 

 cumulative effects) over each sampling quarter. 



62 



