months. At the start of each exposure period, 

 one rack of panels was removed for processing 

 and a new rack with fresh panels was deployed. 

 ITiroughout this report the exposure periods will 

 be referred to using the following abbreviations: 

 Feb-Aug, May-Nov, Aug-Feb and Nov-May. 

 Each abbreviation refers to the month of panel 

 deployment followed by the month of panel col- 

 lection. 



Sample Processing 



After collection, panels were either refrigerated 

 at 5 "C and processed immediately or frozen and 

 processed at a later time. Primary cover, as a 

 percentage, was estimated for each organism that 

 occupied more than 1% of the panel surface, e.g., 

 barnacles, biyozoans, tunicates and some algae. 

 Beginning in 1980, cover was estimated for 

 frecspace, mud and the dead tests of fouling spe- 

 cies, to complete the description of total primary 

 cover for each panel. Numerical abundance was 

 determined for barnacles and mussels by counting 

 the individuals on each panel. If the number of 

 individuals per panel exceeded 100, six subsamples 

 of 1 X 1 inch were randomly selected, three from 

 the upper half and three from the lower half of 

 the panel. In 1981, only the asbestos side (cur- 

 rently replaced by plexiglass) of each exposure 

 panel was used for determining numerical abun- 

 dance of fouling species; therefore, numerical 

 abundance data for barnacles and mussels were 

 not recorded for the wood panels in 1981. 



The abundance of woodborers was determined 

 after the panel had been scraped of fouling species. 

 All individuals of the genera Limnoria and Chelura 

 were counted when densities were less than 100 

 individuals per panel; otherwise, the subsampling 

 scheme previously described for barnacles and 

 mussels was used. Subsampling was always con- 

 ducted evenly between the top and the bottom 

 halves so that approximately 100 individuals were 

 collected from each panel. After assessing the 

 limnorid and chelurid abundances, panels were 

 frozen and subsequently examined by means of 

 X-ray photography (80 kV, 5 mA, for 1.2 min). 

 The radiographs were used to count the number 



of shipworms. Teredo navalis and T. hartschi, and 

 visually estimate the percentage of wood lost per 

 panel. The percentage of wood lost was deter- 

 mined by rating the general proportions of bright 

 areas, caused by various densities of shipworm 

 tubes and the dark areas caused by various degrees 

 of wood-loss. To determine the species of 

 shipworms collected, shipworms were randomly 

 removed from the panels until all or at least 100 

 individuals were identified from each site. 

 Shipworms smaller than 5 mm in length were 

 classified as Teredo juveniles because their pallets 

 were too small and undeveloped to allow accurate 

 identifications. 



Data Analysis 



The first set of exposure panels for which re- 

 sults are presented in this report was deployed in 

 November 1978, and the last set was collected in 

 August 1987. Each exposure period is represented, 

 therefore, by six panels replicated over years, a 

 maximum of four times during 2-unit operation, 

 and one or two times during 3-unit operation. 

 I'he actual numbers of panels processed for each 

 site/exposure period combination are presented in 

 Table 1 . Unit 3 began commercial operation on 

 April 23, 1986, and although some intermittent 

 operation and testing of circulation pumps oc- 

 curred before this date, we have considered all 

 sampling periods from Nov-May of 1979 to Nov- 

 May of 1986 as 2-unit operation. The monitoring 

 of all panels was suspended from November 1981 

 to February 1985 to investigate the life histories 

 of two shipworms. Teredo navalis and T. hartschi 

 in relation to seawater temperature. 



All averages and other summary statistics for 

 each taxon were computed by exposure period. 

 Temperature data were averaged by month. Per- 

 centage of primary cover and counts of individuals 

 were summarized as means. Data provided in 

 histograms were summarized by year within each 

 exposure period, while corresponding tables com- 

 pare data collected before vs. after 3-unit opera- 

 tion. This format represents the annual variability 

 as well as the differences between 2-unit and 

 3-unit operational periods. 



230 



