with values near the extremes (high or low) in 

 each sampling quarter. Abundances at GN during 

 this period were well within the range established 

 in previous years, and exhibited much less tem- 

 poral variation. 



After regression analysis removed variation at- 

 tributable to abiotic factors (23% at WP and 

 45% at GN), no significant trends in rhynchocoel 

 abundance were evident. In addition, no signifi- 

 cant differences were evident among years at either 

 sampling station. 



Haploscoloplos fragilis 



This burrowing deposit-feeder is capable of in- 

 gesting large sand grains and deriving nutrients 

 from the bacterial flora developed upon them. In 

 addition, morphological adaptations allow this 

 species to rapidly burrow through loose uncon- 

 solidated sands so that this species can inhabit 

 areas of high sediment movement (Myers 1977). 



Haploscoloplos fragilis is a typical dominant of 

 the clean sandy beach communities found at GN 

 and WP where it ranked either 2nd or 3rd in 

 terms of abundance during the last two years. 

 Quarterly densities from September 1985 to June 

 1987 ranged from - 26/core at GN and - 

 8/core at WP (Fig. 7K-L). At both stations, this 

 species was most abundant in September and 

 least abundant in March or ,Iune. Peak abun- 

 dances at GN during 1986 and 1987 were generally 

 higher than previous years (except 1985), while 

 at WP they were below those obtained in previous 

 years. 



After removing natural variation (72% at GN 

 and 86% at WP), there were no significant long- 

 term trends in density at either station, nor were 

 there any significant differences among sampling 

 years at GN. At WP, the 1986 and 1987 means 

 were significantly different from only 1984. 



Species Diversity 



Annual mean species diversity (H') during the 

 past two years ranged from 1.1 - 2.0, evenness (J) 

 from 0.4 - 0.6, and species numbers from 6-17 

 (Table 3). Lowest diversity of aU stations occurred 

 at GN in 1986 (11'= 1.1) reflecting the generally 

 lower numbers of species and low evenness (due 

 to high densities of Paraonis fulgens). All indices 

 at GN for 1987 were within the range of previous 

 years. At JC, H', S and N were generally higher 

 than other stations in both 1986 and 1987 with 

 1987 totals reflecting high abundances of 

 Scolecolepides viridis, He.disle diversicolor and 

 Fahrica sabella. 



Overall there were no major shifts in parameters 

 used to describe diversity of intertidal communities 

 in 1986 or 1987 relative to the pre-operational 

 period. Althougli the values of IT, S, and N 

 were lower at WP in both 1986 and 1987, they 

 were still within the historical range. 



Cluster Analysis 



Cluster analysis produced a dendrogram which 

 showed the very low similarity (-26.0 %) of the 

 JC community to that of WP and GN (Fig. 8). 

 This spatial pattern is consistent with previous 

 years and is due to differences in both patterns 

 of species abundance and composition (NUSCO 

 1987). During pre-operational and operational 

 years, similarity among the WP (potentially im- 

 pacted) and GN (reference) communities (Group 

 I) was higher with all years grouping at 45%. 

 This group further divided into three subgroups: 

 Group A contained most GN years; Group B 

 most WP years; and Group C, the 1986 GN and 

 WP years. Within the WP cluster (Group A), the 

 1987 collection linked with 1982-84, and separated 

 from other years within the group due to similarly 

 low numbers of Scolecolepides viridis, Polydora 

 ligni and Capitella spp.. Within the GN cluster, 

 1987 chained onto collections made in 1984 and 

 1985; these shared similarly high numbers of 

 oligochaetcs and low numbers of rhynchocoels. 

 Group C contained 1986 GN and WP collections 



Benthic Infauna 



