lected at GN were at or above the previous years' 

 ranges. Similarly, the numbers of polychaete spe- 

 cies and abundance of polychaetes, molluscs and 

 arthropods at IN were above the previous ranges. 

 Most values for JC were within the range of pre- 

 vious years and the most notable change was in 

 the high numbers of arthropods observed in 1987. 

 At EF, the overall trend was for lower numbers 

 of species and abundances. 



Community Abundance 



The log-transformed quarterly mean abundances 

 of subtidal communities since 1 980 are presented 

 in Figure 10. Densities during 1986-87 ranged 

 from 67 - 208/core at EF,, 181 - 455/core (expo- 

 nential of values in Figure 10) at GN, 27 - 

 503/core at IN and 138 - 600/core at JC . Except 

 for June 1987 at IN and GN, values obtained in 

 the la.st two years were within the range established 

 in previous years. Although within the range, 

 abundances at EF, IN and JC in March 1986 

 were among the lowest ever recorded for these 

 stations. Levels near the upper extremes also oc- 

 curred in the last two years (e.g., December 1985, 

 March and June 1986 at IN, September 1986 at 

 JC, and June 1987 at GN). No comparable peaks 

 were evident at EF. 



Multiple regression analyses, which accounted 

 for 43, 54, 57 and 44% of the total variation at 

 EF, GN, IN and JC, respectively, revealed no 

 significant trends in community abundance at any 

 subtidal station. However, pair- wise comparisons 

 revealed significant interannual differences at all 

 stations. 



since 1980, but both were significant from only 

 1983. The 1987 mean was also significantly higher 

 than 1980. 



Number of Species 



In 1986-87, mean quarterly number of species 

 comprising subtidal communities from 1986-87 

 ranged between 19-37/core at EF, 21-37/core at 

 GN, 12-29/core at IN and I8-37/core at JC (Fig. 

 11). Lowest number of species during this period 

 occurred in March 1987, at GN and EF and in 

 March 1986, at IN and JC. Highest values at all 

 stations were obtained in June 1987. All quarterly 

 values, except in June 1987 at IN, were within 

 the range observed during the 1980-85 period. 



After removing variation attributable to explan- 

 atory variables, significant trends were evident in 

 species number at IN, EF, and GN. An increasing 

 trend in annual adjusted means was evident at 

 IN; the 1986 mean was significantly higher than 

 only 1984, but 1987 was significantly higher than 

 ail previous years except 1985. At EF, a significant 

 increasing trend was during the pre-operational 

 period; this trend reversed in 1986, which was 

 significantly lower than 1984 and 1985. The 1987 

 mean was also lower than 1985 but not 1984. At 

 GN, a significant increasing trend began in 1984. 

 Values obtained in in 1986 were significantly 

 higher than those obtained from 1981-1983, while 

 1987 was significant from only 1981. .lordan 

 Cove was the only community where a significant 

 trend in species number did not occur and pair- 

 wise comparisons of annual means revealed sig- 

 nificant differences only between 1987 and 1982. 



At EF, the 1986 mean was significantly lower 

 than 1984, when peak abundances occurred. 

 Abundances at JC in 1986 were significantly lower 

 than 1980, 1984 and 1987. The 1987 mean was 

 also significantly higher than 1982. Of all com- 

 munities, IN exhibited the largest change in abun- 

 dance during the last two years; 1986 and 1987 

 were significantly higher than annual means from 

 1981-1984. Annual mean densities at GN were 

 near (1986) or exceeded (1987) the range observed 



Community Dominance 



A list of all taxa ranking among the top ten in 

 abundance for 1986 and 1987 and over the 1980-85 

 period is presented in Table 5. In the last two 

 years, 6 of 7 top ranked organisms (according to 

 the BIV) were the same at JC and GN and four 

 of these were also among the top dominants at 

 EF. Mediomastus amhisela was the most consis- 

 tently dominant organism at GN, JC and IN in 



Benthic Infauna 



