TABLE 26. The seasonal 5-mean density (no. per 500 m ) 

 of cunner larvae collected at stations EN and NB by year. 



vived to the larval stage. Further, survival in re- 

 cent years was even less than usual and probably 



EN, suggesting that larval densities may be lower 

 near the MNPS intaJkes compared to other areas. 

 Larval abundance began declining in 1984 and 

 has not returned to historic levels. However, a 

 similarly large decrease in larval abundance, which 

 occurred from 1977 to 1978, was followed by an 

 increase to relatively high densities, in the early 

 1980s. Low larval abundance in 1984 may have 

 been due to predation or other factors operating 

 concurrently on the entire summer 

 ichthyoplankton assemblage, because larval den- 

 sities of tautog and anchovies were low that year 

 aswell(NUSCO 1987b). A comparison of cunner 

 and tautog larval abundance showed a remarkably 

 similar pattern at both EN and NB (Tables 23 

 and 26). At EN, lowest densities occurred in 

 1978, 1984, and 1986, and the greatest densities 

 occurred in 1981. At NB, the greatest abundances 

 were in 1983 and lowest in 1986. However , these 

 similarities between the two species were not ev- 

 ident in their egg abundance estimates. 



Cunner egg survival was reported as low (about 

 5%) by Williams et al. (1973), based on the ex- 

 amination of embryonic development in field- 

 collected eggs; this survival estimate did not take 

 into account the possible additional loss due to 

 predation. An index of egg survival wat, estimated 

 by the ratio of the annual 6-mean for larvae to 

 eggs (see Table 5 for egg 5-mean) at EN for 

 cunner, and also tautog for comparison (Table 

 27). These low ratios indicated that few eggs sur- 



produced fewer juvenile recruits from 1984 

 through 1986. Egg densities seem unrelated to 

 the corresponding larval densities and no com- 

 pensatory relationship was found. Whatever fac- 

 tors affected survival of cunner larvae also affected 

 tautog larvae; ranks of the larval to egg ratios 

 were nearly the same for the two species and they 

 were significantly correlated (Spearman rank cor- 

 relation, r = 0.83, p<0.01). 



Cunner were taken only frequently enough at 

 three (FN, JC, and NB) of the six trawl stations 

 to use the 5-mean for describing abundances (Ta- 

 ble 28). At aU three stations there was a decline 



TABLE 28. Seasonal 5-mean catch (CPUE) and and 95% 

 confidence intervals of cunner caught by trawl at selected 

 stations. 



Season used for calculating 5-mean was May-August at 

 IN, May-September at JC and April-November at NB. 



Pish Ecology Studies 285 



