Table 34. Yearly median densities (n per 500 m^) of winter flounder larvae in en- 



trainment samples during their season of occurrence and annual entrain- 

 ment estimates with 95% confidence intervals for MNPS Units 1 and 2 

 from 1976 through 1985. 



The 10 yr of entrainment data provided an opportunity to examine factors that would affect annual 

 abundance and the timing of peak entrainment. The Gompertz function (Equation 6) was fit to the data 

 to compare the a parameter as an index of abundance to the median entrainment density and to compare 

 annual estimated dates of peak abundance (Table 35). A high correlation was found between the median 

 and the a parameter (Spearman's rank correlation coefficient = 0.855; p = 0.0016), showing that the a 

 parameter was a good index of abundance. The annual abundance of entrained larvae appeared to be 

 related to annual egg production in the river (Fig. 32). A linear regression was fitted to the total egg 

 production index (Table 11) versus the a parameter. The 1980 total egg production index was excluded 

 from the analyses because of the previously discussed problems with data from that year. The relationship 

 between egg production and entrainment abundance suggested that natural mortality from hatching to the 

 time of entrainment was similar for the 8 yr examined. 



Ill 



