larger than 20 cm have been measured and sexed. 

 Non-measured fish were classified into various 

 length and sex groupings, depending upon the 

 year; at minimum, all fish caught can be classified 

 as smaller or larger than 15 cm. Since 1977, the 

 sex and reproductive condition of the larger winter 

 flounder were determined either by observing eggs 

 or milt or by the presence (males) or absence 

 (females) of ctenii on the caudal peduncle scales 

 of the left side (Smigielski 1975). Following mea- 

 surement or classitication, all fish 15 (1977-82) or 

 20 (1983-87) cm or larger were marked with a 

 number or letter made by a brass brand cooled 

 in liquid nitrogen; the mark was changed weekly. 

 Fish recaptured were noted and remarked with 

 the brand designating the week of their recapture. 

 In 1 976, fish were fin-clipped and recaptures were 

 marked with a numbered spaglietti tag. The log- 

 likelihood ratio test (G-test of Sokal and Rohlf 

 1969) was used to compare the proportions of 

 winter flounder marked and recaptured in each 

 category of sex, length interval, and station. The 

 probability level chosen to reject the null hypoth- 



esis in this and other statistical tests ^ven in this 

 section was p^0.05. 



Estimates of abundance of all winter flounder 

 1 5 or 20 cm and larger in the Niantic River during 

 the spawning season were obtained from the mark 

 and recapture data using the Jolly (1965) model. 

 The actual computations were done using a com- 

 puter program (Davies 1971) of Jolly's model. 

 As a result of a comprehensive review of our 

 mark-recapture methodology, a composite index 

 of relative abundance was developed to describe 

 the number of adult winter flounder spawning in 

 the Niantic River each year (NUSCO 1986a, 

 1987). This index was computed by averaging 

 the central weekly estimates of N (scaled as thou- 

 sands) from the Jolly model, with the first £ind 

 last estimates excluded when sufficient data were 

 available. These Jolly estimates are less reliable 

 and were eliminated from the computations in all 

 years except when the number of values used 

 would have been less than three. The standard 

 error of N (a) was determined as: 



(i)7Var of(Ar,) -I- Var of(Arj) + Var o^Wj) 



(1) 



where N is the weekly Jolly estimate of population size 



Using observations of reproductive condition 

 from 1981-87, probit analysis (SAS Institute Inc. 

 1985) was used to estimate the length at which 

 50% of all females were mature. An index of the 

 number of females reproducing in the Niantic 

 River each year since 1977 was created by esti- 

 mating their abundance in each 1-cm length in- 

 crement starting with 26 cm. Fecundity (annual 

 egg production per female) was estimated from 

 length-frequencies and a length-fecundity relation- 

 ship determined for Niantic River winter flounder 

 (NUSCO 1987). Annual average fecundity was 

 determined from the sum of all individual egg 

 production estimates divided by the number of 

 spawning females for each year. A relative index 

 of annual egg production (in billions) was found 

 by summing all individual estimates. 



The catch of winter flounder taken in the trawl 

 monitoring program from 1976 through 1987 (see 

 Fish Ecology section for methods) was also used 

 as an index of abundance. CPUE was expressed 

 as the annual 5-mean (Pennington 1986; see the 

 Delta Distribution section of this report for details 

 concerning this abundance index) during a calen- 

 dar year to facilitate comparisons with other re- 

 gional abundance indices. For other analyses, 

 length-frequency information was used to parti- 

 tion the catch into size groups smaller and larger 

 than 15 cm. 



Both the annual median CPUE of adult winter 

 flounder taken in the Niantic River during the 

 spawning season (1976-87) and the trawl moni- 

 toring program annual 5-mean (1976-86) were 



152 



