TABLE 1. Annual mark and recapture data from Niantic River adult winter flounder abundance studies during 

 the spawning season from 1976 through 1987. 



Minimum size for marking was 15 cm during 1976-82 and 20 cm thereafter. 



data of impinged fish were also available. Routine 

 impingement monitoring was discontinued in mid- 

 December 1987, upon agreement between NU 

 and CT DEP. 



Results and Discussion 



Adult and juvenile studies 



Abundance of winter flounder near 



Millstone 



The Niantic River winter flounder population 

 is deraographically open and therefore subject to 

 immigration, emigration, natural death, and re- 

 moval by fishermen (White et al. 1982). Mark 

 and recapture surveys designed to estimate abun- 

 dance of open populations using the stochastic 

 model of Jolly began in 1976 (NUSCO 1987). 

 The Jolly model is an extremely powerful general 

 formula that uses all the information provided by 

 the mark and recapture experiment and provides 

 the most efficient abundance estimates for open 

 populations as long as basic assumptions are ap- 

 proximately met (Cormack 1968; Southwood 

 1978; Begon 1979). Application of the Jolly 

 model to the Niantic River winter flounder pop- 



ulation was discussed previously in NUSCO 

 (1986a, 1987). 



The 1987 survey had the latest start since 1980 

 due to extended ice cover in the river, which re- 

 sulted in only 5 weeks of sampling (Table 1). 

 Although the percentage of recaptures (5.7%) was 

 similar to the range observed during 1982-86 

 (4.7-7.0%), fewer fish were marked than in any 

 previous year. A condition peculiar to 1987 was 

 the large amount of kelp and detritus found in 

 the mid to lower river chaimel (stations 2 and 4), 

 which entirely precluded sampling at the latter 

 station after the first week of study. During most 

 of the year, few winter flounder were found in 

 the navigation channel (stations 1 and 2) and 

 catch in the basin of the upper river (station 51) 

 was also less than in previous years; most fish 

 were in the western arm of the river (stations 

 52-54). However, during the fourth week of the 

 survey, many fish withdrew from the upper river 

 arm into the basin following a storm. 



For the 1987 survey, log-likelihood ratio tests 

 indicated no significant differences in the propor- 

 tions of marked and recaptured fish classified by 



158 



