suits of a probit analysis showed that the length 

 of 50% sexual maturation of Niantic River females 

 during 1981-87 was 26.8 cm with a 95% confi- 

 dence interval of 26.3 to 27.2 cm. On an aimual 

 basis, values ranged from 25.1 cm in 1983 to 29.4 

 cm in 1981; most of these fish were 3 or 4 years 

 old. Based on our observations, many males 

 mature at even smaller sizes (10-12 cm) and 

 younger ages (2 years) than females. 



Winter flounder spawning was followed by 

 noting the weekly change in the percentage of 

 gravid females larger than 25 cm in the Niantic 

 River. Generally, most spawning was completed 

 by late March or early April as relatively few 

 gravid females remained in the river (Fig. 12). Ice 



in the river prevented starting population surveys 

 earlier in January or February, so for most years 

 approximately two-thirds of the females examined 

 during late February or early March had spawned 

 before sampling began. In most cases, spawning 

 appeared to be correlated with water temperature. 

 In relatively cold years, proportionately fewer fe- 

 males spawned during the earlier portion of the 

 survey, whereas in warmer years more were spent 

 at the beginning of sampling (NUSCO 1987). 



The annual proportion of females larger than 

 26 cm was used with the Jolly index of abundance 

 for that year to obtain a relative index of female 

 spawners. Spawning females made up 20 to 51 % 



FEBRUARY 



MARCH 



APRIL 



Fig. 12. Percentage of adult female winter flounder in 



1983 through 1987. 

 of the population during the past 1 1 years (Table 

 10). Percentages for 1977-82 were lower because 

 they were based on abundance estimates for all 

 fish larger than 15 cm and thus included more 

 immature fish in the total. The value for 1980 

 (20%) was particularly low and was an indication 

 that comparatively more small, immature fish 

 were caught during that survey. 



spawning condition by week in the Niantic River from 



Annual average fecundity estimates were cal- 

 culated using a length-fecundity relationship de- 

 scribed in NUSCO (1987) with annual length- 

 frequency data. Values have been relatively con- 

 sistent with somewhat greater averages found 

 since 1982, when the start of the surveys was 

 advanced into February. During the past several 

 years, most females larger than 40 cm were found 



Winter Flounder Studies 



171 



