sex or length. The percentages of fish recaptured 

 at stations 51 (9.6%) and 52 (8.1%) were signif- 

 icantly greater than those at the other stations 

 (2.5-3.9%). In most other years when significant 

 differences were found, greatest percentages of re- 



captures were from the lower river stations ( 1 or 

 2), most likely because any marked winter flounder 

 moving out of the river would have had a greater 

 probability of being caught near its mouth. 



TABLE 2. Weekly catch data used for estimating the Jolly index of winter flounder abundance during the 

 spawning season in the Niantic River. 



Using the methodology previously described, 

 annual .lolly composite abundance indices were 

 calculated to describe relative abundance of winter 

 flounder in the Niantic River during the spawoiing 



season. The weekly catch data (Table 2) were 

 used with the Jolly model and the computed 1987 

 index of abundance was 10.0 ± 3.8 (Table 3). 

 This represents a slight increase over the 1986 



TABLE 3. The Jolly index of abundance for winter flounder larger than 20 cm during the 1987 

 spawning season in the Niantic River. 



index, which was a 12-year low (Table 4; Fig. 3). 

 Sampling intensities must be relatively high to 

 obtain acceptable estimates of Jolly model param- 

 eters (Cormack 1979; Auckland 1980; Nichols et 

 al. 1981; Hightower and Gilbert 1984). Estimates 

 of population size (N) are biased to some degree 

 and for sampling intensities of 5 to 9% (which is 



similar to the Niantic River studies), N may have 

 low accuracy, depending upon the absolute pop- 

 ulation size (Gilbert 1973; Carothers 1973; 

 Hightower and Gilbert 1984). Examination of 

 simulations done by Hightower and Gilbert ( 1 984) 

 indicated that Jolly abundance estimates of Niantic 

 River winter flounder may have been accurate to 



Winter Flounder Studies 



159 



