TABLE 17. Estimated larval winter flounder total mortality from hatching to 7 mm. 



increased ability to avoid the net, mcreased swim 

 speed, lowered availability, and perhaps increased 

 alertness with age. Therefore, sampling was dis- 

 continued in the Niantic River at the end of sum- 

 mer as water temperatures began to decline and 

 before young winter flounder left the shallows. 



In most years, abundance of young winter 

 flounder at stations LR and WA peaked in early 

 June, most likely when larval recruitment began 

 to be offset by mortality (Fig. 30). In 1986 at 

 WA, an initial peak in early June was followed 

 by a normal decrease (Fig. 31). However, abun- 

 dance increased to another peak in late July before 

 again declining. Weekly catches were also more 

 variable at WA in 1986 than in 1987, but the 

 converse was true at LR. The reason for the 

 variability in densities may have been due to 

 small-scale differences in young winter flounder 

 distribution. At WA, field notes indicated that 

 positions of tows varied more by replicate and 

 from week to week in 1986 than during 1987. 

 Greater effort was made to maintain a more uni- 

 form tow path at WA in 1987. In earlier years, 

 the area available for tows at LR was well-defined 

 by surrounding beds of eelgrass {Zostera marina). 

 By contrast, eelgrass was nearly absent there in 

 1987 and this led to increasing lateral variability 

 in tow position. At times, differences among 

 replicates were striking and the movement of a 

 few meters from the rock rip-rap along the high- 

 way often meant large differences in the number 

 of young caught as well as in the amounts of 

 algae, detritus, and other organisms. The increase 



in density seen from August to September in 1987 

 at LR was due, in part, to a slight shift in tow 

 path closer to the shoreline. Some differences in 

 catch may also have been attributed to varying 

 net efficiency because of bottom type or algae 

 and detritus loads. 



Comparison of moving averages of weekly 

 abundance among years showed that initial den- 

 sities of young were considerably higher in 1987 

 than in previous years and at LR generally re- 

 mained so throughout the season (Fig. 32). Abun- 

 dance in 1986 was also greater than in 1983-85 

 and peaked later during the season. By August, 

 relatively high abundance at WA in 1987 declined 

 to levels seen previously, with densities in 1986 

 appearing to have been only marginally greater. 



Growth 



Growth of young was illustrated by changes in 

 weekly mean length (Fig. 33). After a relatively 

 rapid increase from May through July, further 

 growth occurred at a slower rate throughout the 

 remainder of summer with little or no increase in 

 weekly means during September. Growth was 

 less variabile than densities were and weekly 

 means had relatively small 95% confidence inter- 

 vals. Unlike previous years, when mean lengths 

 at LR significantly exceeded that at WA by 20 

 mm or more in summer (NUSCO 1987), smaller 

 differences in growth were noted in 1986 and 

 1987, with no significant differences found between 

 stations during the latter year. 



198 



