Examination of the length -frequency distribution of larvae collected from 1983 through 1985 showed 

 a separation between the first three developmental stages by predominant 0.5-mm size-classes (Fig. 20). 

 Stage 1 larvae were primarily in the 2.5 to 3.0 size-classes (84%), Stage 2 were 3.0 to 4.5 (88%), Stage 3 

 were 5.0 to 7.5 (87%), and Stage 4 were 6.5 to 8.0 (83%). These predominant size-classes for each 

 developmental stage were similar in each of the years (NUSCo 1984, 1985, 1986a), indicating that stage 

 development and length were closely related. Due to this relationship, larval developmental stages can be 

 estimated from length measurements for data collected prior to larval classification by stage in 1983. 



A comparison was made of the length-frequency distribution of larvae collected in the Niantic River 

 and Bay during 1981-85 (Fig. 21). Like the spatial distribution of developmental stages (Fig. 14), smaller 

 larvae dominated in the river and larger larvae in the bay. The 3.0-mm and smaller size-classes comprised 

 over 50% of the larvae collected in the river during the 5-yr period, even though the collections in 3 of 

 the 5 yr were made with 333-|im mesh nets and many of the smaller larvae were undersampled because 

 of net extrusion. Based on the large decline from the 3.0- to the 4.0-mm size-class, highest mortality 

 probably occurred at that size. Larvae in this length reinge were a combination of Stage 1 during yolk 

 absorption and Stage 2 at first feeding. This apparent time of high mortality may represent the larval 

 winter flounder "critical period", a concept first hypothesized by Hjort (1926) and discussed by May (1974) 

 for marine fishes. They suggested that starvation may be a compensatory factor. This period "of high 

 mortality was not as evident in the catch curve of winter flounder larvae presented by Pearcy (1962) for 

 the Mystic River. But the dominant size-class in his catch curve was 3.5-mni and his use of a 363 — |j,m 

 net may have resulted in an undersampling of smaller larvae. The slight increase in percentage starting 

 with the 6.0-mm size-class in the river and bay probably represented decreased growth in length for Stage 

 3 and 4 larvae during metamorphosis with a concurrent increase in body depth. Laroche (1981) reported 

 that for winter flounder the percentage of body depth at the pectoral fin base to standard length increased 

 from 9% for yolk-sac larvae to 31% for transformed larvae. The increasing percentage of larvae from 

 smaller to larger size-classes in the bay again indicated that spawning primarily occurred in the river and 

 larvae were gradually flushed into the bay. In the bay over 50% of the larvae collected were in the 5.0-mm 

 and larger size-classes. Based on the length frequency-developmental stage relationship (Fig. 20), they 

 were mostly Stage 3 and 4 larvae. The decline in percentage at about 6.5 to 7.0 mm in the river and bay 

 represented the transition to demersal juveniles that were less susceptible to capture with a plankton net. 



83 



