in Long Island Sound fronn May to October, 

 and foimd greater numbers in spring and 

 autumn. Herman (1963) obtained eggs and 

 larvae in Narragansett Bay, R.I., from May 

 through August and in late October, and he 

 postulated a split spawning season in that 

 locality, with autumn spawning being more 

 productive. Scattergood, Trefethen, and Coffin 

 (1951) reported that spawning fish occurred 

 in Maine and Massachusetts waters in August 

 and September. Higham and Nicholson (1964) 

 found fish in spawning condition in purse 

 seine catches north of Long Island from June 

 to October and concluded that spawning in 

 those waters reached a peak in October, with 

 the suggestion of an earlier peak in May, 



It seems clear, then, that two distinct 

 groups of adults occur together from Long 

 Island northward--one which spawns in spring, 

 and the other in autumn. In this study, samples 

 of spawners and nonspawners were collected 

 simultaneously from pound net catches in Long 

 Island and Cape Cod waters in spring, audit is 

 known that mixed schools of spring and autumn 

 spawners live together in these waters through 

 summer (Higham and Nicholson, 1964). They 

 are subject, therefore, to the same environ- 

 ment. If the reproductive cycle were governed 

 primarily by environment, then it would be 

 expected that the fish would mature and spawn 

 at the same time, and variations in these 

 responses would be random. Morphological 

 differences also would be obscured. Results 

 of the present study show, however, that ir- 

 respective of the locality in which autumn 

 spawners were taken, their mean vertebral 

 counts were similar. This was also true for 

 spring spawners. But comparisons of the mean 

 counts of autumn spawners with those of 

 spring spawners showed significant hetero- 

 geneity. 



The seasonal difference in spawning times 

 of individuals in northern waters indicates a 

 fundamental difference in response to the 

 environment. Spring and autumn spawnings 

 (and early development of young) apparently 

 take place under entirely different temperature 

 conditions. In figure Z are plotted the m.onthly 

 nneans and the daily ranges of surface water 

 temperature for 1957-59 in localities where 

 samples of fish used in the present study were 

 collected. Temperatures near Cape Cod and 

 Long Island not only were similar, but the 

 means in April and May were from 4° to 13°C. 

 lower than those in September and October, 

 with overlap in ranges occurring only in May 

 (Ambrose Lightship was off station in May each 

 year.). Temperatures near Long Island in 

 September and October were similar to those 

 off North Carolina inNovember and December. 

 This suggests that the southward migration of 

 autumn spawners from Long Island to the 

 Carolinas may be governed by their physiologi- 

 cal response to tenmperature. This argument 

 seems to be supported by the lack of significant 



^ 25 



g 20 

 q: 

 ^ 15 



^ 10 

 (t 5 



UJ 



BUZZARDS LIGHTSHIP 

 Lot. 4I''2I'N. Long. 71" 03' W. 



AMBROSE LIGHTSHIP -• 

 Lot. 40° 27'N. Long. 73° 49' W. 



J I I I I L 



J I 



15 - 



10 -■ 

 5 



FRYING PAN SHOALS LIGHTSHIP 

 Lot. 33° 28' N. Long. 77° 33' W. 



■ ' ■ I I 1 1 — 



_i- 



-L. 



F Nfl A M J J A 

 MONTH 



N 



Figure 2. — Monthly range and mean temperatures at 

 three lightship stations along the Atlantic coast of the 

 United States, 1957-59. 



variation in the meristic counts of juveniles 

 that were presumed to have originated from 

 the autumn spawning from Long Island south- 

 ward (Sutherland, 1963), 



There is evidence of other biological dif- 

 ferences between autumn and spring spawners. 

 For example, the mean ages and lengths of 

 spring spawners differed from those of autumn 

 spawners (table 1). Allowing for the advance in 

 age that takes place betweenautumn and spring 

 each year (arbitrarily on January 1), I found 

 that both the mean age and mean length of 

 autumn spawners off North Carolina were 

 significantly lower than those of the spring 

 spawners taken in Cape Cod and Long Island 

 waters in spring. There was, however, no 

 significant difference between autumn spawners 

 and nonspawners taken in northern waters in 

 spring. This suggests that autumn spawners 

 either suffered a greater fishing mortality or 

 were recruited at a younger age and at a 

 shorter length than spring spawners, or both. 



10 



