(4) differential genotypic variation between 

 year classes. 



Since differences between the nnean verte- 

 bral counts of three year classes were not 

 significant, variation between year classes 

 was not considered in the following analyses. 



Comparison of Spawners and Nonspawners 

 within and between Localities. 



The first major hypothesis tested in this 

 study was that the spawners and nonspawners, 

 which occurred together in Cape Cod and Long 

 Island waters in spring, constituted two dis- 

 crete groups of fish. Vertebral counts of 

 spawners from each of the two localities 

 first were tested for evidence of significance 

 between sample means. Results of these 

 analyses gave no indication of heterogeneity 

 within localities. Further analysis indicated 

 that spawners from the two localities probably 

 were drawn from a homogeneous population. 

 Similar comparisons of nonspawners, both 

 within and between localities, led to the same 

 conclusions. Accordingly, the vertebral counts 

 of spawners from the two localities were 

 pooled and compared with the pooled counts 

 of nonspawners from both localities. A highly 

 significant F value of 47.74 (Pq q, = 6.64) 



indicates that it is unlikely that the samples 

 of spawners and nonspawners collected in 

 Cape Cod and Long Island waters in spring 

 were nnennbers of a homogeneous population. 



Vertebral counts of 1,989 spawners ranged 

 from 45 (10 fish) to 49 (1 fish), with the 

 majority having 47 vertebrae (1,441 fish); 

 mean counts of samples varied from 46.830 

 to 47.000. Vertebral counts of 2,014 non- 

 spawners also ranged from 45 (1 fish) to 49 

 (1 fish), with most having 47 vertebrae (1,509 

 fish); mean counts of samples ranged from 

 46.882 to 47.150. The difference between the 

 grand means of spawners (46.914) and non- 

 spawners (47.027) was 0.113 vertebra. 



The foregoing findings led to the second 

 major hypothesis, nannely, that the non- 

 spawners which occurred in Cape Cod and 

 Long Island waters in spring and the spawners 

 which occurred in Long Island and North 

 Carolina waters in autunnn constituted a 

 homogeneous population. 



Mean counts of samples within each locality 

 were tested for homogeneity and showed no 

 significant differences in either locality. The 

 autumn spawners from Long Island and North 

 Carolina were then compared and gave no 

 indication of heterogeneity. Finally, both 

 groups of autumn spawners were connpared 

 with the nonspawners taken in Cape Cod and 

 Long Island waters in spring, and differences 

 were not significant. 



Vertebral counts of 407 autumn spawners 

 from Long Island ranged from 46 (36 fish) 

 to 49 (1 fish), with most having a count of 47 



(315 fish); mean counts of samples ranged 

 between 46.949 and 47.107. A total of 1,638 

 autumn spawners from North Carolina had 

 counts ranging from 45 (2 fish) to 49 (1 fish), 

 with the majority having 47 (1,226 fish); mean 

 counts of samples varied from 46.972 to 

 47.101. The grand means for Long Island and 

 North Carolina autumn spawners were 47.052 

 and 47.030, respectively, as compared with 

 d count of 47.027 for nonspawners taken in 

 Cape Cod and Long Island waters in spring. 



Finally, comparison was made of the mean 

 vertebral counts of spring spawners and 

 autumn spawners (nonspawners taken in Cape 

 Cod and Long Island waters in spring were 

 included with autumn spawners from Long 

 Island and North Carolina). As might be ex- 

 pected, results of the analysis showed highly 

 significant heterogeneity between the nneans 

 (F = 69.22; Pq q, = 6.64). The grand mean 



vertebral count for 1,989 spring spawners 

 was 46.914 and for 4,059 autumn spawners, 

 47.031, a difference of 0.117 vertebra. 



The conclusions drawn from the foregoing 

 analyses of vertebral counts were that 



(1) spring spawners in Long Island and Cape 

 Cod waters constituted a homogeneous group 

 which was distinct from nonspawners, 



(2) autumn spawners in Long Island and North 

 Carolina waters and nonspawners in Long 

 Island waters in spring (autumn spawners) 

 constituted a honaogeneous group, and (3) 

 spring and autumn spawners constituted two 

 discrete groups of fish. The term group is 

 used here, as defined by Marr (1957), to de- 

 scribe "a fraction of a population with dis- 

 tinctive characteristics, the nature of which 

 (phenotypic or genotypic) has not yet been 

 determined." 



Comparison of Juveniles and Spawners of the 

 Same Year Class 



From a study of mean vertebral counts, 

 June (1958) and Sutherland (1963) found that 

 juvenile Atlantic menhaden inhabiting estuarine 

 nurseries along the Atlantic coast were sep- 

 arable into two groups--one occurring north 

 of Long Island, N.Y., and the other south of 

 there. Of further interest in the present study, 

 therefore, is a comparison of the mean verte- 

 bral counts of spring and autumn spawning 

 adults of a given year class with juveniles of 

 the same year class (age 0) from these two 

 general nursery areas to determine if some 

 relationship could be established. 



In table 4, the mean vertebral counts of 

 spring spawners in three year classes 

 (1955-57) are compared with the mean counts 

 of juveniles of the same year class at age 

 from estuarine nurseries north of Long Island, 

 and autumn spawners with juveniles from 

 estuarine nurseries south of Long Island. Of 

 six possible connparisons, all but two (both 



