Table 1. — Mean vertebral numbers of male and female juvenile Atlantic menhaden 



fran different localities, 1956-59 



Date and locality- 



Males 



Mean 



S(X-X)2 



Number 



Females 



Mean 



S(X-X)2 



Number 



1956 



Childs Ptiver, Mass 



Reeves Bay, N.Y 



Indian Mver, Del 



Little River, N.C 



St. Catherines Sound, Ga. 



1957 



Childs River, Mass 



Saugatuck River, Conn. . . 



Indian River, Del.""- 



Neuse River, N.C 



Edisto River, S.C 



1958 



Weir Creek, Mass 



Ball Creek, Va 



Indian Field Creek, Va. . 

 Broad Creek, N.C 



1959 



Indian River, Del 



Ball Creek, Va 



Indian Creek, Va 



Broad Creek, N.C 



46.66 

 47.18 

 46.83 

 47.12 

 47.06 



46.79 

 46.50 

 47.15 

 47.02 

 46.90 



46.80 

 47.05 

 47.07 

 47.05 



47.05 

 47.03 

 46.91 

 47.00 



14.83 



6.55 



4.50 



12.40 



12.81 



11.60 



12.86 



9.72 



3.90 



21.79 

 23.95 

 21.4^ 

 12.00 



54 

 44 

 18 

 42 

 47 



60 

 65 

 58 

 39 



77 

 75 

 64 

 55 



46.52 

 47.14 

 47.01 

 47.09 

 47.13 



46.82 

 46.62 

 46.97 

 46.98 

 47.00 



46.76 

 47.03 

 47.10 

 47.10 



46.86 

 47.04 

 47.03 

 47.09 



11.98 

 10.86 

 22.99 

 14.57 

 12.08 



12.55 

 14.98 

 6.97 

 10.98 

 10.00 



10.87 



4.97 



13.62 



23.39 



2.59 

 4.96 

 4.97 

 9.64 



40 

 56 

 81 

 58 

 53 



44 

 55 

 37 

 48 

 48 



38 

 34 

 42 

 59 



22 

 25 

 36 



44 



P Less than 0.05. 



used in analysis of vertebral variation between 

 the sexes. Of the 18 samples tested, 14 posi- 

 tive and 4 negative regression coefficients 

 were obtained, two of which differed sig- 

 nificantly from zero (1 of these was positive 

 and significant at the 5-percent level of prob- 

 ability and the other negative and significant 

 at the 1-percent level). These results indi- 

 cate that fish length may be a source of 

 variation; however, since no consistent pattern 

 was evident, length was not considered in 

 subsequent analyses. 



Variation among samples and 

 localities 



Vetebral counts of 17,024 juveniles ranged 

 from 44 (1 specimen) to 49 (7 specimens) 



with the greatest number having 47 vetebrae 

 (12,281). Vertebral counts, by samples, locali- 

 ties, and dates of collection, are given in 

 appendix tables 1 through 4. 



The mean numbers of vertebrae in all sam- 

 ples of the 4 year classes were tested for 

 homogeneity by analysis of variance (table 2). 

 Heterogeneity of means within and between 

 localities was indicated by the variance ratios 

 of the 1956 and 1957 year classes. Figure 2 

 shows that, on the basis of mean vertebral 

 numbers, the fish were separated into two 

 major groups,^ one extending north of latitude 



^Marr (1957) defined group as "a fractionof a popula- 

 tion with distinctive characteristics, the natureof which 

 (phenotypic or genotypic) has not yet been determined." 



