Frequency of first occurrence of the haemal brace on vertebrae 



Nos. 14 to 18 



Locality 



14 



15 



Vertebra number 



16 



17 



18 



British Columbia 



Southern California 



Soledad 



Viscaino 



Cape 



Gulf 



Totals 



161 



1,163 



L 222 



14 



The differences between the four northern 

 regions are not significant . Differences between 

 Viscaino and Cape regions were "extreme" and 

 differences between the two southern regions, 

 Cape and Gulf, were pronounced. Differences 

 between samples from the same area were not 

 significant for four localities, were on the bor- 

 derline of significance for southern California 

 (P = .01) and were significant for Viscaino 

 (P .0001). The variation in Viscaino was asso- 

 ciated with differences between age classes, the 

 divergent sample being of small fish of the year . 



Morphometric comparisons . - -Head and 

 fork lengths were measured for most samples 

 (not taken on British Columbia specimens). When 

 head length was plotted against fork length, no 

 departure from rectilinearity was found, and the 

 regression lines were of the type Y = a + bx . 

 Statistics describing the regressions of head 

 length on fork length for each area are given in 

 the following table: 



to reflect differences between age groups and/or 

 recruits from different spawning grounds. The 

 samples from other regions are not different 

 "within regions" . However, all five regions were 

 found to differ significantly from one another in 

 head length, the differences are far greater than 

 those associated with meristic characters . From 

 this morphometric study Roedel concluded that 

 each region was characterized by distinct groups 

 of fish; those in California and Soledad with rela- 

 tively short heads, those to the south with 

 progressively longer heads. 



In summary, fish from the four northern 

 areas were not considered separable on the basis 

 of vertebral characters . Head length data were 

 not available for British Columbia. For the re- 

 maining three regions, the comparisons indicate 

 that the fish were in all probability drawn from 

 three statistically distinguishable stocks. Vis- 

 caino and Cape fish were sharply set apart in all 

 respects. The Cape fish were characterized by 



Differences among southern California 

 samples are significant. This could be inter- 

 preted as evidence for two or more genetically 

 separable populations, or it could be considered 



a more posterior appearance of the vertebral 

 structures and by relatively longer head. Gulf 

 fish were strongly differentiated from Cape fish, 

 and form a unit apart on the basis of Roedel's study. 



64 



