250 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



TSH/WYTSCHA 

 GOFIBUSCHA  • 



KISUTCH 

 ALPINUS 



 MALMA 



—  MARSTONI 



— AUREOLUS 



S. KAMLOOPS 



— e. WHITEHOUSEI 



GAIRDNERI 

  NAMAYCUSH 



FONTINALIS 

 SALAR 



# =MEAN 

 1 = STANDARD DEVIATION 



TOTAL NUMBER OF VERTEBRAE 



Figure 5. — Mean number of vertebrae. 



Gill rakers 



Counts of gill rakers made by different investi- 

 gators are soniewliat more comparable than are 

 those of the vertebral counts. Even here, how- 

 ever, there seems to be some question concerning 

 the comparability of counts between fish of differ- 

 ent sizes. Thus Wilder (1952, p. 187) says that 



all the gill rakers on both limbs of the first gill 

 arch were counted including rudimentary rakers 

 sometimes present on large trout. He also writes 

 that — 



fhe exceptionally low raker count for Boeabec trout is 

 possibly a result of the low average size (115 mm. SL) 

 of the flsh In this sample as there is some evi- 

 dence to indicate that raker count increases with size in 

 salmonoids. . . . 



Foerster and Pritchard (1935b) write concern- 

 ing young Oncorhynchus — 



From Table 1, in which is presented a summary of the 

 average numbers of gill-rakers for each %-inch length 

 group for all species, it appears that in the very early 

 stages up to a length of 1% inches, there is an increase in 

 the number of gill-rakers with increase in size. Such a 

 change might be attributed to the overlooking of some of 

 the rudimentary rakers on the very small arches, but in 

 view of the fact that all counts were carefully made 

 under comparatively liigh magnification, it is unlikely 

 that such an error would have occurred. 



The available gill-raker counts for Oncorhyn- 

 chus are given in table 19. Obviously, the count 

 of 0. nerha. is significantly higher than that of gor- 

 iusclui, which in turn is significantly higher than 

 the counts of the remaining three species. Because 

 the counts for Oncorhynchus are all for mature 

 adults returning from the sea on a spawning 

 migration, the factor of size of fish on gill-raker 

 count may be entirely disregarded. 



If we disregard the two smaller samples of 

 tshaioyfsclia (li and 17 specimens), the largest 

 differences between means of samples of the same 



Table 18. — Numlier of vertebrae in North American Salmomdae 



' See bottom part of table 15 for treatment of these data. 

 > Based on adjusted values for Oncorhynchtis. 



Note.— Insofar as possible was put on basis of total number of vertebrae; 



for details see tables 15-17. 



