248 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



Mottley's counts are chiefly on fry or finger- 

 lings 20 to 75 mm. in length. He stained the tis- 

 sues with alizarin and counted the last stained 

 centrum; since the urostyle did not stain it was 

 not counted. He writes — 



In making a comparison with the data of other investi- 

 gators, however, it should be noted that in the caudal 

 region, if the centra were stained as discrete blocks they 

 were counted separately, if the separation was not com- 

 plete the.v were counted as one. 



Because the last two or three vertebrae were not 

 always separated in the very small fish, he found 

 a slight tendency toward a lower vertebral count 

 in the smaller fry. Therefore, although his data 

 can be used for interspecific comparisons in Salmo, 

 they must be used cautiously in making compari- 

 sons with species of other genera. 



The maximum mean difference between any 2 

 of the 11 samples of Sahno gairdneri is 1.22 verte- 

 brae ( 64.56 minus 63.34 ) . Obviously S. gairdneri 

 and clarhl differ significantly from either salar or 

 trutta. "VVliether clarki and gairdneri or salar and 

 trutta can be distinguished by vertebral count can- 

 not be answered without additional data. 



For the genus Oncorhynchus, all available 



counts except those for two small samples of adult 

 tshawytscha were made by Foerster and Pritch- 

 ard (1935b) on unstained young ranging from 

 % inch to 3 inches in length. According to their 

 statement it would appear that their counts do not 

 include the three upturned vertebrae in the tail. 

 Furthermore, there is some reason to suspect that 

 the number counted is related to size. Table 15 

 gives the estimate of the statistical parameters 

 for the five species and it may be noted that the 

 variance was highest (7.84) for tierka, which has 

 the smallest fry, and smallest (2.20 and 1.44, re- 

 spectively) for gorhuscha and tshawytscha, which 

 have the largest fry. 



For nerka, the distribution of \'ei-tebral counts 

 is negatively skewed so that the mean, 63.73, is 

 about 2 counts below the mode (about 65.5). In 

 the bottom part of table 15 are shown the result- 

 ing estimates of the parameters for four species 

 of Oncorhynchus, when the counts causing this 

 extreme negative skew are disregarded. Altliough 

 tshawytscha^ shows the highest average count it 

 would seem unwise to use vertebrae as a distin- 

 guishing cliaracter between species of Oncorhyn- 

 chus until further data are available. 



Table 15. — Count of vertehrae in genus Oncorhynchus 



Note. Believe these are 3 vertebrae short of total number, as Foerster and 

 Pritchard say, ". . . the segments beginning with the one immediately 

 behind the skull and ending with the one immediately in front of the long 

 vertebrae projecting up into the tail can be counted". 



I Foerster and Pritchard (1935b); Cultus Lake, British Columbia, except 

 goTbuscha which were from Masset Inlet, British Columbia. 



" Townsend (1944); Oregon. 



s Recapitulation of estimated sample parameters rejecting counts below 

 62 vertebrae (see text). 



