262 



FISHERY BULLETIN OF THE FISH AND WILDLIFE SERVICE 



.50 



O 



ffi 



K 



O 



« 40 



ffi 



< 



^. 35 



O 



I .30 



.2 5 



11..20 

 O 



5'5k 



LlJ 



s 



TSHAWYTSCHA 



• A-fr^a 



GORBUSCHA 



KISUTCH 



NERKA 



SALAR 

 _• L 



1 2 3 4 5 6 7 



AVERAGE FISH WEIGHT IN KILOS 



FiQUBE 17. — Avei-age weight of fry after absorption of the 

 yolk compared with the average total weight of the 

 species. 



In considering egg size in relation to fish weight, 

 however, it is obvious that Oncorhyixchus can be 

 distinguished even more clearly by this character. 

 Thus, in figure 17, in which the weight of fry with 



the yolk absorl>ed is plotted against the average 

 weight of the fish, S. mlar has small fry for the 

 size of the parent fish. In fact all five species of 

 Oncorhynchus except gorhuscha- fall in a straight 

 line. The larger size of the fry (and of course 

 the egg) of gorbuscha may be related to the ex- 

 treme degree of anadromy in this species, whereby 

 the fry emerge from the gravel as soon as the yolk 

 is absorbed and migrate seaward at once. 



DISTRIBUTION IN RELATION TO TEMPERATURE 



Species may range over a wide area and yet 

 avoid extreme conditions by changing spawning 

 seasons and by occupying different ecological 

 niches. A further complication is the tendency of 

 isolated populations to change genetically. 

 Despite these difficulties the overall picture shows 

 that some of the species are definitely arctic or 

 subarctic, whilst others range far to the south. 

 The approximate latitudes given in table 31 are 

 not too descriptive of the actual temperatures en- 

 countered because of the great differences in both 

 sea-water and fresh-water temperatures at com- 

 parable latitudes on different coasts and the com- 

 plicating factor of the lowering effect of altitude 

 on fresh- water temperature. 



Table 31. — Limits of ranges of North American Salmonidae, ranked according to temperature of water frequented 



In order to obtain a picture of the effect of 

 temperature on distribution, I have disregarded 

 latitude in favor of generalized temperature iso- 

 therms. The mean surface ocean temperatures 

 (see Davidson and Hutchinson, 1938) differ con- 

 siderably at comparable latitudes on the eastern 

 and western shores of the continent. In table 31, 



the water temperatures at the extreme ranges of 

 the distribution have been ranked subjectively by 

 species. This empirical method shows definite 

 trends when the species are grouped according to 

 their temperature distribution ( averaging both ex- 

 tremes of the range) . 



