452 



FISHERY BXJLLETIN OF THE FISH AND WILDLIFE SERVICE 



2. That the annual sex ratio remains constant. 



3. That the regression of number of eggs on 

 size of fish does not vary between years. 



4. That the size and/or age at maturity does 

 not vary between years. 



5. That the number of eggs is a function of 

 fish size independent of age. 



6. Tliat there is no annual variation in the 

 proportion of the eggs retained by the females 

 in spawning. 



The foregoing assumptions are usually not 

 fully satisfied so that the variability of an ap- 

 proximate measure of reproductive potential in 

 critical experiments may be so large as to obscure 

 the very factors, the effects of which the biologist 

 is seeking to measure. 



This variability between numbers of mature 

 adults and actual reproductive potential has long 

 been recognized, and biologists have attempted 

 to discount it by substituting an estimation of 

 the total annual egg deposition for number of 

 adults as being a better measure of reproductive 

 potential. This paper is confined to an analysis of 

 the factors causing variation in the relation be- 

 tween number of eggs and number of mature 

 adults. 



After making the necessary allowance for dif- 

 ferences in size of fish, a wide range in fecundity 

 still exists between races of the same species from 

 different localities. For instance, McGregor 

 (1922, 1923a) found that the king salmon of the 

 Sacramento River have a far higher fecundity 

 than those of the Klamath River. Thus, if y 

 is number of eggs and x is length of the fish in 

 centimeters, the formulae for the regressions of 

 number of eggs on length are — 



Klamath River Log y'= .00682 .Y + 3.01 116 



Sacramento River Log K=. 00319 A' + 3.56836 



The Klamath River fish (65 specimens) ranged 

 from 61 to 107 centimeters in length (average, 

 82.6), with a geometric mean of 3,754 eggs. The 

 Sacramento River fish (50 specimens) ranged 

 from 59 to 110 cm. (average, 92.4) and had a 

 geometric mean of 7,298 eggs. At 85 cm., the 

 calculated geometric means for the two popula- 

 tions are 3,894 and 6,912 eggs, an increase of 78 

 percent in number of eggs for the Sacramento 

 River fish when compared with king salmon from 

 the Klamath River. 



The question arises as to the causes and the 



biological significance of such a great difference in 

 fecundity between populations of the same species. 

 It is recognized that harsher ecological situations 

 impose lower survival rates on some races. 

 Assuming that the number of eggs can be in- 

 creased by selection (as seems to have been done 

 for domesticated strains of trout), then the 

 number of eggs may well differ genetically between 

 various wild races of salmonids. In the case in 

 point there is good reason to believe that the 

 variation in egg number is not caused by variation 

 in the marine environment since, as McGregor 

 pointed out (1923b), Sacramento River and 

 Klamath River king salmon occur together in the 

 ocean troll catches. 



That the fecundity of fish of the same length 

 may even differ widely between populations 

 spawning in different portions of the same river 

 system is shown by Aro and Broadhead (1950) 

 for the sockeye salmon of the Skeena River. 

 For 3 years, 1939, 1948, and 1949, the female 

 sockeye of small Lakelse Lake (5.5 sq. mi.) 

 averaged 58.9 cm. in length (58.1-59.6) with an 

 average of 3,816 eggs (3,699-3,888); while for the 

 3 years of 1946, 1947, and 1949, the female sockeye 

 of the large upriver Babine Lake (171.8 sq. mi.) 

 averaged 58.5 cm. (57.1-60.1) in length with an 

 average of only 3,181 eggs (3,056-3,389). 



In assessing the significance of differences in 

 fecundity between fish of various localities, it 

 becomes important to measure the variation within 

 localities. Some of the important factors within 

 localities to be considered are — 



1. Size of \ the fish in relation to number of 



eggs- 



2. Age at maturity. 



3. Size of the eggs. 



4. Seasonal trends in fecundity in the same 

 locality. 



5. Annual variation in fecundity. 



RELATION OF SIZE OF FISH TO NUMBER' 

 OF EGGS 



Combining his own observations with those of 

 Titcomb (1897), Ricker (1932) states that the 

 relation between number of eggs and- length of 

 fish is curvilinear for the eastern charr, or brook 

 trout, Salvelinus fontinalis. The number of eggs 

 varied from 80 in a 5.1 -inch charr to 5,630 in a 

 22-inch charr. However, Allen (1956) points out 



