CHAMBERS ET AL.: CORRELATIONS BETWEEN EARLY LIFE HISTORY TRAITS 



temperatures and developmental rates. Second, 

 the small particle substrate to which eggs adhere 

 is motile, mixing in all three dimensions with 

 wave action. Third, onshore winds promote 

 hatching and/or release of larvae from these 

 beaches, yet their frequency is highly variable 

 within and between years (Frank and Leggett 

 1981a, 1983: Leggett et al. 1984; Taggart and 

 Leggett 1987b). From the maternal perspective, 

 the time of the transition of her offspring from 

 beach residence to the pelagic mode is unpredict- 

 able owing to these three sources of variation. It 

 appears unlikely that a single optimal solution 

 exists for developmental rate, size at hatching, 

 or energy reserves. Additional examples of in- 

 trapopulation and intraclutch variation in early 

 life history traits, and of theoretical bases for the 

 maintenance of this variation in unpredictable 

 environments, is provided in Capinera (1979) 

 and Kaplan and Cooper (1984). 



Female Effects on ELH Traits 



Variation in egg quality is a sum of genetic and 

 environmental factors. We have no evidence that 

 egg size is heritable in capelin. We did not detect 

 a male effect or heritable variation in size at 

 hatching in another study involving 11 male 

 capehn each mated with 3 females (Chambers 

 and Leggett unpubl. data), yet there, as here, 

 significant differences among females were evi- 

 dent in both initial egg and hatchling sizes. Our 

 determination of a direct relationship between 

 condition and lipid indices of females, and initial 

 egg yolk volume probably reflects a predominant 

 environmental component to the observed varia- 

 tion in egg quahty (yolk volume) and its corre- 

 lates. The degree of influence from environ- 

 mental sources of maternal effects on ELH traits 

 probably depends largely on conditions experi- 

 enced by a female during the time that energy- is 

 being acquired and converted towards oogen- 

 esis, particularly her feeding rate (Hislop et al. 

 1978) and encountered temperatures (Tana- 

 sichuk and Ware 1987). If so, these largely non- 

 genetic maternal effects have the potential to 

 modulate the expression of critical ELH traits 

 and modify the survival pattern that would 

 otherwise be exhibited if the initial provisioning 

 of yolk reserves were at par throughout the pop- 

 ulation. Well-endowed individuals could thus be 

 buffered against selection in the yolk-dependent 

 period with, perhaps, residual effects later in 

 larval life. The inverse correlation between size 

 of mother and age at smoltification of her 



progeny in sockeye salmon, Oncorhynchus 

 nerka, as reported by Bradford and Peterman 

 ( 1987) may be such a residual effect. 



Covariation Within Individuals and 

 Variation Within Groups 



Relationships we observed between ELH 

 traits clearly show that simple correlations 

 based on group data may support, but may also 

 be inconsistent with inferences drawn from anal- 

 yses at the individual level. Only estimates of 

 variation and correlations between traits that 

 are based on individual-level observations are 

 gi'ounded on the same scale that natural selec- 

 tion primarily acts (Sober 1984). Caution must 

 be exercised when extrapolating from group dif- 

 ferences or correlations based on averages to 

 influences of one trait on another, to potential 

 advantages conferred on individuals, to dy- 

 namics of populations, or to evolution of life his- 

 tories. 



ACKNOWLEDGMENTS 



This paper is dedicated to the memory of Dr. 

 Reuben Lasker. The work was supported by 

 operating and strategic grants to WCL from the 

 Natural Sciences and Engineering Research 

 Council of Canada. We are grateful of the help 

 provided by D. Methven and the staff of the 

 Marine Science Research Laboratory, St. 

 John's, Newfoundland. 



LITERATURE CITED 



Bagarinao, T. 



1986. Yolk resorption, onset of feeding and survival 

 potential of larvae of three tropical marine fish species 

 reared in the hatchery. Mar. Biol. (Berl.) 91:449-459. 



Bailey, K. M. 



1984. Comparison of laboratory rates of predation on 

 five species of marine fish larvae by three planktonic 

 invertebrates: effects of larval size on vulnerabil- 

 ity. Mar. Biol. (Berl.) 79:303-309. 



Bengtson, D. A., R. C. Barkman, and VV. J. Berry. 



1987. Relationship between maternal size, egg diam- 

 eter, time of spawning season, temperature and 

 length at hatch of Atlantic silverside, Meiiidia 

 memdia. J. Fish Biol. 31:697-704. 



Blaxter, J. H. S. 



1969. Development: eggs and larvae. In W. S. Hoar 

 and D. J. Randall (editors). Fish physiologj', Vol. 3, p. 

 177-252. Acad. Press, N.Y. 



1988. Pattern and variety of development. In W. S. 

 Hoar and D. J. Randall (editors), Fish physiology, 

 Vol. IIA, p. 1-58. Acad. Press, N.Y. 



521 



