Comportion of Iht Growth RotO 



0( Inb'Od onO OutOrtO Lorvot 



Of Ifn ft«ntfico" 0>»tt' 



6 B 10 12 



Ag« of Lorwoe <n Ooys 



Figure 5. — Comparison of the growth rate/of inbred and 

 outbred larvae of the American oyster. Crassostrea vir- 



i;inicii. 



inbred oysters is undoubtedly attributable to the 

 sporadically and unpredictable poor water quality at 

 the Milford Laboratory, much of it represents the 

 severity of the effect of enforced inbreeding of this 

 oyster. This is particularly significant in that the 

 inbreeding cultures were initiated with unusually 

 large numbers of eggs, as compared to the far smaller 

 population numbers available for studies with 

 species of higher organisms. 



Imai and Sakai (1961) detected and made some 

 measurements of inbreeding depression in C. i,'i,i;as. 

 Lines of C gii^'n-'i were lost in the third generation 

 of full-sib crosses. 



SELECTIVE BREEDING OF C. VIRGINICA 



Quantitative, commercially important traits are 

 controlled predominantly by exceedingly large 

 numbers of either of two types of genes. The effect of 

 one type is additive: the effect of the other is non- 

 additive. It is the additive type that responds to 

 molding and change by selective breeding. The herita- 

 bility of a trait is a measure of this additive genetic 

 variance as separated from the other type and from 

 the total phenotypic variance. Heritability estimates 

 can be used to predict progress by selection (see 

 formulae in Fig. 6). 



Theoretical heritability estimates should be read- 

 ily obtainable in the oyster in one generation for any 

 number of commercial traits by crossing the divided 

 lots of eggs of several females by several different 

 males. Contemporaneous cultures of such crosses 

 should provide enough full-sib. and maternal- and 

 paternal-half-sib families for an analysis of variance. 



From the results of this analysis heritability esti- 

 mates can be derived directly. 



This was attempted several times to estimate 

 heritability of growth rate for laboratory-reared C . 

 virginicti larvae and spat. Unfortunately, in spite of 

 its merits, this method did not prove successful at the 

 .Milford Laboratory. This is because these very 

 high larval mortalities experienced at the laboratory 

 so much of the time seldom leave enough contem- 

 porary culture families for statistical comparisons. 

 However, in a commercial-scale pilot hatchery at 

 the University of Oregon operating under better 

 environmental conditions heritability estimates for 

 several different characteristics of C. gigiis were 

 obtained recently (Lannon, 1972). It is also proba- 

 ble that C. gigas is more vigorous and easier to 

 handle in artificial culture than C. virginicci. 



In one series of these diallel crosses of C. virginica 

 at Milford enough larval families did survive long 

 enough in sufficient numbers to obtain a rough esti- 

 mate of heritability of larval growth to the larval age 

 of 2 wk. This estimate, 24%, is in the medium-to- 

 low range. 



Realized heritability estimates, as opposed to 

 such theoretical heritability estimates, can be ob- 

 tained from selection experiments. Selection exper- 

 iments can also tell something about the duration of 

 the response to selection, and about the limits of 

 selection response for different coetTicients of in- 

 breeding, and selection differentials. Such experi- 

 ments would fare better under the variable culture 

 conditions at Milford than the diallel crosses of the 

 theoretical estimates. 



BASIC FORMULA FOR SELECTIVE BREEDING 



Annual Progress for Trait 

 Being Selectively Bred For - 



Heritability = 



Heritability x Selection Differentiol 

 Generotion Interval 



Variation Oje to Additive Genetic Vorionce 

 Total PVienotypic Voriance 



Selection Differential - 



Difference Between Selected Oysters and Average of All Oysten 

 Fratn Which Selected 



Generation Intervol = 



Average Age of Oysters at Time of Reproduction 



Figure h. — Basic formulu for selecli\e breeding 

 of ov sler. 



80 



