The Inheritance of Egg Shell Color W. L. BLOW, C. H. BOSTIAN AND E.^W. GLAZENER North Carolina State College, Raleigh, N. C. ECONOMIC studies have shown definite consumer preference based on egg shell color. In North Carolina the State College Extension Service has observed that brown-shelled eggs which are sorted on the basis of color sell more rapidly than brown-shelled eggs which are unsorted and sold in mixed shades of brown. In most strains which lay brown-shelled eggs, variation in color is so great that sorting of different shades is not practical. The possibility that a uniformly dark shell color might be secured by breeding was considered. In order to determine the limitations of selective breeding, it is desirable to know something of the nature of the inheritance of egg shell color. Hurst (0) and Warren (0) have reported that the progeny of matings involving white- and dark-shelled breeds of chickens laid eggs of an intermediate color. Benjamin (0) found that dominance was not involved in the inheritance of egg shell color in White Leghorns and further concluded that both parents contribute equally to the inheritance of this character. Punnett and Bailey (0) developed a theory for the inheritance of shell color which involved one major and several minor factors. The progenies of matings made by KopeS () of Leghorn males X Orpington females were slightly lighter with respect to egg color than the progeny of the reciprocal cross which indicated that Contribution from the Department of Poultry Science, North Carolina Agricultural Experiment Station as Journal Paper. (Received for publication November, ) sex-linkage was involved. According to Axelsson () shell color is presumably controlled by several pairs of genes, part of which are sex-linked. Two or more factors were assumed by Hays () to be involved in the inheritance of egg shell color. The results of reciprocal crosses involving Rhode Island Reds and White Leghorns were accepted by Hall () as evidence of "sex-linkage. PROCEDURE The data in this study were assembled from the Barred Plymouth Rock and S.C. Rhode Island Red flocks at the North Carolina State College poultry farm. A total of females from dams and sires were classified for shell color from the pullet flocks housed in and. The breeds were housed separately in commercial type houses with access to green feed in enclosed lots. A North Carolina State College breeding mash was fed with a grain supplement. Eggs were classified for color intensity during a three day trapnesting period each week from the time the birds first reached sexual maturity. The data for the pullets hatched in were collected from a trapnesting period of two days each week. A standard composed of egg shells of varying intensities of pigmentation was used for the classification of color intensity. This standard was composed of eleven eggs, numbered from one to eleven. The tints were graduated from the lightest shade approaching white, beginning with number one, to the darkest shade which was cinnamon, numbered as eleven. The Downloaded from http://ps.oxfordjournals.org/ at Penn State University (Paterno Lib) on March, 0
W. L. BLOW, C. H. BOSTIAN AND E. W. GLAZENER standard set was replaced periodically since the color of egg shells is not permanent. Classification of the individual eggs for color was accomplished by matching them with the nearest appropriate shade in the standard. Phenotypes of the egg shell color character of the birds in this experiment were based on the eggs classified for color intensity during the first six weeks of observations for each bird. The mode rather than the mean of the observations during this period was selected as Egg color of dams 0 pie size. For this reason no frequency distribution or averages were determined for the sire progenies. Analysis of variance and regression analysis (Snedecor, ) were used to arrive at estimates of the heritability of differences. An analysis of variance for the egg color data is presented in Table. This technique involves a separation of the parental components of variance from the other sources of variation in the appropriate mean squares. The parental variances with certain assumptions im- TABLE. Frequency distribution of egg shell color for daughters by dams 0 ESS color of daughters the phenotype of the color character since this value tends to minimize extreme deviations from the true intensity caused by personal errors in classification. METHODS OF ANALYSIS 0 0 0 0 Total No. 0 0 Daughters Av. shell color......... Since no observations were made on egg shell color prior to, dam records were available only for those pullets hatched in. A frequency distribution based on the egg shell color of these birds and their dams is presented in Table. The egg character of the sires involved in this experiment quite obviously cannot be typed except from the performance of their female relatives. Large errors can occur in this type of inference; and when only small numbers of sires are available as there were in this study, the chances of errors being canceled are limited by samplicit may in turn be used to estimate the variance resulting from genes which act in an additive manner. If all dam families were of equal size, k and k' would be equal and their value would be the number of pullets per dam family. If in addition each sire family included the same number of dam familes k would be the number of pullets per sire. Hetzer et al. () gave formulae for the appropriate values of the k's when these conditions do not hold. The variation existing between birds which have the same sire and dam is measured by E. Since E+D represents the variation between paternal half-sibs, D is the statistic which measures the contribution of the dam to the likeness of the maternal progenies. Unrelated birds are characterized by a variation which is Downloaded from http://ps.oxfordjournals.org/ at Penn State University (Paterno Lib) on March, 0
THE INHERITANCE OF EGG SHELL COLOR represented by E+D+S, and S measures the paternal contribution since unrelated birds are this much more variable than birds which have the same sire but different dams. It is these maternal and paternal contributions which are used to estimate the additive genetic variance. These estimates of the additive component of variance may be influenced significantly by dominance or sex-linkage if such are involved in the inheritance of Source of variation Total variation Between years Between breeds within years Between sires within years and breeds Between dams within sires Between progency of the same dam and sire ance from the dam variance. The paternal effects, which include variation resulting from sex-linked genes, can be estimated by subtracting the dam variance from the sire variance. Of course whichever difference occurs may be due to errors of random sampling; the question is whether or not these differences are significantly different from zero. Heritability of differences is an expression of the fraction of the differences be- TABLE. Analysis of variance of data for egg shell color D.F. ** Mean squares are highly significant. k =. E =.0 k'=. D=0.0 k =. S=0. egg shell color. The estimate of the dam variance contains one-fourth of the additive genetic variance plus one-fourth of the variance due to dominance deviations. Likewise S, the sire variance, which estimates one-fourth of the additive genetic variance may also contain a component due to the involvement of sex-linkage with some of the genes influencing shell color. Environmental conditions which are not constant for different parental progenies may also contribute to the size of the dam and sire variance; however, such effects are generally of relatively small importance. The remainder of the genetic variance is contained in E with an additional part caused by environmental fluctuations which are strictly random with respect to individuals. Sum of squares,.....,0. Mean square.0.0.**.**.0 Components of the mean square E+k'D+k S E+kD An estimate of the strictly maternal effects such as dominance deviations can be obtained by subtracting the sire varitween a selected and unselected group of parents which can usually be expected between their progeny. This estimate is determined by the relative size of the additive portion of the genetic variance and the total variance. Genetic differences resulting from dominance are not involved. Since the sire and dam variance each contain one-fourth of the additive variance, four times these variances can be used as estimates of the additive portion. These estimates of the additive genetic variation are based on the assumption that the nonadditive portions of the parental variances are small. With that assumption implicit, estimates are made of the heritability of differences of egg shell color. These estimates of heritability are derived from the following formulae: D and S E+D+S E+D+S E Downloaded from http://ps.oxfordjournals.org/ at Penn State University (Paterno Lib) on March, 0
W. L. BLOW, C. H. BOSTIAN AND E. W. GLAZENER An additional estimate of heritability is provided by using twice the intra-sire regression of offspring on dam, a statistical technique elucidated by Lush (0). DISCUSSION It is apparent from the phenotypic relationships presented in Table that the dams are exerting considerable influence on the egg shell color of their offspring. With one exception an increase in shell pigmentation of the dams was associated with increased shell pigmentation in the offspring. With these averages as a criteria of the effectiveness of individual selection, indications are that rapid progress can be made in developing a darker shell color in breeds laying brown-shelled eggs. The egg shell data is partitioned into the degrees of freedom associated with the corresponding sums of squares in Table. When the F test was applied to the mean squares for dams and sires it was found that both F values were significant beyond the % level. The levels of significance indicate that real differences in egg shell color do exist between dam families and between sire families, differences which may be explained on a genetic basis. Appropriate computations based on a division of the parental mean squares into their respective components gave estimates of.0 and. for the dam and sire variances respectively. The larger size of the dam variance in this analysis suggests that dominance may be involved in the inheritance of egg shell color. The question is whether or not the difference in the size of the two variances is significantly larger than would be encountered in random sampling. The significance of this difference may be resolved by determining the joint probability of getting a mean square for sires as small as. associated with a mean square for dams as large as. if the parental variances are really equal. The joint probability of these two occurrences is less than one in twenty. This indicates that the difference is not due to an error in random sampling and that dominance is involved in the inheritance of egg shell color. No evidence is available in this analysis of variance that sex-linkage is involved. Examination of the parental variances does not preclude the likelihood of sexlinked effects since confounding of sexlinked and dominance effects in the parental variances prevents a clean separation of the variation due to these causes. Since the existence of sex-linked effects can not be disproved from this analysis, such effects may possibly be involved. The parental variances are used to arrive at estimates of the heritability of differences. If the dam variance is used, the heritability may be estimated at., a figure which may be considered an over-estimate due to the probable involvement of dominance. The estimate of heritability based on the sire component of variance is.0. An additional estimate of heritability is furnished by twice the intra-sire regression of offspring on dam. This estimate, based on dam-progeny groups, is.. The weighted average of the heritability estimates from the sire variance and the intra-sire regression may be accepted as the most reliable estimate of the heritability of shell color from these data. Since the intra-sire regression of offspring on dam is distributed with a variance estimated at.000 and the variance of the sire variance is estimated at.0, approximately thirty-one times as much weight is given to the estimate of heritability based on the intra-sire regression. The weighted average is. which approaches the estimate of heri- Downloaded from http://ps.oxfordjournals.org/ at Penn State University (Paterno Lib) on March, 0
THE INHERITANCE OF EGG SHELL COLOR tability derived from the intra-sire regression. This estimate of heritability indicates the rate of progress to be expected by selection among individuals. Thus if. is approximately the heritability of differences, the most probable change in egg shell color in the next generation would be approximately per cent of the deviation of the egg shell color of the birds selected as parents from the average of their generation. SUMMARY Estimates of the heritability of egg shell color in breeds laying brown-shelled eggs were made using the data from females from dams and sires. Dominance was found to be involved in the inheritance of egg shell color. All shell color genes are not necessarily affected by dominance, nor is dominance necessarily complete. No evidence of sex-linkage was available in this study. This does not preclude the likelihood of the existence of sex-linkage since dominance and sex-linked effects are confounded with the additive variance in the parental variances. The heritability of differences of egg shell color in breeds laying brown-shelled eggs is estimated at.0 from the sire variance. The estimate of heritability from the dam variance is. which is considered an over-estimate due to the presence of dominance. Twice the intrasire regression of offspring on dam gives an estimate of. for heritability of shell color. The weighted average of the estimates based on the sire variance and intra-sire regression is.. ACKNOWLEDGMENT The authors wish to thank Dr. R. E. Comstock, Department of Experimental Statistics, for his assistance in the techniques of analysis. REFERENCES Axelsson, Joel,. Variation and heredity of some characters in White Leghorns, Rhode Island Reds, and Barnevelders. Lunds Universitets Arsskrift. N. E. Bd., No.. Benjamin, Earl W., 0. A study of selection for size, shape and color of hens' eggs. Cornell University Memoir. Hall, G. O.,. Egg shell color in crosses between white- and brown-egg breeds. Poultry Sci. : -. Hays, F. A.,. Inheritance of egg size and egg character. Mass. Agr. Exp. Sta. Bull.. Hetzer, H. O., G. E. Dickerson, and J. H. Zeller,. Heritability of type in Poland China swine as evaluated by scoring. J. Animal Sci. : 0-. Hurst, C. C, 0. Experiments with poultry. Reports to the Evol. Comm. Roy. Soc. Rpt. II: p.. KopeC, Stefan,. An experimental study of xenia in the domestic fowl. J. Genetics :-. Lush, Jay L., 0. Intra-sire correlations or regressions of offspring on dam as a method of estimating heritability of characteristics. Amer. Soc. Anim. Prod., Proc. :-0. Punnett, R. C. and P. G. Bailey, 0. Genetic studies in poultry. II Inheritance of egg-color and broodiness. J. Genetics 0: p.. Snedecor, George W.,. Statistical Methods. Ames, Iowa. The Iowa State College Press. Warren, D. C, 0. Crossbreeding of poultry. Kansas Agr. Exp. Sta. Bull.. Downloaded from http://ps.oxfordjournals.org/ at Penn State University (Paterno Lib) on March, 0