UNIVERSITY OF CALIFORNIA COLLEGE OF AGRICULTURE AGRICULTURAL EXPERIMENT STATION BERKELEY, CALIFORNIA BREEDING CHICKENS FOR MEAT PRODUCTION V. S. ASMUNDSON and I. MICHAEL LERNER BULLETIN 675 November, 1942 UNIVERSITY OF CALIFORNIA BERKELEY, CALIFORNIA CONTENTS PAGE Growth of the poultry meat industry in California 4 Market requirements for meat birds 4 Grades 5 Color preferences 6 Explanation of terms used in breeding 6 Principles of selection 7 Mass selection 9 Pedigree breeding 9 Family selection and progeny testing 10 General remarks on selection 11 Problems common to all breeding programs 12 Factors affecting meat quantity and quality 13 Growth rate 14 Growth patterns 17 Seasonal effects 18 Conformation 18 Percentage of edible flesh 23 Fat in the muscles 23 Kate of feathering 25 Differences between breeds 26 Differences within varieties 27 Differences in back feathering 29 Breast defects 29 Crooked keel 30 Breast blisters 30 Flesh pigments 32 Economy of gain 33 Crossbreeding 34 Hybrid vigor 35 Sex linkage 37 Eecommended breeds and crosses 39 Broilers 39 Fryers and light roasters 40 Heavy roasters and capons 40 Breeding programs and changing market requirements 41 Literature cited 43 BREEDING CHICKENS FOR MEAT PRODUCTION 1 V. S. ASMUNDSON 2 and I. MICHAEL LEENER 3 In the last few years a persistent demand for information on breeding chickens for meat production has arisen in California. While experi- mental results in this field are as yet limited in scope, a considerable I fc 50 Q 40 > 10 3 § 1 \ 1 \ 1 \ /&7 / 4 * \ / / \ \ / / \ \ / / \\ / t \\ / > \\ / / i // \ 7 V t \ \ /9c 19 +'' ^ ^ / N s Ja 1ft zb.Ai fa A ir.M 1 auJi >neJi i/uAl 'Q.St ptC ktA bi/.D ec. Fig. 1. — Hatchery sales within the state of California. Note the increase in sale of baby chicks during the summer and fall months in 1939 over 1927. This increase is due in part to out-of-season hatching for fryer production. (Data from Tinley and Voorhies, 1940.) literature on isolated phases of the subject has been already built up. This bulletin, issued to meet the immediate needs of California poultry- men and breeders, reviews and interprets the pertinent literature in the light of California conditions. 1 Keceived for publication March 24, 1942. 2 Associate Professor of Poultry Husbandry and Associate Poultry Husbandman in the Experiment Station. 3 Assistant Professor of Poultry Husbandry and Assistant Poultry Husbandman in the Experiment Station. [3] 4 University of California — Experiment Station GROWTH OF THE POULTRY MEAT INDUSTRY IN CALIFORNIA Much of the poultry marketed in California is a by-product of the market-egg industry. There are several facts, however, that point to increased production of poultry for meat. These may be briefly sum- marized as follows : ( 1 ) while, according to census information, birds on hand (at any given time) in California decreased slightly from 1934 to 1939, the number of birds raised increased by over six million; (2) hatches are being spread more uniformly throughout the year, owing partly to an increasing tendency to replace pullets as needed and partly to the year-long production of fryers (fig. 1) ; (3) an increased per- centage of dressed poultry receipts on the Los Angeles market are from California points ; (4) increased interest in and inquiries about poultry for meat production. MARKET REQUIREMENTS FOR MEAT BIRDS The requirements for meat birds on California markets are, with the exceptions noted below, those common to meat birds wherever marketed. For that reason the classification and tentative specifications adopted by the Bureau of Agricultural Economics, United States Department of Agriculture 4 have been followed, with slight modifications in the mini- mum ages of fryers and roasters to allow for more rapid growth com- monly obtained in specialized fryer production. Exact classifications differ on the various markets ; for example, all birds weighing from 2 to 4 pounds may be referred to as fryers on some markets. Dressed chickens are classified according to kind, age, sex, and weight as follows : a) Young birds: Broilers. Young chickens, approximately 8 to 12 weeks old, of either sex, weighing not over 2% pounds and sufficiently soft-meated to be cooked tender by broiling. Fryers. Young chickens, approximately 9 to 20 weeks old, of either sex, weighing more than 2% but not more than 3^2 pounds, and sufficiently soft-meated to be cooked tender by frying. Roasters. Young chickens, approximately 4 to 9 months old, of either sex, weighing over 3% pounds, and sufficiently soft-meated to be cooked tender by roasting. 4 United States Bureau of Agricultural Economics. Classification and tentative specifications for United States standards and grades for dressed chickens. 8 p. Re- vised March, 1938. (Mimeo.) Bul. 675] Breeding Chickens for Meat Production 5 Stags. Male birds, of any weight, with flesh slightly darkened and toughened and with comb and spur development showing the bird to be in a state of maturity between that of roasting chickens and cocks. Capons. Unsexed male birds weighing over 4 pounds, usually 7 to 10 months old, and with soft tender flesh. b) Old birds: Fowl. Mature female birds of any age or weight. Cocks. Mature male birds of any weight, with darkened and tough- ened flesh. Hens culled from the laying flocks are marketed as fowl, while surplus Leghorn cockerels are marketed as "broilers." Most of the birds produced especially for meat are raised to fryer or light roaster size, or to about 3 to 4 pounds. On some markets, such as New Orleans, a 2-pound broiler is preferred, but in California, fryers are preferred, although there is little difference in the price paid per pound for fryers and for broilers. Poultry are also classified in other ways according to methods of prepa- ration and handling of the birds prior to their sale to the consumer. Grades. — There are four quality grades for dressed chickens. These grades are based on the quality of the bird when killed and on the effi- ciency of the methods of killing and dressing. l.TJ. S. Special, or U. S. Grade AA. Young, fine-grained, soft-meated birds with broad full-fleshed breast, the entire carcass fully covered with fat and with skin soft and glossy lying close to the flesh. These must be well bled, well dressed, and free of pinf eathers. No flesh or skin bruises allowed, and only slight skin abrasions or discolorations permitted, none of which shall be on the breast. No crooked breasts or other deformities allowed. A broken or disjointed wing above the wing tip or a broken or disjointed leg are not permitted. No open tears or torn skin permitted. Birds must be dry-picked or semiscalded, and dry-packed. 2. U. S. Prime, or 77. S. Grade A. Young, soft-meated birds, with well- fleshed breast, the entire carcass well covered with fat, and with soft glossy skin. They must be well bled, well dressed, and practically free of pinfeathers. No flesh bruises and only very slight skin bruises, abrasions, or discolorations permitted, none of which shall be on the breast. No crooked breasts or other deformities allowed. Brokens wings above the wing tips or broken legs not permitted. Torn skin not per- mitted. No sewn skin permitted on the back. Birds with crops properly removed and sewn up may be included in this grade. Birds must be dry-picked and semiscalded and dry-packed. 6 University of California — Experiment Station 3. V. 8. Choice, or U. S. Grade B. Young, soft-meated birds, with fairly well-fleshed breast, and with carcass fairly well covered with fat. They must be fairly well bled and dressed, and may show few scattered pin- feathers over the entire carcass. Slight flesh or skin bruises, abrasions, or discolorations permitted, but not more than three such defects on the breast. Abrasions or tears over 2 inches in diameter not allowed unless properly sewn up. Dented or slightly crooked breast bones or other slight deformities permitted. One broken wing or one broken leg in the flesh permitted if the bone does not protrude through the flesh and if not showing excessive bruise or blood clot. 4. V. S. Commercial, or TJ. S. Grade C. Young birds with poorly fleshed breast and with carcass poorly covered with fat. They may show evidence of poor bleeding and have numerous pinfeathers over the entire carcass. Abrasions and discolorations permitted and hunchbacks or other de- formities allowed if birds are fairly well fleshed. Birds badly bruised so as to make any appreciable part of the carcass inedible or birds emaci- ated or showing external evidence of disease not permitted. Color Preferences. — Well-fleshed and feathered fryers of the yellow- skinned colored breeds are preferred in California, but no premium is paid for yellow-skinned birds nor deduction made for white-skinned birds. A higher price per pound is usually paid for colored than for white-plumaged birds. This unjustified discrimination is one of the reasons why White Plymouth Rocks and other white-plumed birds have failed to become popular in California although they are widely used in other states for broiler, fryer, and light-roaster production. EXPLANATION OF TERMS USED IN BREEDING The following section is reprinted from Bulletin 626 (Taylor and Lerner, 1938 ) 5 of the California Agricultural Experiment Station. It may be suggested that the reader familiarize himself with the contents of that companion bulletin before proceeding with the present publication. Inheritance of characters is based on the transmission of genes from parent to offspring. Genes are defined as the determiners of hereditary characters. They occur in pairs (with a few exceptions), one member of each pair being derived from the sire, and its mate from the dam. One member of the pair may be different from its mate, as in the case of a cross between two breeds.. For example : A bird resulting from a cross between a purebred rose-combed and a single-combed bird will receive a gene for rose comb from one parent and for single comb from the other. The crossbred bird will exhibit a rose comb, since the rose-comb gene is dominant and prevents the expression of the single-comb gene, which is known as a recessive. If two such crossbred birds are mated 5 See "Literature Cited" for complete data on citations, which are referred to in the text by author and date of publication. Bul. 675] Breeding Chickens for Meat Production 7 together, in the next generation both rose- and single-combed individuals will appear in the ratio of three rose-combed birds to one single-combed bird. Of the rose-combed birds, one third will be pure-breeding, and two thirds will not breed true (possessing like their parents, one rose-comb gene and one single-comb gene). The birds which possess two like genes and breed true are homozygous for this gene. In the example given, the pure-breeding rose-combed and the single-combed birds are homozygous for comb-shape genes. Those rose-combed birds which possess two unlike genes will not breed true and are heterozygous. Various genes are usually designated by arbitrary letters. Capital letters indicate the dominant condition, and small letters the recessive: the gene for rose comb, for example, is represented as E and that for single comb as r. The actual constitution of an individual with respect to these genes is known as the genotype, which is desig- nated by the genes of which it is composed : the genotype of the single-combed bird, for example, is represented by rr, that of the pure-breeding rose-combed bird by EE, and that of the heterozygous rose-combed bird by Er. The appearance of the bird with respect to these characters is called the phenotype. In the example given, there are only two phenotypes — single-comb, to which the bird with the rr genotype belongs, and the rose-comb, to which the birds with both the EE and Er genotypes belong. The two genotypes that show the rose-comb phenotype can be distinguished only by breeding tests. If a rose-combed bird, bred to a single-combed bird gives only rose-combed offspring, then it must belong to the EE genotype, but if it gives some single-combed offspring, then it must belong to the Er genotype. At times, when two unlike genes are present in the pair governing the expression of a character, neither gene completely supresses the action of its mate. The case is then one of incomplete dominance. When this occurs, the mating of two such heterozy- gous individuals produces three phonotypes in the offspring, which correspond to the three genotypes present. In cases where characters are produced by the action of a large number of genes, such are known as multiple genes. Sex-linlced inheritance can probably be best explained by an example. A simple sex-linked character in poultry is the barred plumage of the Barred Plymouth Kock. When a male of this variety is crossed with a black female, all the progeny is found to be barred. In a reciprocal cross, however, when a black male is mated to a barred female, in the first generation, the females are found to be black, while the males are barred. Sex-linked crosses for early sex identification are based on this principle. The reason for the behavior of the sex -linked genes lies in the fact that the male possesses a paired complement, as in the case of all other genes, while the female has only one member of a sex-linked gene pair. Thus the male genotype in the case of the pure Barred Plymouth Bock is BB, while the female genotype is B-. In the case of the black birds, the male's constitution is bb and the female's b-. In the second cross as described above, bb x B- (black male x barred female), the male progeny will be heterozygous barred (Bb) and the females will be black (&-). PRINCIPLES OF SELECTION There are essentially five steps in the development and application of a breeding program : 1. Development of methods of measurement of the characteristics of individuals. 8 University of California — Experiment Station 2. Determination of the characteristics and measurements of the ideal type. 3. Actual measurement of individuals. 4. Analysis of the individual records on basis of the group they belong to (family, strain, or breed) . 5. Selection of superior representatives to be mated. As far as breeding chickens for meat production is concerned, the greatest difficulty has been in relation to the first two steps. Yet they are both indispensable. While at the present time, in the absence of complete information, makeshift substitutes are being used, rapid progress in the improvement of meat birds cannot be expected until the necessary knowl- edge is acquired. Many ideas exist as to what characteristics an ideal meat bird should possess. In the majority of cases, these ideas are purely subjective. They can be illustrated on actual birds, but cannot be trans- lated into the language of measurement. In a character such as egg pro- duction the number of eggs laid by a bird in a given period of time represents a figure which can be compared to similar figures obtained for other birds. In the case of meat type such an objective index of worth is not available. This in part may be attributed to the failure of the pack- ing industry to provide the breeders with sufficiently precise specifica- tions, but largely it is because of the complexity of the problem involved. Considerable research work is being carried on at the present time on just what constitutes a superior dressed carcass and how to recognize it in a living bird. When a practical solution is obtained for this problem the work of the meat breeder will be greatly facilitated. What is defi- nitely needed is the determination of the typical measurements and characteristics of birds which approach the ideal agreed upon. These measurements must, of course, reflect the visual appearance, the per- centage of edible flesh, and the growth of the birds. The third step of the breeding program represents the actual application of the first two steps. The methods used in the fourth and fifth steps outlined are common to all breeding programs. The type of procedure depends on the level of operation. Thus the commonest and earliest method used is that of mass selection. In a more advanced breeding stage the consideration of ances- try of the individuals selected for breeding purposes is also needed. This may be designated pedigree breeding since it involves the identification of stock hatched and the keeping of pedigrees. Finally, the most ad- vanced breeding stage is that comprising family selection and progeny testing. Each of the successive stages implies the incorporation of the Bul. 675] Breeding Chickens for Meat Production 9 methods of the earlier stages and, in addition, the use of special methods for the higher level of breeding operations. Mass Selection. — This method is based on breeding from stock superior in the characteristics for which selection is practiced. In other words, the selector chooses the best of his birds to mate with each other. The individual excellence of the breeding birds selected is the only criterion used. Nothing has to be known with regard to the origin of the birds, nor whether or not they transmit their desirable characteristics to their off- spring. It is obvious that improvement under this system of breeding is very uncertain. The appearance of the bird does not always correspond to the inherited factors which the bird possesses. Thus, two rose-combed birds may produce single-combed birds in their offspring. Similarly two white birds mated may yield colored progeny. Thus, the appearance of the bird (its phenotype) does not necessarily correspond to the inherited factors which the bird possesses (the genotype) and transmits to his or her offspring. Consequently, selection on the basis of the phenotype does not necessarily lead to the results obtained by selection on basis of the genotype. In the next two stages of the breeding problem, both pheno- typic and genotypic selection are practiced. In addition to the method involving the mating of the best to the best, another type of breeding which is often used is that of compromise mating. This involves the use of birds excelling in one or more characters in mating with birds which are deficient in these particular characters but which possess other superior characteristics which in turn may be lacking in their prospective mates. Pedigree Breeding. — At the next level of breeding procedure, an at- tempt to evaluate the genotype of the birds to be selected is made. Since the genotype cannot be determined from the bird's appearance, reliance has to be placed on breeding performance either of the bird itself or of the bird's ancestors. It is the latter type that is involved in the simplest form of pedigree breeding. When the characteristics of the parents of a bird are known to be desirable and then the bird itself duplicates these qualities, one is justified in assuming that the probability of the given bird carrying concealed undesirable factors in its genotype is less than when no information on the ancestry is available. If information on more than one generation is also on hand, such probability is further reduced. Consequently, the principle of selection at this stage is that the birds used in mating should show an uninterrupted lineage of desirable an- cestors. Indeed, it was this method of selection that has resulted to a great extent in the improvement of most of our domestic animals includ- ing poultry. It must, however, be clear that this method provides only an 10 University of California — Experiment Station estimate or approximation to the knowledge of the bird's genotype. Full sisters and brothers differ both in phenotype and genotype, owing to segregation of genes as explained in the bulletin already referred to. The simple form of pedigree breeding where selection is made on basis of ancestry does not take account of this fact. Family Selection and Progeny Testing. — A more precise way of eval- uating a bird's genotype is to base an estimate of breeding worth on the actual performance of the progeny of a given mating. Whatever are the phenotypes of two mated birds, if they have desirable genotypes, they will produce superior offspring. Of course, it is generally more likely that a superior phenotype is associated with a superior genotype. That is to say, a desirable looking bird may have a desirable or an undesirable genetic constitution ; a bird lacking in desirable characters will seldom if ever have a superior genotype. Hence, the problem of the selector in this breeding phase is to test the superior appearing individuals for their genetic constitution. If they carry the desirable genes, their off- spring will exhibit the desired qualities. Such breeding birds should then be remated as long as they reproduce themselves efficiently. Their offspring in turn should be progeny-tested and the operation repeated as long as a breeding program is carried on. In the absence of a progeny test, the sib test may be used. The word sib is a contraction of sister and brother and is used to indicate the group of both when no distinction as to sex is relevant. Thus all of the offspring of a given male mated to a given female are designated in relation to each other as sibs or siblings. It should be apparent that when a bird has uni- formly good sibs, it is likely that the genotypes of the parents were su- perior (the sib test of a bird is, of course, at the same time the progeny test of the parents). The superior sibs may then be subjected to a prog- eny test themselves. It may be noted that in the transition from the second to the highest level of breeding a shift in emphasis has been made. Whereas in mass selection and in the simpler form of pedigree breeding it was the indi- vidual that was considered, in the progeny and the sib tests the family as a whole is being evaluated. The word family is used here to designate a group of sibs. The term sire family may be used to denote the offspring of one male and several females (full and half-sibs). The key to this shift in emphasis lies in the segregation of genes, as mentioned before. Since no given individual is likely to breed true for all of the many characters which make a superior bird, there is always segregation in its offspring. Accordingly, the offspring vary in degree of excellence, but by considering their average worth, a fair estimate of the genotypes Bul. 675] Breeding Chickens for Meat Production 11 of the parents may be obtained. It follows clearly that the larger the family is, the more accurate is such an estimate. There are, however, two qualifications in this general rule. First, it can be shown statistically that the reliability of the estimate depends not on the number of indi- viduals but on the square of that number. This means that to double the precision of the estimate a fourfold increase in number of progeny is necessary; to triple the precision of the estimate, the number of offspring must be increased ninefold, and so forth. Secondly, an increase in the number of progeny raised means that the hatching season has to be lengthened. In the case of characters affected by date of hatch (such as rate of growth, sexual maturity), the environmental or nonhereditary variation is increased. Unless this added variation is corrected for or taken in account, the variability in the characteristics of a family be- comes too great for an adequate genotype analysis. Thus a compromise must be reached between increasing nonhereditary variability by length- ening the hatching season and decreasing the number of progeny tested from each mating in a shorter season. General Remarks on Selection. — To summarize the discussion on selec- tion, it may be said that an efficient breeding program comprises pri- marily the recognition of individuals of superior worth in the sense of their ability to transmit desirable characters to their offspring, and of mating such individuals to produce a desirable combination of charac- ters in the next generation. A non-trap-nesting breeder or multiplier relies on the visual appearance of the birds for identification of superior stock. This method is not so efficient as the one employed by a trap-nester who can marshal the information about the ancestry and the progeny of the birds concerned to aid him in the procedure of such identification. The multiplier makes his selections anew every year. The trap-nesting breeder is enabled to repeat the successful matings of previous years and to utilize the better families and strains generation after generation. Both the multiplier and trap-nest breeder must keep records. The prob- lem of the fryer breeder, for instance, is to produce birds which are of superior quality at the time of marketing for fryers. Since it is not possible to tell by looking at an adult bird whether it was early or late feathering, rapid or slow in its early growth, records must be made on the prospective breeding birds at the appropriate ages with respect to such characters. The breeder who pedigrees his stock is able to identify the individual birds by means of wing bands. The multiplier or nonpedi- greeing breeder must use a scheme of leg banding to ensure that the identity of the superior birds at the marketing age will not be lost by the time such birds are ready for the breeding pen. 12 University of California — Experiment Station PROBLEMS COMMON TO ALL BREEDING PROGRAMS Any sound poultry breeding program must take into consideration the fundamental factors of hatchability, egg production, and viability. All of these affect the number of birds that can be produced, and a high level of all three is essential for efficient production. As is the case with any character, the actual hatchability, egg production, and viability observed in a flock are the results of interaction of hereditary and environmental factors. Sound feeding and managemental practices are prerequisite to any production program. In all of the present discussion this is implied. Hence, no specific repetition of this fundamental point will be made, with a few exceptions (see, for instance, the discussion of crooked keels) . The principles involved in breeding for improved hatchability, pro- duction and viability as well as the methods used are dealt with in detail in Bulletin 626 (Taylor and Lerner, 1938). The only reiteration neces- sary here is to emphasize the complex nature of all of the characters. Thus, hatchability is influenced by many genes, acting, perhaps, at dif- ferent times in the course of the embryo's development. Selection against lethal genes (genes, which cause embryo mortality) is a necessity in flocks where such genes occur. Egg production is also not a unit char- acter : early maturity, lack of winter pause, high rate, lack of broodiness, and high persistency must all be selected for to ensure improvement in egg records. Similarly, the factors for viability may be specific ; that is to say, that selection against mortality from neoplastic diseases (fowl paralysis, tumors, iritis, and associated conditions) will not necessarily lower mortality from other causes. An elementary selection program may use as a standard of selection the net result of the action and interaction of all these various factors. This means that a breeder may use the percentage of hatchability, the average egg production and the percentage survival as the only bases upon which he selects his breeding birds. Such a program may lead to im- provement, but a more certain way would consist of an attempt to analyze which one of the particular characters is the limiting factor in the breed- er's progress. For instance, a given family average production may be low because of late maturity; another family with the same average record may exhibit low rate of production. By identifying such com- ponent factors, a breeder is enabled to exercise much more intelligent judgment in making up matings, than he can in the more elementary pro- cedure, and progress should be correspondingly accelerated. Exactly the same principle holds in relation to meat production. A breeder may select on the basis of a total quality estimate of a family or Bul. 675] Breeding Chickens for Meat Production 13 strain. He may also consider the particular factors which are respon- sible for the differences in quality between groups of birds. There are also differences in efficiency of production and in yield, as well as in quality. The fundamental problem for the meat breeder, in the light of the previous discussion, is then to be able to identify the bases of these differences. So far as the inherited factors are concerned, economic pro- duction of a large number of birds of superior quality depends, in addi- tion to hatchability, egg production, and viability, on such characters as growth rate and pattern, conformation, rate of feathering, and free- dom from defects. These together with a number of other pertinent characters will be discussed in the following section. FACTORS AFFECTING MEAT QUANTITY AND QUALITY Much technical information has been gathered concerning the factors that affect meat quantity and quality. Some of these factors are of great importance and deserve careful study. Others are of negligible impor- tance. The amount of work done has not always been in proportion to the need for information about a particular factor ; hence, we have much corroborative or contradictory information about some phases and little about others. The following discussion naturally reflects these discrep- ancies in our knowledge of the subject. The weight of the newly hatched chick depends primarily on the weight of the egg from which it hatched. The correlation coefficients measuring this relation reported in the literature vary from + 0.73 ± 0.013 (Benjamin, 1920) to + 0.95 (Graham, 1932). Galpin (1938) found small but significant variations in the correlation between egg and chick weight at different seasons, although the seasonal changes in chick weight resulted at least partly from changes in egg weight. Halbersleben and Mussehl (1922) reported that chicks weighed on the average 64 per cent of the weight of the eggs. It is thus obvious that the weight of chicks cannot be expected to exceed two thirds of the weight of the eggs set. When the weights of day-old chicks from different hens are grouped together, no statistically significant difference can be demonstrated be- tween the weight of males and females. Munro and Kosin (1940) have, however, shown that the males weigh slightly more than the females by comparing the weights of chicks of the two sexes from the same hen and in the same hatch. Initial chick weight has no influence on the rate of growth of chicks up to 8 to 20 weeks of age. The weight of the chick about 24 hours after it hatches is therefore of no importance to the producer of broilers or fryers, provided the chicks were hatched under normal conditions. If the chicks 14 University of California — Experiment Station are small because the humidity was excessively low or the temperature too high, this may result in too great a mortality among them. Such con- ditions usually also result in low hatchability and, therefore, should be carefully avoided. While the weight of day-old chicks is of little practical importance so far as the fryer producer is concerned, it is nevertheless desirable to produce chicks of average size, weighing about l 1 /^ to 1% ounces (35 to 43 grams) each. There are three reasons for this : (1) Small chicks may be out of small strains of birds and may prove slow-growing. Usually, how- ever, differences in the weight of chicks merely reflect differences in the weight of eggs from pullets and hens of different ages. (2) Medium- sized chicks are hatched from medium-sized eggs, which hatch better than eggs that are much below or above the average in weight. (3) Aver- age-sized chicks make a better first impression on the buyer than under- sized chicks. GROWTH RATE The size of an immature animal depends primarily on two factors : the length of time the animal has been growing, and the rate at which it grew. Initial weight in the case of chicks has, as it has been already stated, little influence. While there may be inherited differences in the duration of growth between birds of different breeds, strains, or families, the im- portant genetic variable from the standpoint of a poultry meat producer is rate of growth to market age or market weight. Inherited ability to grow fast or slow varies from breed to breed, as well as within a breed ; this accounts for size differences at different ages. There have been many figures published to illustrate this point. Perhaps the most extensive studies on breed differences have been conducted at the Missouri Agri- cultural Experiment Station. Kempster (1941) has presented the aver- age weights of five breeds of pullets obtained in the course of ten years. It is not desirable to put much reliance on averages based on totals of many hatches produced in different years. Furthermore, Kempster fails to give the number of birds on which the computations are based. Never- theless, his figures are reproduced in table 1 because they present a typical picture of the course of growth of the pullets of the different breeds. They are, of course, specific to the strains of birds and to the con- ditions of feeding and management under which the birds were raised. The main point to be noted from the figures is the gradual decrease in gains after the age of 12 weeks. This decrease appears to be related to increasing size and occurs in most strains of birds under ordinary condi- tions ; this is of great importance to the meat producer. As will be shown Bul. 675] Breeding Chickens for Meat Production 15 later, not only do the increments in weight decrease after 12 weeks but the efficiency of gains is consistently reduced with age. Growth rate may be measured in many different ways. Thus it may be expressed by the formula : Wo—W 1 Rate= X 100 i/ 2 (W 2 +W 1 ) where W 2 is body weight at the end of the period considered, and W x the weight at the beginning of the period. This formula may be explained TABLE 1 Growth of Pullets from Hatching Time to 40 Weeks of Age* Age in weeks Single Comb White Leghorn White Plymouth Rock Rhode Island Red New Hampshire White Wyandotte Weight Gainf Weight Gain Weight Gain Weight Gain Weight Gain 4 pounds 0.09 41 1.01 1.67 2.20 2.56 2.98 3.26 3.41 3.46 3.50 pounds 0.32 .60 .66 .53 .36 .42 .28 .15 .05 0.04 pounds 0.09 0.39 1.07 1.88 2.55 3.09 3.59 4.22 4.69 4.92 5.04 pounds 0.30 .68 .81 .67 .54 .50 .63 .47 .23 0.12 pounds 0.09 0.41 1.07 1.89 2.63 3.19 3.73 4.33 4.88 5.24 5.46 pounds 0.32 .66 .82 .74 .56 .54 .60 .55 .36 0.22 pounds 0.10 0.48 1.25 2.27 3.17 3.85 4.38 4.89 5.25 5.40 5.54 pounds 0.38 0.77 1.02 0.90 0.68 0.53 0.51 0.36 0.15 0.14 pounds 0.09 0.44 1.14 2.10 2.94 3.45 4.03 4.69 5.06 5.22 5.00 pounds 0.35 8 .70 12 ,. 16 .96 .84 20 .51 24 .58 28 .66 32 .37 36 .16 40 -0.22 * Data from Kempster (1941). t Gains are obtained by direct subtractions of weight at previous weighing. best by assuming that the weight of the bird is equivalent to so much capital increasing at a certain rate of interest. If the interest is added to the principal at the end of a given period, what is known as simple inter- est obtains. If the interest is compounded continuously throughout the period, the compound-interest formula may be used. For the sake of sim- plicity, one may add the interest once at the moment when the principal plus the interest is halfway between the original and the final capital. The formula given above permits the calculation of the interest rate on this basis. Applying it to growth data on Barred Plymouth Rocks and Single Comb White Leghorns the rate of growth is found to be as ex- pressed in table 2 (Asmundson and Lerner, 1934). The figures shown are expressed in terms of per cent for the complete period so that it may be said, for instance, that the rate of growth of Leghorn males from 2 to 8 weeks of age was 140.9 per cent. It may be 16 University of California — Experiment Station seen from this table that the early growth of the Plymouth Rocks exceeds that of the Leghorns, and that the males grow faster than the females. In the last period of growth the Leghorns actually increased at a higher rate than the Plymouth Rocks, yet in these particular strains at 24 weeks of age the latter outweighed the former by better than 1% pounds in the case of the males, and by somewhat less than 1% pounds in the case of the females. This point is of double importance : the early differences in growth rate determine to a large extent final size differences; and the producer of broilers or fryers is particularly interested in adequate early growth, since he desires to market his product at the earliest possible TABLE 2 Average Growth Rates for Barred Plymouth Rocks and Single Comb White Leghorns* Males Females Age Barred Plymouth Rocks Single Comb White Leghorns Difference Barred Plymouth Rocks Single Comb White Leghorns Difference weeks 2-8 per cent 148.6 98.2 33.6 per cent 140.9 90.5 34.3 per cent 7.7 7.7 -0.7 per cent 145.6 91.2 32.6 per cent 136.5 82.4 34.5 per cent 9.1 8-16 8.8 16-24 -1.9 * Data from Asmundson and Lerner (1934). age. An important consideration involved is, however, the fact that within a breed, strain differences may exist which are not necessarily reflected in the final size of the birds. These may be designated as differ- ences in growth patterns and are of great importance. They will be dis- cussed in the next section. A demonstration of inherent growth rate differences within a breed has been provided by the study of Asmundson and Lerner (1933). Six families of Single Comb White Leghorns from the same flock, raised together under the same management were found to fall into three distinct classes with respect to their growth rate from 2 to 8 weeks of age. The three rapid-growing families had a rate of 143.36 per cent, two inter- mediate families had a rate of 139.86 per cent, and one slow family a rate of 133.21 per cent. The differences between the three classes were significant. Schnetzler (1936) has demonstrated that such differences are amenable to selection. He weighed a group of Barred Plymouth Rock chicks at 8 weeks of age, and when they reached maturity made two matings, one Bul. 675] Breeding Chickens for Meat Production 17 of the fast-growing birds and one of the slow-growing birds. The off- spring produced by the two groups showed distinct differences in their weight at 8 weeks of age. This is a clear indication that even mass Selec- ts 4'A 3'A .S5 2'/z I. I'A '/4 ) A BJ>R. / / / / S.CML-A-t / / 6 £ S.CML-4-2 / / / j ^5L ZHL ■&2 / T // '/ A f V' 6 '/ V / i >' gt- yz R nrn° H P 0/mr xttth Qnr £ i Sssbred cochsre /s 4 5 w pounds / 2 3 /./t/e weight Fig. 10. — Kelation of weight gains to feed consumption. (Adapted from Titus, 1940.) The implication is, of course, that the sooner the animal reaches market- ing weight, the more profitable it is, a conclusion which, doubtlessly, is obvious. Table 8, prepared from data presented by Jull and Titus (1928) , illustrates this particular point. It is readily seen that the gain per unit of feed consumed decreases with age (last two columns of the table). Titus (1940) has presented a graph which shows this decrease to follow a straight line (fig. 10) . CROSSBREEDING Crossbreeding is the mating of birds from different breeds, such as Plymouth Rock x Rhode Island Red, Cornish male x Plymouth Rock fe- male, and others. It is commonly used also to denote matings of birds Bul. 675] Breeding Chickens for Meat Production 35 from different varieties, such as White and Barred Plymouth Rocks. Outbreeding is the mating of unrelated birds of the same variety. Crossbreeding is widely used in the production of poultry for meat. In general the results obtained vary according to the genetic constitu- tion of the breeds and strains crossed. For this reason there may be important differences in the results obtained from crossing two different strains of one variety with a strain of another variety. HYBRID VIGOR When different breeds are crossed it is frequently observed that fer- tility, hatchability, livability, and rate of growth are increased. This phenomenon, which is known as heterosis, or hybrid vigor, is usually the only valid reason for crossbreeding. If no hybrid vigor is obtained when breeds are crossed, there is usually no advantage in making the cross. The one exception to this may be the crossing of Cornish male on females of other breeds to produce a more attractive carcass. The crosses that have been made may conveniently be divided into three groups: (1) crosses of comparatively small breeds, such as the Leghorn, with heavier breeds, such as the Plymouth Rock, New Hamp- shire, Rhode Island Red, Jersey Giant, Australorp ; (2) crosses between the heavier breeds or different varieties of those breeds; (3) crosses of Cornish or other heavy nonfighting game-type males onto females of other breeds. The Cornish females are not used because they are usually poor layers. The results of crossbreeding may be considered in relation to fertility, hatchability, chick mortality or survival, growth of the chick, feed re- quired per pound of gain, and feathering. The data in tables 9 and 10 indicate that crossbreeding has little or no effect on fertility where aver- age fertility is obtained in the strains used. On the other hand, hatch- ability generally is improved by crossbreeding. This has been shown by the work of several investigators (Warren, 1930; Byerly, Knox, and Jull, 1934; Funk, 1934; Hess, Byerly, and Jull, 1941). At the National Research Center (Knox, 1939) it was found that if the hatchability of the parent breeds was low there was a considerable increase from cross- breeding, whereas crossbreeding resulted in little or no increase in hatchability when the hatchability of eggs from the parent breeds was high. The mortality to 8 weeks among crossbred chicks is lower, on the aver- age, than the mortality among the chicks of the breeds and strains crossed. There are exceptions (Bice and Tower, 1939; Horlacher, Smith, and Wiley, 1941) but the best comparative data (Warren, 1930) where 36 University of California — Experiment Station males of the same breed were alternated with males of a different breed, leave no doubt that on the average a higher percentage of the crossbred chicks survive. Here, as in the case of hatchability, there is considerable improvement in survival if the viability of the parental strains crossed is low but little or, in some cases, no improvement is obtained by crossing if the viability of the parental strains is high. The rate of growth of crossbred chicks on the average exceeds that of the parent breeds to broiler or fryer weight when the heavy breeds are crossed. When Leghorns and general-purpose breeds are crossed the crossbreds may exceed both parental varieties to 8 or even 12 weeks of TABLE 9 Summary of Eesults of Crossbreeding in the Arkansas Experiments^ Breeds or crosses Fertile eggs Hatch of fertile eggs Chicks surviving to 8 weeks of age Average weights at 8 weeks Average weights at 12 weeks Feed per pound of gain to 12 weeks Bare- back chicks at 8 weeks Leghorn X Leghorn Leghorn X Cornish, Wyandotte, or Australorp. . per cent 89.7 76.5 83.6 82.4 per cent 65.7 85.0 80 3 81.7 per cent 92.5 89.3 86.9 94.1 pounds 1 20 1.21 1.18 1.28 pounds 2.17 2.25 2.24 2.42 pounds 4.66 4.41 4.84 4 34 per cent 1.45 12.12 12.67 25.90 * Data from Horlacher, Smith and Wiley, 1941. t Rhode Island Reds, Rhode Island Whites, Barred and White Plymouth Rocks. age, but after that the heavier breeds are likely to weigh more. The re- sults of crossing the Japanese Shamo Game or the Dark Cornish on other breeds may be expected to be the same so far as growth is concerned as if one of the commoner American or English breeds were used. The breast conformation of the crossbreds will, however, resemble that of the Cornish or Game breed and they will have the compact feathering char- acteristic of the Cornish (Jaap, 1941). Heterosis or hybrid vigor, as measured by growth, will depend on the strains crossed (Knox, 1939). The most rapid-growing strains of birds do not always give the best results when crossed. Nevertheless, when the same male was crossed with females of two different breeds the progeny of the hen from the more rapid-growing breed made the better gains to 8 weeks of age (Warren, 1930). Crossbreds utilize feed more efficiently on the average than birds of the parental strains (Hess, Byerly, and Jull, 1941). The data in tables 9 and 10 show also that the crossbreds required less feed to produce a pound of gain. They presumably utilize feed more efficiently than pure- breds because of more rapid growth and not because they are more effi- Bul. 675] Breeding Chickens for Meat Production 37 cient converters of feed into flesh than the parental breeds, although this has not been fully established (see the section "Economy of Gain"). It may be assumed, nevertheless, that rapid growth and low mortality are the most important factors in efficient feed utilization. Leghorns have fewer bare-backs than the crossbreds or heavy breeds (see data of Horlacher, Smith, and Wiley in table 9), as would be ex- pected. Crossbreds from Leghorn hens mated to heavy-breed males feather out at an earlier age on the average than crossbreds between heavy breeds. The results apparently depend partly on the heavy breed used, since the crossbreds from Leghorn female x Australorp male were TABLE 10 Summary of Eesults of Crossbreeding in the Hawaii Experiments* Breeds or crosses Fertile eggs Hatch of fertile eggs Chicks surviving to 8 weeks of age Average weights at 8 weeks Feed per pound of gain to 8 weeks per cent 75.6 76.9 76.9 74.4 per cent 53.6 65.7 60.3 63.0 per cent 86.6 92.7 92.4 96.3 pounds 1.29 1.46 1.37 1.50 pounds 3.97 Leghorn X Japanese Shamo Game 2.58 3.41 Rock and Red females X Shamo Game males 2.67 ' Data from Bice and Tower (1939). t Japanese Shamo Games, Barred Plymouth Rocks, and Rhode Island Reds. slower in feathering than most of the crossbreds between heavy breeds. The slowest-feathering birds were from White Wyandotte hens mated to Rhode Island Red males (82.81 per cent bare-backs at 8 weeks of age), a result that may again reflect strain, rather than breed differences, al- though the progeny of the reciprocal cross had very few bare-backs (4.05 per cent) at 8 weeks. Apparently heterosis, as such, is not a factor, or is only a minor factor, in rate of feathering. SEX LINKAGE Sex-linked plumage-color factors can be used to distinguish males from females. The most readily available crosses are from black males mated to barred females (for example, Jersey Black Giant male mated to Barred Plymouth Rock female) and "gold" males mated to "silver" females. For the latter cross, females having the Columbian, silver-laced, or silver-penciled patterns, such as the Columbian Plymouth Rock, Light Brahma, Light Sussex, Silver Laced Wyandotte, or Silver Penciled Wyandotte, are mated to red, buff, or black-red males such as the Rhode 38 University of California — Experiment Station Island Red, Buff Orpington, or Dark Cornish. Males of the latter breeds can also be mated to Barred Plymouth Rock females. When black males are mated to barred females the male progeny have the characteristic white head spot of barred birds, while the females are without the spot. Similarly, the male progeny of a silver (Columbian) female mated to a gold (red) male will be silver or white in color, while the female will be predominantly red or buff in color. Since some pattern genes, such as those carried by silver-penciled varieties, obscure the differences between males and females, not all the varieties that fall into the various groups are equally useful for crossing. The most clear-cut results are obtained by mating black males to barred females — although any nonbarred male may be used except White Leg- horn, White Plymouth Rock, and some White Wyandottes; and by mating red, buff, or black-red males to Columbian-pattern females. To avoid crossbreeding and yet use the sex-linked genes for distin- guishing day-old males from females, so-called "auto-sexing" varieties have been developed. These usually have the dominant sex-linked gene for barring superimposed on brown or red. None of the varieties de- veloped so far are of any special interest to the producer of meat birds. Barred Plymouth Rock males can be distinguished from females at the day-old stage with an accuracy of better than 90 per cent ; hence they may be regarded as a reasonably good auto-sexing breed and variety, although it was not, of course, developed for that purpose. Since the development of methods for sexing day-old chicks by cloacal examina- tion, the use of sex-linked crosses has become of less practical importance. Age at first egg is an inherited character determined by sex-linked and autosomal (non-sex-linked) genes. Early sexual maturity is domi- nant to late sexual maturity; hence, it would be advisable to use males from early-maturing strains if the pullets are to be kept for egg produc- tion. Actually, however, egg production depends on so many non-sex- linked factors that no one has a major effect. The report of Knox (1939) and others indicate that the crossbreds are usually intermediate but that sometimes they exceed both parents. The average annual egg production of hybrids and the parental breeds as reported by Warren (1930) is as follows : Eggs per bird White Leghorn 173.8 Jersey Black Giant 162.3 Hybrids 212.9 Rhode Island Red 168.9 White Leghorn 211.6 White Leghorn $ x Rhode Island Red