231‘ if“;v ‘3‘”. VS Wi‘f‘ffifi‘é A C 5.33 iShtifim "3 73‘ ”Stash: {er i410 Same sf (.5. 3 NaECEflGAN STAT: cm. VLL. Gauge Bambi Swag; This is to certify that the , , I ‘~ (9v. thesis entitled ~ ~ '_ VARLATIONS WITHIN A STRAIN or RHODE ISLAND REDS I » J \ IN THEIR REQUIRIIIITIT FOR VITAMIN 1312 I - 7 , .. presented by E ' George Bentley Sweet Q I ‘ ' r ' has been accepted towards fulfillment v " ' of the requirements for l" 1.1 . fl Master of Science degree in Poultry Husbandry I “ e. l I . - sr’ EL . ', .1 I i . _ Major professor .2, Date September 25, 1952 ( 4 |-. t ‘J g o v v . ’ t'l o ‘ \ ‘ > .,~ , 0169 V o}’~ '. I, —— - _. —v— .. w - . s . " >1 .‘ ‘1 A 1 I I . \ I J / 'n‘ . I,’ ' ’ l i l r, I" i I I / .ll \- I I! l \ I - | ‘3 I , (\ (.- r. ‘ I It. /’ ' \ , I f t .- . ‘ I I w a - ' . , l \ ~_ ’ l ‘. } ~ 'I"\ 2 I ,_ l. l t . I .J' l I Mi fl... - $.4— n.. . . *-O—” o VARIATIONS WITHIN A STRAIN OF RHODE ISLAND REDS IN THEIR REQUIREMEKT FOR VITAHIN B12 by GEORGE BENTLEY SWEET Mm».— A THESIS Submitted to the College of Graduate Studies of Nichigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of vMASTER OF SCIENCE Department of Poultry Husbandry 1952 THESQS ‘J‘ 7".» 7', t AC 1’3? 0 :‘.’LED GET ZEN T S C’The author wishes to express his deep appreciation to Dr. Albert C. Groschke,* formerly of the Department of Poultry Husbandry, for suggesting this investigation and for his assistance in organizing and carrying on the experiment. Grateful acknowledgement is also due to Dr. William D. Eaten, Professor of kathematics, for his immeasurable and patient help in developing a statistical analysis for the data and for the generous gift of his time. The cooperation of Dr. Lloyd R. Champion of the De- partment of Poultry Husbandry and his guidance in edit- ing this work and assisting in the experimentation are deeply appreciated. Many thanks are due to David Libby for his assistance and helpful discussions of the experimentation. The author also wishes to thank C. G. Card and J. A. Davidson of the Department of Poultry Husbandry for their constructive criticisms during the writing of this study. *Now with the Borden Company, New York, New York. Introduction Purpose of Study Previous Work CONTENTS Inheritance as a factor influencing efficiency of food utilization Experimental Procedure I. II. III. Results I. II. III. IIIa. IV. IVa. Vb. VI. VIa. Hearing of Parental Stock and Selection of Breeders Plan of Nat Statistical and Discussi Variations Variations Variations Variations Variations Variations Variations Approach 1 Variations ings Analysis on Between Hens Between Hatches Between Pens Between Like Sires Between Like Sires Between Like Pens Between Unlike Ratings Between Unlike Matings Approach 2 Variations Between Unlike Matings Approach 5 Variations Variations Between Unlike Sires Between Unlike Pens Summary and Conclusion Bibliography PAGE 29 50 32 55 4O INTRODUCTION The basis of all knowledge is experience. In science it is experimentation. The raw materials of experimenta- tion are variations. Causes of variations in biological studies are differences in heredity, with which the indi- viduals Start life, and differences in environment, known and unknown, to which they were exposed during their de- velopment. The interaction of these two main causes may have a joint effect which can not be ascribed to either one alone. The lack of ability to reproduce experimental results can probably be attributed to this effect of interaction, especially in environmental experiments which have no pre- cautions controlling the individuals' heredity either by uniform selection of individuals or by breeding for uni- formity. Variation that is ascribed to environment alone may be the combined results of both heredity and environment. Lack of uniformity of any biological factor in indivi- duals of a population is certainly expected. Selection of uniform individuals to be used in an experiment arouses the curiosity. Can that uniformity be obtained by breeding? Basic genetic principles are involved; specific genotypes allowing the development of a degree of expression of a certain characteristic. The cause of the lack of uniformity is the heterozygosis of the individuals. By phenotypic selection, relatively homozygous individuals can be ob- tained in some future generation. A great deal of work is being carried on today con- cerning hereditary resistance and susceptibility to cer- tain diseases. Hereditary differences in anatomy causing variations in some physiological processes are also being studied. Whether it is a difference in anatomy or a difference directly caused by a specific genotype, the difference in individual efficiency of utilization of a specific nutrient has received very little attention experimentally. The experiment presented in this thesis concerns itself with a study of this type, namely a study of the variations in two-week weights of chicks fed a vitamin B12 deficient diet. PURPOSE OF STUDY' The purpose of this study was to determine if varia- tions in two-week weights of chicks on a vitamin B12 de- ficient diet could be partially attributed to heredity; and if so, to test the possibility of developing strains relatively resistant and relatively susceptible to a de- ficiency of vitamin B12 in the diet. PREVIOUS WORK Inheritance as a.Factor Influencing Efficiency of Food Utilization An intensive genetic study on the efficiency of utilization of a specific nutrient was made by Lamoreux and Butt (1948). Based on relative resistance or suscep- tibility of chicks to a riboflavin deficiency, selective breeding was practiced for six generations. The analysis of the progeny mortality and weight data showed lower mor- tality and higher weights in the resistant strain as com- pared with the mortality and weights of the susceptible strain. Chicks from both strains made comparable growth gains when fed an adequate diet. These researchers be- lieved that between the two strains there were genetic differences which were specifically concerned with the utilization of riboflavin. The data of an incomplete study by Lerner and Bird (1948) supported the results of Lamoreux and Butt with respect to genetic differences in the chicks' response to a riboflavin deficiency. Many nutritional researchers have demonstrated or suggested breed and strain differences in the dietary re- quirement of certain nutrients. These differences have been especially prominent in biological assays in which the response of the animals is used to determine the poten- cy of a specific nutrient within the supplement fed the animals.. Biological assays Which utilize chicks as experimental animals have another variable to contend with, i.e., the amount of the specific nutrient in the egg at hatching time as it may vary between breeds, strains, fami- lies, or even individuals within a family. Selecting on the merit of efficiency of food utiliza- tion, Morris, 32 al., (1935) inbred nine generations of rats. A line for high efficiency and a line for low efficiency were developed. Each generation was compared statistically with the second generation and significant differences were found. The average level of efficiency of food utilization appeared fairly uniform in both lines after the sixth genera- tion. The evidence of this study supported the belief that complex heritable factors influence the efficiency of food utilization. Strains of rats fed a low vitamin D, high calcium diet were found by Gowen (1936) to have an average life of five to six months, half of the average life span of rats on a A. 0 H. (D '55 C1. D; H. CD (‘1‘ Cf 5" CO :3 normal diet. Nales lived longer on this defi the females, while on a non-deficient diet the females lived longer. The evidence of this experiment points to distinct genetic differences in the dietary requirements of the dif- ferent strains. However, resistant and susceptible strains to vitamin D dietary deficiency were not developed. Ringrose and N rris (1956) found strain differences in chicks used for the biological assay of vitamin A. U Breed differences in resistance to a deficiency of thiamine were observed by Lamoreux and Hutt (1939). In each of four experiments the ability of White Leghorn chicks to survive on a diet deficient in thiamine was greater than that of Rhode Island Reds. The progeny of a White Leghorn male and Rhode Island Red females were inter- mediate between the parental lines with respect to a defi- ciency of thiamine. White Leghorns w re also found to be superior in re- sistance to manganese deficiency when compared with New Hampshires in a study by Gallup and Norris (1939). Within the two breeds there were also strain differences in manga- nese requirements. Golding, et al., (1940) fourd White Leghorns superior to Barred Rocks in the low requirement of manganese. Studies in the nutrition of the mouse by Fenton and Covgill (1947) showed that one highly inbred strain of mice had a low requirement for riboflavin, while another highly inbred strain had a high requirement. Indications resulting from a study by Bethke, et al., (1936) show that the variability of vitamin G in eggs from different hens is due in part to strain differences. The majority of the variability came from the various levels of vitamin G in the diet of the hen. Hens' diets also affect the vitamin E content of the egg (Barnum, 935) and the vita- min A content of the egg (Ellis, 1955). a modified macro-fermentation assay procedure, Using ‘J Scrimshaw, at al., (1945) found that the mean thiamine con- tent of axis from White Lejhorns was significantly greater than the thiamine content of eggs from Rhode Island Reds and Barred Rocks, the latter two having comparable stores of thiamine in their eggs. All these hens were kept under -' comparable conditions and had the same thiamine content in their diets. Jackson, 33 31., (1946) found variations in the con- centration of riboflavin in the eggs from various breeds and strains on the same ration. There were also variations be- tween hens of the same strain. Bird, at 31., (1946) have clearly shown that the mani- festations of animal protein factor deficiency in the hens' diet are poor hatchability and low viability of their chicks, indicating that the stores of this factor in the egg are decreased when the diet of the hen is deficient in this fac- tor. Studies by Bird, 33 31., (1947) show that chicks vary widely with respect to their stores, at hatching time, of the animal protein factor and their ability to withstand a deficiency. Hens fed a deficient diet for eleven months were found to vary in the hatchability of their eggs. Hens characterized by high and by low hatchability were not found to vary with respect to the growth promoting pronerties of their excreta or their efficiency of food utilization. The proEeny of hens characterized by high, intermediate, and low hatchability, respectively, showed high, intermediate, and low viability and growth rate to six weeks of age. The effect of dietary supplements upon the growth of chicks .varied with.the maternal diets and with the ability of the dams to withstand the dietary deficiency. EXPERIMENTAL PROCEDURE I. BEARING OF PARENTAL STOCK AND SELECTION OF BREEDERS The initial preparations for this study began in Jan- uary, 1951, when the replacement stock of Rhode Island Beds for that year was hatched. The parental stock (P1) was housed in nine different pens, with fifteen females and one male per pen. Seven hundred.F1 chicks were hatched in Jan- uary and seven hundred.F1 chicks were hatched in February. The chicks were wingbanded and brooded in batteries on a standard chick starting mash. Mortality was recorded daily and individual weights, to the nearest gram,.were ob- tained at the end of two weeks. At four weeks of age the chicks were removed to a brooder house. Six weeks later the flock was culled heavily. The birds were ranged on good pasture from.May to August, at which time the breeders were selected and housed. Selection was based on pedigree estimates of the breeders. These estimates were obtained from the records of the two-week body weights of the breeder and body weights of its sibs from the two replacement hatches. Two-week body weight records of the chicks from.two previous hatches fed a vitamin B12 deficient diet also contributed to the estimates. All P1 individuals were classified as Resistant (R) or Susceptible (S) based upon the reaction of their progeny obtained from four hatches. The F1 individuals were lO classified as R if their sibs were in the top 20% of the weight range and suffered less than 25% mortality, or S if their sibs were in the lowest 20% of the weight range and had more than 25% mortality. II. PLAN OF MATINGS The plan of each experiment was a modification of the conditions of Fisher's (1925-1944) "Latin Square" design. The modification is that these were polyallel matings in- volved in each experiment and analyses were made within the polyallel mating arrangement instead of rows, columns, and treatments. Contrary to procedure used in Drosophila, i.e., the female parent is listed first, the symbols used herein follow the procedure of the animal breeders where the male parent is listed first. The subscript number refers to a specific male and the subscript letter refers to a specific pen. The designs of the experiments are shown in Figure l on the following page. 108 Figure l - Designs of Experiment I and Experiment II Experiment I Pen A (S) Pen B (S) Pen C (R) Pen D (R) EEEEE’ Male Female Male Female Male Female Male Female 1 and 2 51 x Sa 82 x Sb R1 x Rc R2 x Rd 5 and 4 82 x Sa 81 x Sb R2 x R0 R1 x Rd 5 and 6 R2 x 39. R1 x Sb 52 x Re 31 1: Rd '7 and 8 R1 x Sa R2 x Sb 51 x Re 82 x Rd Eeriment II Pen E (S) Pen F (S) Pen G (R) Pen H (R) m Male Female Male Female Male Female Male Female 1 and 2 S3 x 83 R4 3: Sf R5 x Rg S4 x Rh 3 and'4 R3 1: Se S4 x Sf 33 x RE; R4 x Rh 5 and 6 R4 x Se 35 x Sf S4 x Rg R5 x Rh 7 and 8 S4 x Se R5 x Sf R4 x Rg S5 x Rh 11 Two pens, each.housing 20 R pullets, and two pens, each housing 20 S pullets, were set up in Experiments I and II. Two R cockerels and two S cockerels were rotated in each of the four pens of each experiment. Two hatches were obtained from each mating. Males were removed from the pens and placed in individual cages on approximately the 28th day after their initial entry into the pen. Two days later, semen was collected from the next male to be introduced in the particular pen and pullets of that pen were artificially inseminated with 0.2 cc. of semen. If the volume of semen was not sufficient on the first collection, the male was returned to the battery and the remaining in- seminations took place the following afternoon. The male was then placed in that pen after all the pullets had been inseminated. Eggs laid after a lapse of 36 hours were saved for hatching. It is generally accepted that the newly in- troduced sperm fertilizes the ova (Warren and Gish, 1943; Warren and Kilpatrick, 1929). Eggs were set at intervals of approximately twelve days. All chicks were pedigreed and hatchability, fertility and abnormal embryos were recorded. The chicks were placed in batteries equipped with raised bottoms and fed a basal diet deficient in vitamin B12, (Table 1). The wing band numbers of all dead chicks were recorded each day and the survivors were weighed at two weeks of age. Table l - Composition of Basal Chick Diet 12 Soybean Oil meal...............o........... Ground yellow corn......................... Dehydrated alfalfa meal...................o Steamed bonemeal............0...’.......... Oyster shell flour......................... B-Y'feed (500 micrograms riboflavin per gram)..........0......C................. FiSh Oil (400 D 2000 A)ooooooooooooooooooo salt (iOdized).O..........;.....‘C......... ChOline chloride........................... NicOtinic ECidoooooQOooooooococo-o.oooooooo Manganese sulfate...............o.......... *d O E o. to 0am oo o o (DCDCHDCDC) iJCRUHDCD ... 01-04' I CDCHJCfiGDUI'CfiCDC)$WD ODO NHm 100.027 Calculated analysis: Protein.................... 2606 % 03101umoooooooooooooooooooo 105 % PhosphorUSooooooooooooooooo 0.8 % 13 III. STATISTICAL ANALYSIS As stated previously, the design of the matings was to simulate a "Latin Square" where each cell was analyzed respectively and later combined in all pertinent compari- sons. The Analysis of Variance according to the procedure of Fisher (1924), as presented by Snedecor (1946), was used to test significance. However, in the analyses of variance only one criteri- on of measurement can be used. The R line had greater weights and very little mortality, while the S line had low weights and a high mortality. It was recognized that mor- tality is a necessary criterion of selection and can not be eliminated in any analysis. Therefore, the experimenter chose the one criterion of weight, assuming the weight of every dead chick as 30 grams. It is logical that the chick showing vitamin B12 deficiency symptoms will decrease in initial weight; the initial weight of a chick is approxi- mately 40 grams or 65% of the weight of the egg (Jull and Quinn, 1925). It is postulated that because no chick weights were recorded under 31 grams, the assumed weight_of 50 grams is reasonable for all chicks that died during the two week period. It is logical that the use of the figure of 30 grams as a constant for all dead chicks is superior in an analysis to giving these chicks no weight at all and yet including them in the average. If a female of a pen was represented in a hatch by 14 only one chick that chick's performance was eliminated from the data. One hundred and sixty-two analyses of variance were made of all pertinent comparisons of the breeders' perfor- mance involving four thousand chick weights of Experiments I and II. In Experiments I and II the data were analyzed in- itially to obtain differences of progeny performance be- tween hens of a pen. Further, in Experiments I and II two consecutive hatches of chicks were produced by each of the sixteen different pen matings to obtain sufficient numbers of chicks for proper statistical analysis. Thirty-two analyses of variance were made of the two-week weights of the chicks from two consecutive hatches from each of the sixteen different pen matings. Each pen mating consisted of one male and twenty females. On an average, only eleven of the twenty females of a pen produced chicks in each hatch. The design of the sixteen different pen matings and their duplicate hatches may be seen in Figure 1. These initial thirty-two analyses of variance gave estimates of variance of progeny weights between hens of the same pen. The data from the analysis of variance of each in- dividual mating of one hatch were combined with the data of the analysis of variance from the same mating of the next consecutive hatch. A new analysis of variance was made to calculate the estimate of the average variance between the hens of a pen from the data of two hatches; 15 thus having gained an estimate of variance between con- secutive hatches involving the same mating. In order to obtain a final analysis of variance giving comparisons between the sires of both classifications (R and S), the statistics of the combined consecutive hatches had to be compared in a special manner. It is designated as Approach No. l and is illustrated in Figure 2. Approach No. 2 (Figure 3) is necessary to obtain comparisons between the pens of both classifications (R and S). Approach No. 5 (Figure 4) is necessary for comparisons between the unlike matings R x B vs. S x S and R x S vs. S x R. Figure 2 - Approach No. 1 15a s.v.* Bet. Like Pens i.e. 52 x 3,3L vs. 82 x sb (8 Comparisons in the Experiment) E2 I 3a]( NI 31 1 Sb S.V. Bet. Like Sires i.e. SI x Sab vs. 32 x Sab (4 Comparisons in the Experiment) 33 Sb Sa Sb rslxsal_ _ _LS. [_ _ _Tsz beJ l___|Sleb| lszxsE[_-.j S.V. Bet. Unlike Matings i.e. 312 x Sab vs. 812 x Red (2 Comparisons in the Experiment) 8a 3b R0 Rd l 31 X Sal 32 X Sb’| [—32 XSaI $1 beI V30 isg xRolslxfid—J I31XRCISZ XRdI S.V. Bet. Unlike Sires - Includes all cells of the Latin Square design in the Comparison Sa Sb Rc Rd Sa Sb Rc Rd SlxsaSZXSb______ _____ -RIXRCRZXRG ngSasleb______ ______ -szRcfllde -___-_ngRcSldevsoRngaRI-L Sb..——--- ____ ___slchsszd RleaRz Sb"“’""‘" u '6' S.V. - Source of variation Figure 3 - Approach No. 2 15b S.V.* Bet. Like Sires i.e. 81 x Sa vs. Sg x Sa VS. (8 Comparisons in the Experiment) '32 x Sbgl VS. 51 x Sb] S.V. Bet. Like Pens i.e. 312 1 Sa vs. 312 x Sb (4 Comparisons in the Experiment) a Sb VS. m m S S.V. Bet. Unlike Matings i.e. 812 x Sab vs. R12 x Sab (2 Comparisons in the Experiment) Sb Isl x Sal 32 x sbl 32 1: Sat 51 x SbJ V30 Sb R2 x Sal R1 x sbl [R1 x Sal 32 x Sb] S.V. Bet. Unlike Pens - Involves all the cells of the Latin Square Experiment Sa Sb Re Rd 81 x Sa 52 x Sb R1 x R0 R2 x Rd Sg x Sa 31 x Sb R2 x Rc R1 x.Rd R2 x sa R1 x Sb vs' 32 X RC 51 x Rd R1 x 83 R2 x SD. 81 x R0 82 x Rd *- S.V. - Source of variation Figure 4 - Approach No. 5 ’ 150 s.v.* Bet. Pens - (Same as Figure 2) S.V. Bet. Like Sires - (Same as Figure 2) S.V. Bet. Unlike Matings i.e. 812 x Sab vs. R12 x Rod (2 Comparisons in the Experiment) S Sb RC R a d Isl x safsz 3: 53 [RI 3: RC I122 x RE] I32 xSal SIbeJ VS' |R2 ch lRlxRQ S.V. Bet. matings of extreme classification and reciprocal matings 38 3b Rc Rd lesal SgXSb [121ch Iszch lszxsal slxsb lazxsc |R1de| VS. Rgxsalfilxsb 15:2ch Isldel EileaIszSb Islch IsszdI *S.V. - Source of variation 16 RESULTS AND DISCUSSION In the P1 population both the R individuals and the S individuals were found to be heterogeneous with respect to transmitting the hereditary factors for a uniform body weight to all their offspring. The degree of phenotypic expression within families ran from resistant (body weight in upper 20 per cent of weight range, less than 25 per cent mortality) through intermediate and even to susceptible (body weight in lower 20 per cent of weight range, more than 25 per cent mortality). However, classification of the Pl individuals was based on the average of their F1 progeny performances. The criteria of selection of the F1 breeders were pedigree estimates of the genotypes of the individuals. The decision to keep or reject a bird for breeding was based on the average of its sibs' two-week weight and mortality records. The estimated intensity of resistance for any one individual was in all probability not equal to the intensity of resistance of its sibs. Thus, individual merit of those selected was lowered every time a breeder, which would be rejected by progeny test results, was accepted because it had unusually excellent sibs. Therefore, it is expected that the F1 breeders would also demonstrate heterogeneity with respect to producing chicks that would have relatively equal weights at two 17 weeks of age when fed a vitamin B12 deficient diet. From the analyses of the progeny tests of Experiment II, the pedigree estimates of male R2 and male S2 were found to be erroneous. Fortunately, however, they were interchange- able, R2 became 32 and 52 became R2. The data from a study by Bird, §£_al., (1947) showed that the amounts of vitamin B12 stored in eggs at hatching time varied. These researchers suggested that the variation within a family in the ability of chicks to withstand a vitamin B12 deficiency was due to heredity. All other environmental factors were relatively equal for all chicks or controlled by statistical treatment. It is suggested that the variability of body weights within families is due either to a lack of uniform deposition by the hen of the vitamin in the egg, an inherited character- istic in the chick affecting the efficiency of vitamin B12 utilization, or a combination of both. Since the deposition of the vitamin in the egg is affected only by the female, the difference in the progeny performance of the R males and the S males can be ascribed solely to heredity if all other effects, including that of the females, can be sta- tistically eliminated. The analyses of variance of chick weights from Experi- ment I are presented in Tables 2, 3 and 4. Each of the three analyses of variance has a common estimate of error, estimate of variance between hens, estimate of variance 18 between hatches, and total variance. The approach of the analysis of variance in Table 2 is to facilitate the com- parison of unlike sires. The approach of the analysis of variance in Table 3 is to facilitate the comparison of unlike pens. In Table 4, the approach of the analysis of variance is to facilitate the remaining comparison of unlike matings. The above description of Tables 2, 3, and 4 of Experi- ment I also describes respectively Tables 5, 6, and 7 of Experiment II. Some elaboration on the terminology used in the analyses of variance is warranted. Variation between the performances by progeny from hens of a pen is referred to as variation between hens. Variation in the performance by progeny from a specific pen mating of one hatch and the performance by progeny of that same specific pen mating of the next consecu- tive hatch is referred to as variation between hatches. Variations in the performances by progeny of females of one pen and the performance by progeny of the females of another pen is referred to as variation between pens. The varia- tion in performance by progeny of males is referred to as variation between sires. The variation of performance by progeny of particular types of matings is referred to as variation between matings. 19 I. VARIANCES BETHEEN HENS WITH RESPECT TO THEIR PROGENY PERFORMANCES Eight of the thirty-two analyses of variance of Experi- ment I showed no significant difference in the variance between hens of a pen, and seven of the thirty-two analyses of variance of Experiment II data showed no significant difference in the variance between hens of a pen. These non-significant differences in the variance between per- formances of hens of a pen, even though these hens had been selected by pedigree estimates, show the relatively homo- zygosity among hens with respect to producing offspring of relatively equal average weights at two weeks of age on a vitamin 812 deficient diet. In the final analysis of variance of the data of Ex- periment I and of the data of Experiment II (Tables 2, 5, 4, 5, 6 and 7) the individual sum of squares of each analysis of variance were totaled and divided by the total degrees of freedom to give an estimate of average variance between hens. In Experiment I (Tables 2, 5 and 4) the estimate of average variance between hens was 1271.0, which was shown to be highly significant (F: