The Genetic Control of Variance in Comb and Testes Weights of Young Male Chickens R. GEORGE JAAP, MARION W. MURRAY* AND ROGER W.

354 E. D. WALTER AND J. R. AITKEN using 640 birds. These birds were forcemoulted between their first and second laying years. Feed restriction during the rearing period caused a change in the pattern of production, in that the birds commenced production later but subsequently laid at a higher rate than full-fed controls. Early egg size was considerably increased and total feed consumption was decreased. Rearing and laying house mortality and final body weight were unaffected. Feed restriction (12%) in the laying period effected a considerable saving in feed but this was offset by a drop in egg production. Production in the second laying year was not affected by feed restriction on range or in the first laying year except for the birds which were restricted in both periods. The latter group partially compensated for their low production in the first year by laying at a higher rate than the other groups in the second year. Egg weights were not affected by feed restriction in the laying house. REFERENCES Davis, W. M., and A. B. Watts, 1955. The effect of VARIATION in weight of tissues which respond to hormone stimulation may arise from: (1) variation prior to hormone stimulation, (2) variation in amount of hormone to which the tissue is exposed, and (3) variation in sensitivity of * Present address: Wye College, Wye, Kent, England. ration treatments during the various stages of the chicken's life upon the subsequent performance of laying hens. Poultry Sci. 34:1231-1232. Gowe, R. S., A. S. Johnson, R. D. Crawford, J. H. Downs, A. T. Hill, W. F. Mountain, J. R. Pelletier and J. H. Strain, 1960. Restricted versus fullfeeding during the growth period for egg production stock. Brit. Poultry Sci. 1: 37-56. Maclntyre, T. M., and J. R. Aitken, 1959. The performance of laying hens reared on restricted and full feeding programs. Can. J. Animal Sci. 39: 217-225. Milby, T. T., and D. H. Sherwood, 1953. The effect of restricted feeding on growth and subsequent production of pullets. Poultry Sci. 32: 916. Milby, T. T., and D. H. Sherwood, 1956. The influence of feed intake during the growing period on the subsequent performance of laying hens. Poultry Sci. 35: 863-869. Schneider, A. J., B. B. Bohren and V. L. Anderson, 1955. The effect of restricted feeding on several genetically controlled characters in the fowl. Poultry Sci. 34: 691-702. Singsen, E. P., L. D. Matterson, A. Kozeff and L. D. Stinson, 1954. The effects of feeding high and low efficiency rations to growing pullets on their subsequent laying performance. Poultry Sci. 33: 1081. Sunde, M. L., W. W. Cravens, H. R. Bird and J. G. Halpin, 1954. The effect of complete and incomplete growing diets on subsequent performance of the laying hen. Poultry Sci. 33: 779-784. The Genetic Control of Variance in Comb and Testes Weights of Young Male Chickens R. GEORGE JAAP, MARION W. MURRAY* AND ROGER W. TEMPLE Ohio Agricultural Experiment Station, Columbus, Ohio (Received for publication May 19, 1960) tissues of different individuals. Both initial variation and that due to hormone stimulation are affected by genetical and environmental differences between individuals. This report is concerned with early identification of genetic effects on the weight responses of the testes and the comb. These may, or may not, characterize the responses expected from each

VARIANCE OF COMB AND TESTES WEIGHTS 355 chicken throughout its life span. The small amounts of androgen or gonadotrophin which may be present in the blood stream prior to 11 days after hatching would appear to be below the threshold for stimulation of comb and testis size in large bodied breeds. Although Breneman (1941) obtained evidence of gonadotropic and androgenic effects on the testes and combs of White Leghorn males as early as the 10th day after hatching, these effects appeared later in Rhode Island Reds. Therefore, in the large bodied birds used for this report, variation in weight of these tissues at hatching and at 11 days is considered to be relatively free from hormonal effects. Both genetical and environmental effects on comb and testis size at or shortly after hatching have been indicated in the reports of Jaap (1935), Munro el al. (1943), Jaap and Thompson (1944), Dorfmann (1948), Jaap and Robertson (1953) and others. The question arises whether genetic effects in absence of hormone stimulation are similar to those affecting the hormone response. Use of the baby chick for hormone assay (c.f. Emmens, 1950) has provided much information on size and weight variation produced by varying quantities of gonadotrophins and androgens. The comb of the newly hatched chick has proved to be quantitatively sensitive to varying amounts of androgen. Likewise, the testes respond in a quantitative manner to known amounts of gonadotrophin (Breneman el al., 1959). A second question is concerned with the relative importance of genotype on the response to known amounts of hormone stimulation. Several reports have indicated some genetic control over hormone response. Dorfmann (1948) has demonstrated that the androgenic response of the White Leghorn chick's comb is greater than that of the Barred Plymouth Rock chick. Campos and Shaffner (1952) estimated the heritability of weight increases of the androgen-stimulated comb at 14 days after hatching to be 0.13 when the chicks were simultaneously injected with androgen and estrogen. Jaap and Robertson (1953) demonstrated that size of the comb prior to androgen stimulation affected the magnitude of the weight increase of the comb. The differences in response of six out of seven inbred lines of Brown Leghorns was attributable to differences in comb size at the beginning of the androgenic stimulation. The breed differences in response demonstrated by Dorfmann (1948) could have been due to initial comb size because the comb of White Leghorn chicks is larger than that of Barred Plymouth Rocks at hatching. Campos and Shaffner (1952) used one breed, New Hampshires, to demonstrate the heritable effects on the weight of hormone-stimulated combs. Chapman (1946) and Kyle and Chapman (1952) have developed lines of rats with high and low ovarian response to gonadotrophin. Casida el al. (1952) have attributed the genetic differences in response of the ovary in these two lines of rats to a larger initial ovarian size in the high response lines. While the testis of the male chick has been used frequently in assay of gonadotrophins, no study has been made of the genetics of this response. It appears reasonable to assume that genetic effects may act directly on cell growth and multiplication as well as on the secretory activity of the interstitial cells of the testes. USE OF RANDOMBRED SYNTHETIC STRAIN A large-bodied meat-type strain has been developed from crossbred ancestry. This population, designated as White Gold, has been maintained by a restricted

356 R. G. JAAP, M. W. MURRAY AND R. W. TEMPLE type of random mating. The breed composition and mating methods used to maintain the White Gold Randombred were described briefly by Goodman and Jaap (1960). Gold colored sires and silver colored dams, which were, in turn, progency of silver colored sires and gold colored dams, were used to obtain male chicks which could be readily identified by their white (silver) down color at hatching. The White Gold population was closed in 1955. Most of the data included in this report were collected during the 1956 through 1958 period. Since this strain was utilized immediately after it was closed, genetic effects in this population may differ from those of a breed population which has been closed for a number of years. Unpublished evidence from body weights at eight weeks of age and from the frequencies of pea versus single combs indicate that the population remained relatively stable during the 1956-1959 period. EXPERIMENTAL METHODS Both the sires and dams were maintained in individual wire cages. Sets of two gold sires and two silver dams were artificially mated in all four combinations. Chicks were hatched every three weeks, allowing time to change sire-dam combinations between hatches. The 2X2 sets of progenies have been designated as a diallel set. Two full brothers were used from each mating. A total of eight male chicks from the four matings of two sires and two dams formed each diallel set. As soon as all four progenies were obtained new sires and dams were used to obtain additional sets of progenies from diallel matings. The male chicks were killed by cervical fracture at one, or at eleven, days after hatching. Hormone treatment for measuring response was administered during the ten-day growing period. The combs and testes were removed with small curved scissors, cleaned of extraneous tissue on a moist paper towel and weighed immediately on a Roller-Smith torsion balance. Comb and testes weights were correlated with body weight. The regression of gland on body was statistically significant in all tests. Comb and testes weights were adjusted individually to a constant body weight by means of regression coefficients calculated for each segment of the data. Variance among the progeny within each diallel set was subdivided into that contributed by the sires, the dams, and sire-dam interaction. The remaining variance (Q) was attributable to differences between full brothers. These variances from the different diallel sets were pooled and the genetic components attributable to sires (S), dams (D) and sire-dam interaction (/) estimated from the mean squares as outlined by Lerner (1950). The additive genetic effects, h 2 a, were obtained by doubling S+D and dividing by S-\-D+I-\-Q. The non-additive genetic effects, hi 2, represents four times the I- fraction of the total of S+D+I+Q. After the data for this report were collected, Goodman and Jaap (1960) estimated that four full sibs per family contributing to 50 diallel sets appeared to be about the minimum size of sample for reasonable estimation of these additive and non-additive heritabilities. Only two full brothers per family were used in the following tests and, in most cases, the number of diallel sets was below 50. Therefore, h 2 and h? are crude approximations of the true genetic effects in this randombred population. Many diallel sets were incomplete and excluded from the diallel analyses. All families of two full brothers were used for estimation of genetic effects due

VARIANCE OF COMB AND TESTES WEIGHTS 357 to differences between full sib families. A superior estimate of the total genetic effects, h 2 b, was calculated from the larger number of full sib families usable in this method. Greater confidence should be placed in estimates of hb 2 than in those of hj+h? from the diallel sets due to the much larger number of full-brother families and to lower errors of estimation expected when the variance is divided into two rather than four subdivisions. Should 2 dominance and epistasis be important h b estimates would be larger than those estimated by h 2 due to inclusion of i of the dominance, f of the additive X additive, \ of the additive X dominance, and \ of the dominance X dominance genetic effects which are excluded from h 2. When dominance or epistasis was present h 2 a plus h? would be expected to estimate the total heritabilities more closely than hb 2 ; however, greater errors of estimation arose in the h 2 a and h? due to the smaller sample sizes. RESULTS Day-old Cockerels.' Body comb and testes weights were available for 543 dayold cockerel chicks. These weights, coefficients of variation, regressions and phenotypic correlations are summarized in Table 1. Both comb and testes weights are more variable than the body weight. Coefficients of variation (Table 1) were 12.5 for body, 71.1 for comb and 92.2 for testes. Heritability of Day-old Comb and Testes TABLE 1. The relation between body, comb and testes weights for 543 day-old White Gold cockerels Coeffi- Regres- Corre-,.,, cient of sion lation Wel S ht varia- on with tion body body Body 39.1 gm. 12.5 Comb 11.37 mg. 71.1 0.184 0.354 Testes 7.43 mg. 92.2 0.160 0.365 Weight. Before making the genetic analyses, the comb and testes weights of each chick were adjusted to a constant body weight of 39 grams by means of their respective regressions as given in Table 1. A considerable portion of the variance in body weight is attributable to the size of egg from which the chick hatches. Some of these egg size differences were due to age of dams. Adjusting comb and testes weights to a constant body weight should remove this non-genetic effect. Thus, heritability was measured for weight variations of the comb and testes at a constant body weight of 39 grams. The population was segregating for single and pea comb types. The frequency of pea combs was about 14 percent. Only the single comb cockerels were used for heritability estimates for comb weight. The summarized data are presented in Table 2. There were 221 families of two full brothers available for calculation of hb 2 for comb weight and 253 families for testes weight. The number of diallel sets available to subdivide additive, h 2, from non-additive, h?, heritability were 14 for comb and 21 for testes. Such a small number of diallel sets may be used merely as an indication of the types of genetic effects contributing to hb 2. Heritability, hb 2, for both comb and testes was high at hatching (.76 and.81, respectively). The diallel data for comb weight suggest that the baby chick's comb may be influenced by both additive and non-additive genetic effects. In contrast, TABLE 2. Heritability estimates for single comb and testes weights at one day of age and at a constant body weight of 39 grams No. No. of 2 of hf diallel h a families sets Comb 221.76 14.34.63 Testes 253.81 21.58 -.38 h,'

358 R. G. JAAP, M. W. MURRAY AND R. W. TEMPLE the negative value of hi 2 obtained for testes weight at one day of age would indicate that all of the genetic effects on testes weight might have been additive. Correlations between Comb and Testes. Two hundred and twenty families, each having two single comb brothers, have been used to calculate the genetic and environmental correlations at hatching between comb and testes weight at a constant body weight of 39 grams. These genetic and environmental correlations, calculated from the differences between and within full-brother families, were as follows: Phenotypic Genetic Environmental.068 -.031.398 It is evident from the low phenotypic correlation between comb and testes weight at a constant body weight that removal of the effects of body weight removes most of the covariance between the comb and testes. This observation supports the original assumption that gonadotropic and androgenic hormone stimulation were inconsequential at hatching. The phenotypic correlation was so low that the negative genetic correlation might have been expected to be larger and nearer that of the positive environmental correlation. The environmental correlation of.398 indicates a tendency for environmental effects which favor larger testes weights to increase comb weight, and vice versa. TABLE 3. The relation between body, comb and testes weights in 11-day-old White Gold chickens Body Comb Testes No. of males 751 713 751 Weight 95.9 gm. 23.8 mg. 26.6 mg. Cockerels at 11 Days of Age. Data from more than 700 cockerel chicks were available to compare the relation between body comb and testes at 11 days after hatch- Coefficient of variation 17.7 31.5 33.8 Regres- Correlasion on tion with body body.312.312.690.576 ing. These are summarized in Table 3. Growth was very rapid, the gain in weight being approximately 57 grams in 11 days. Body and comb weights more than doubled in 11 days and testes weights were more than three times the weight of those at hatching (Table 1). Coefficients of variation (Table 3) were 17.7 for body, 31.5 for comb and 33.8 for the testes weights. Using these coefficients to compare variability at day-old and at 11 days of age, body weight was slightly more variable at 11 days while the comb and testes were much less variable than at hatching. Regressions for both comb and testes on body were greater (P<.01) than those obtained for newly hatched chicks (Table 1). Heritability of 11-day Comb and Testes Weights. A total of 237 families composed of two full brothers per family were used to estimate h? for comb weight and 265 families for calculation of hb 2 for testes weight (Table 4). These heritabilities bases on the covariance between full brothers continued to be high,.68 and.87, at 11 days after hatching. An excessively large estimate was obtained for the non-additive genetic effects on comb weight at 11 days of age (Table 4). Although the number of diallel sets was small (23), they may be interpreted to indicate that the non-additive genetic variance for comb weight by the 11th day of age had increased in relation to the additive genetic effects which may, or may not, have been different from zero. TABLE 4. Heritability estimates for single comb and testes weights at 11 days of age and a constant body weight of 96 grams Comb Testes No. of h* families 237 265.68.87 No. of diallel sets 23 59 -.04.32 1.97.43

VARIANCE OF COMB AND TESTES WEIGHTS 359 The larger number of diallel sets (59) used to calculate ha and h? for testes weights may indicate that the high heritability, hf, resulted from both additive and non-additive genetic effects on growth of the testes after hatching. However, the controls used for gonadotrophin tests (Table 7) did not confirm this observation. The chicks used for the calculations given in Table 4 were from earlier generations (1955 and 1956), while those used for the tests with gonadotrophin were produced between the latter part of 1956 and the end of 1958. Some of the nonadditive genetic effects in the population immediately after it was closed from crossbred ancestry could have been lost during the first few generations of random breeding. DIRECT ANDROGENIC STIMULATION OF THE COMB Procedure. A sesame oil solution of testosterone propionate, containing 0.5 mg. per ml. of solution was used for inunction of the comb. Beginning on the 4th day after hatching and continuing through the 9th day, each chick's comb was covered with the androgenic oil solution on each side. Only enough oil to cover both sides of the comb was used at each application. Jaap and Robertson (1953) had demonstrated that the comb response is proportional to the androgenic concentration of the solution rather than being proportional to the amount of hormone spread upon the comb. Data were collected on the 11th day after hatching in order to compare the results with those from non-treated control chicks used for Tables 3 and 4. Table 5 gives the data summaries for the chicks receiving testosterone propionate solution applied directly to the surface of their combs. Body weight of these chicks was slightly larger but varia- TABLE S. The relation between body and comb weights in 368 White Gold male chicks which received testosterone stimulation applied directly on the comb Body Comb Weight 101.8 48.61 Coefficient of variation 18.6 44.3 Regression on body.630 Correlation with body.567 tion was similar to that of the non-treated chicks (Table 3). The 6-day application of testosterone propionate solution on the comb more than doubled comb weight, 48.6 mg. versus 23.8 mg. Variability of comb weight increased as a result of the androgenic growth stimulation. Also, comparing Tables 3 and 5, the regression of comb weight in mg. on body weight in gm. for the treated chicks was about double that of the non-treated males at 11 days of age,.637 versus.312. This difference in the regressions was highly significant (P<.01). The correlation between comb and body weights after testosterone inunction of the comb was slightly lower,.56 versus.69. Heritability of the Comb Weight under Testosterone Stimulation. The data in Table 5 demonstrated that growth rate of the chick affected its response to testosterone stimulation of the comb. Each comb weight was, therefore, adjusted to a constant body weight of 100 gm. by means of the regression given in Table 5. Heritability estimated from the covariance between full-brother families, hf, was.56 based upon 184 families. This estimate of the genetic effects was lower than the heritability of.78 obtained for the non-treated males. It is probable that most of the increase in variability of comb weight induced by topical testosterone stimulation was environmental, lowering the fraction of the total variation attributable to genetic effects. There were 32 diallel sets available for

360 R. G. JAAP, M. W. MURRAY AND R. W. TEMPLE partitioning the variance in comb weight from androgenic stimulation. The additive heritability, h a 2, was estimated to be.46 and the non-additive, h.?, to be.13. It is interesting that the total genetic effects,.59, estimated from the 32 diallel sets of testosterone-treated chicks closely approximated the ht? estimate of.58. It is possible that more additive effects are expressed as a result of the hormone stimulation. The non-additive heritability was estimated to be much higher in nontreated chicks (Table 4). RESPONSE TO GONADOTROPHIN Procedure. Pregnancy mare serum preparation (PMS) was chosen as the most stable gonadotrophin for use in estimating heritabilities of gonadotrophic stimulation on the chick's testes. Each chick was injected subcutaneously with 2 Cartland- Nelson units (Cartland and Nelson, 1938) of PMS (Gonadogen-Upjohn) per day from the 5th through the 9th day after hatching. This dosage was chosen to produce approximately 100 percent increase in testes weight when the chicks were sacrificed at 11 days of age. Full brother non-treated control chicks were grown intermingled with the injected chicks. Most of the data for non-treated mdle chicks described previously were collected during 1955 and 1956. These data from chicks injected with gonadotrophin were obtained during the period 1956 through 1958. It was thought desirable to use contemporary controls as insurance against possible genetic changes which might have occurred in this randombred White Gold population during the first few generations after it was closed. Average Weights and Variation. Body, comb and testes weights in chicks injected with a total of 10 units of PMS are compared with those for their non-treated controls in Table 6. The PMS injected chicks grew at about the same rate as their noninjected brothers. The total subcutaneous injection of 10 units of PMS increased testes weight 106 percent. As a result of this gonadotrophic stimulation the testes secreted sufficient androgen to increase comb weight by 82 percent. While PMS injection did not produce much change in variability of the body or testes weights, it did result in a large increase in variability of comb weights. This secondary hormone response of the comb resulting from androgen secretion of the PMS stimulated testes produced a coefficient of variation of 49.7 versus 29.5 for the controls (Table 6). Since direct absorption of testosterone from the surface of the comb increased comb variability (Table 5), it is possible that most of this increase in variability as a result of PMS stimulation of the testes was due to varying sensitivities of the comb rather than to varying adrogenic secretion of the PMS-stimulated testes. Whether comb response to androgen circulating in the blood stream is similar to that from androgen absorbed through the comb surface has not been tested. Therefore, it is impossible to determine from these tests whether the testes of different chicks responded by production of different TABLE 6. The relation between body, testes and comb weights in chicks which received 10 units of PMS and their non-treated controls at 11 days of age Treated Control Treated Control Treated Control No. of chicks 663 560 663 560 623 497 Coeffi- Regression ^""wirh Weight cientof on body btx variation weight we; ght Body 94.8 19.1 95.0 17.6 Testes 53.6 32.0 26.0 33.8 Comb 34.9 49.7 19.2 29.5.120**.328.478**.292.177.652.549.691 ** Regression for treated chicks highly significantly different (P<.01) from that of the controls.

VARIANCE OF COMB AND TESTES WEIGHTS 361 amounts of testosterone from the same amount of PMS stimulation. Both the regression of testes on body and the correlation between testes and body (Table 6) in PMS injected chicks were much less than those for the noninjected brothers. Since the correlation between testes and body weight in PMS stimulated chicks was so low, 0.177, it would appear that the testes response to gonadotrophin at this age may be almost independent of body size of the chick. The correlation between comb and body weights in treated chicks (.549) was slightly smaller than that of the noninjected controls (.691). This confirms the previous conclusion that the comb re^sponse to androgen is affected by body size. The increased regression of comb on body demonstrates that each change in body weight allows a greater change in comb weight when the chick is subjected to androgen stimulation. Heritabilities. Heritabilities, h/, 2, of testes and comb weight (Table 7) were high in both the PMS injected and control chicks at 11 days of age. These estimates for PMS injected chicks were slightly higher than those for non-treated contemporary controls and those from earlier generations (Table 4). The genetic effects in the contemporary controls observed during 1956 through 1958 appeared to be entirely additive on testes weight and partially non-additive on comb weight. These estimates do not agree with the earlier estimates collected during 1955 and 1956. It is, therefore, not possible to make any conclusion regarding the relative importance of additive and non-additive genetic effects on testes and comb size in non-treated chicks at 11 days after hatching. More confidence may be placed in the k a 2 and h b 2 estimates for testes weights in the treated chicks. The larger number of TABLE 7. Herilability estimates for testes and single comb weights at 11 days of age in chicks which received 10 units of PMS gonadotropin as compared with non-injected controls No. No. of of fa* diallel h<? fa* families sets Control 281 0.715 36 0.962-0.344 Treated 319 0.925 56 0.796 0.288 Comb Control 244 0.886 26 0.764-0.224 Treated 286 0.947 42 1.094-0.292 * Some of the non-additive effects may have become dissipated with closed flock breeding during first year or two after it was closed from cross ancestry. diallels and the close agreement between ht, 2, and h 2 plus one-half of h?, would indicate that there are some non-additive genetic effects on the testes weight response to gonadotropin. DISCUSSION High heritability of comb and testes weights has been demonstrated in chicks both with and without hormone stimulation. The question of whether genetic effects in the absence of hormone stimulation are similar to those affecting the hormone response cannot be answered from these experiments. From the previously reported results obtained by Goodman and Jaap (1960) it is doubtful whether the diallel mating method is expedient because confidence intervals for the genetic components with so many subdivisions of the variance appear to be excessively large. Only those derived from more than 150 diallel sets would appear to be reliable estimates. Since weights of the comb and testes in these very young chicks are highly heritable, selection experiments might provide more decisive information. In view of the high heritabilities, it should be possible to produce distinct changes in weight of the comb or testes by selection

362 R. G. JAAP, M. W. MURRAY AND R. W. TEMPLE for a relatively few generations. Using randombred populations as a base for these selection experiments, it will be possible not only to measure the effective heritability but also to test for responses correlated with genetic changes in testes or comb weights at these young ages. Should the genetic control over testes and comb response to hormone stimulation remain the same throughout the life of the bird, tissue responses with and without hormone stimulation may be expected to be correlated with later performance in such economical traits as egg production and growth. Identification of these genetic differences at, or shortly after, hatching should make it possible to accelerate progress in breeding for such economical characters. Selection experiments for increased response of the comb to androgen resulting from testosterone and PMS gonadotropin injection are planned in order to explore this method of estimating the various types of genetic effects. Also, it is possible that diallel crosses between populations might yield superior information on the non-additive effects than do diallel matings within populations. SUMMARY Variations in body, comb and testes weights together with regressions and correlations are presented for male chicks at hatching and at 11 days of age. Using a large number of families composed of two full brothers within a randombred population, heritability was estimated from the variance between sets of full brothers. When body weight was held constant, these heritability estimates for weights of comb and testes were: (1) At one day of age: comb.76; testes.81 (2) At 11 days of age: comb no hormone stimulation:.68 and.89 comb testosterone innunction:.56 comb gonadotropin stimulation:.95 testes no hormone stimulation:.87 and.72 testes gonadotrophin stimulation:.92 Subdivisions of these high heritabilities into their additive and non-additive genetic effects were calculated from analyses of diallel matings within this randombred population. Considering those cases where the number of diallel sets was more than SO (more than 200 families) as reliable estimates, non-additive genetic effects were demonstrated to be important for testes weight both with and without PMS stimulation. REFERENCES Breneman, W. R., 1941. Growth of the endocrine glands and viscera in the chick. Endocrinology, 28: 946-54. Breneman, W. R., F. J. Zeller and B. E. Beekman, 1959. Gonadotrophin assay in chicks. Poultry Sci. 38: 152-158. Campos, A. C, and C. S. Shaffner, 1952. The genetic control of chick comb and oviduct response to androgen and estrogen. Poultry Sci. 31: 567-571. Cartland, G. F., and J. W. Nelson, 1952. The bioassay of mare serum hormone. Amer. J. Physiol. 122: 201-207. Casida, L. E., B. R. Casida and A. B. Chapman, 1952. Some differences between two strains of rats developed by selection to differ in their response to equine gonadotrophin. Endocrinology, 51: 148-151. Chapman, A. B., 1946. Genetic and nongenetic sources of variation in the weight response of the immature rat ovary to a gonadotrophic hormone. Genetics, 31:494-507. Dorfmann, R. I., 1948. Studies on the bioassay of hormones. The relative reactivity of the comb of various breeds of chicks to androgens. Endocrinology, 42: 7-14. Emmens, C. W., 1950. Hormone Assay. Academic Press, New York. Goodman, B. L., and R. C. Jaap, 1960. Improving accuracy of heritability estimates from diallel and triallel matings in poultry. 1. Eight-week body weight in closed flocks. 2. Weight of spleen and bursa of Fabricius in a randombred population. Poultry Sci. 39: 938-949. Jaap, R. G., 1935. Gonad-stimulating potency of individual pituitaries. Poultry Sci. 14: 237-246. Jaap, R. G., and R. B. Thompson, 1944. Natal testes weight and early post-natal growth in the

VARIANCE OF COMB AND TESTES WEIGHTS 363 domestic fowl. Growth, 8: 311-319. Jaap, R. G., and H. Robertson, 1953. The chick comb response to androgen in inbred Brown Leghorns. Endocrinology, 53: 512-519. Kyle, W. H., and A. B. Chapman, 1952. Experimental check of the effectiveness of selection for a quantitative character. Genetics, 38: 421-443. Lerner, I. M., 1950. Population Genetics and Animal Improvement. Cambridge University Press, London. Munro, S. S., I. L. Kosin and E. L. Macartney, 1943. Quantitative genic-hormone interactions in the fowl. 1. Relative sensitivity of five breeds to an anterior pituitary extract possessing both thyrotropic and gonadotropic properties. Amer. Nat. 77: 256-273. The Effect of Feeding Antibiotics to Chicks in the Presence of Fresh, Dried and Autoclaved Hen Feces* W. K. WARDEN AND PHILIP J. SCHAIBLE Department of Poultry Science, Michigan State University, East Lansing ACCORDING to Coates et al. (1952) an infectious agent must be present in order for antibiotics to stimulate growth in chicks, since chicks reared in new, clean environment did not respond to dietary antibiotics. This concept is supported by the work of Anderson et al. (1956) who showed that feeding chlortetracycline overcame the depression in growth rate of chicks fed enterococci. More recently, Mameesh el al. (1959) obtained a consistent growth response in chicks fed terramycin when the diet was contaminated with hen feces; a response under these conditions was apparent in only one of six trials with penicillin. Barnes et al. (1959) found that rats fed penicillin exhibited an increased growth rate only if they had access to their feces. In order to investigate further the possibility of a relationship of feces ingestion to antibiotic response and to ascertain if this contamination is associated in any way with the mechanism of antibiotic Michigan Agricultural Experiment Station Journal Article No. 2628. * This research was supported in part by a grantin-aid from Commercial Solvents Corporation, Terre Haute, Indiana. (Received for publication May 19, 1960) action, the following experiments were conducted. EXPERIMENTAL In each experiment 200 grams per ton of antibiotic was fed throughout the test period. Allotment of chicks to replicate pens was made as follows: one-day-old White Rock broiler cockerels (Cobb's strain) were distributed on an equal weight-range basis in quadruplicate lots of 9 or 10 birds each and reared in electrically-heated brooders with raised floors. Except for the controls, replicates of each treatment were assigned to different deck levels and different batteries to minimize possible environmental and positional differences. In order to avoid antibiotic contamination from feed sifting from higher decks, non-antibiotic control lots were assigned to the top decks of the batteries. The composition of the basal all-mash ration fed in these studies is shown in Table 1. Significance of growth effects were measured using an analysis of variance and Duncan's multiple range and multiple F tests (1955). In each experiment some lots of chicks were given access to fresh fecal material