OF HEREFORD AND ANGUS CATTLE'

Transcription

1 HAIR COAT CHARACTERISTICS AND POSTWEANING GROWTH OF HEREFORD AND ANGUS CATTLE' R. P. Gilbert2 and D.R.C. Bailey3 Agriculture Canada Research Station, Lethbridge, Alberta T1J 4B1 ABSTRACT Mid-rib hair coat samples (n = 577) were obtained from 9- to 10-mo-old Hereford and Angus bulls and heifers to examine diet and breed differences in hair coat characteristics and their relationship to 168-d postweaning gain. Each sample was cleaned and subdivided into guard hairs and undercoat, Dependent variables included the number, weight, length, diameter and percentage of medullation of guard hairs and undercoat. Sources of variation included breed, sire nested within breed, year (1965 or 1966), sex of calf, diet fed during the postweaning gain test (2 parts rolled grain:l part chopped hay vs all chopped hay) and the linear regression on age of calf as a covariate. Factor analysis was used to transform the 10 hair coat variables into a set of four factors that accounted for 71% of the total variance of the original variables. Angus cattle tended to have shorter, less medullated coats (Factor l), shorter, larger diameter undercoat hairs and guard hairs with less medullation than Herefords. Sire within breed differences existed for weight, length, and diameter of both types of hairs and all four factors. Compared with the medium-energy-diet, the high-energy diet reduced hair weight per unit of surface area, undercoat number and guard hair medullation. Undercoat density (Factor 3) was reduced by the higher-energy diet (P <.001), whereas guard hair density (Factor 2) was not changed. The number of both types of hairs were lowly heritable (h2 I.07), but all measms of hair coat weight were moderately heritable (24 I h2 5.30). None of the hair coat characteristics or factors were strongly associated with 168d postweaning gain. Key Words: Beef Cattle; Factor Analysis, Hair, Undercoat, Growth Rate, Heritability, Genetic Correlation lntroductlon Smoliak and Peters (1955) observed that bison-cattle hybrids were more willing to graze on open range under adverse winter conditions in western Canada than were Hereford, Angus or Shorthorn cows. In subsequent research, Peters and Slen (1964) examined hair coat differences among cattle, bison and their hybrids and found that bison had much heavier, denser hair coats than did domestic cattle. The bison-cattle hybrids were intermedi- 'Contribution no Resent address: Agricultural Research & Development, Saskatchewan Wheat Pool, Saskatoon SK S7K 3P7. 3~0rrespondi au&or. Received January 26, Accepted June 20, J. Anim. Sci &506 ate to their parental species in hair coat characteristics. With these results in mind, hair coat samples were obtained from unselected Hereford and Angus calves during February of their 168d postweaning gain test to determine whether hair coat differences were related to winter feedlot performance of cattle in western Canada. These calves made up the first two calf crops in a 19-yr selection study for postweaning gain on one of two diets begun in 1%3 (Bailey et al., unpublished data). Although considerable time has elapsed since the collection of these data, current emphasis by animal breeders on matching cattle to the environments in which they are expected to live extends the application of these results to the present. Multivariate methods are appmpriate for the analysis of simultaneous measurements of several traits or characteristics of each animal.

2 Dimension reduction techniques, such as principal component and factor analysis, have been used to evaluate underlying relationships among several traits measured simultaneously in beef cattle, for example, linear measure ments and performance (Brown et al., 1973) and age, weight, gain and pregnancy rate (Maltarechian et al., 1985). The objectives of this study were to evaluate differences in hair coat characteristics and their relationship to postweaning growth of cattle during the winter months in westem Canada and to determine whether consolidation of the variation in the hair coat characteristics by factor analysis aided interpretation of these relationships. Materlals and Methods Experimental Animals. The cattle used in this study were unselected purebred Hereford and Angus bull and heifer calves born in 1964 and Twelve sires were used in each breed each year. All Hereford sires and 11 of 12 Angus sires were used both years. One Angus sire used in 1964 was replaced in Following weaning on November 1 in both years, the calves were assigned to one of eight pens by breed (Angus, Hereford), sex (bull, heifer) and diet. The diets consisted of either a high-energy diet (HED), composed of 44.4% rolled barley, 22.2% rolled oats, 16.7% chopped alfalfa hay and 16.7% chopped Brome hay or a medium-energy diet (MED), composed of 50% chopped alfalfa hay and 50% chopped Brome hay. Each pen was 38 by 23 m surrounded by a 2.1-m solid board fence and contained a heated waterer and selffeeders. Salt and vitamixi mineral supplement were available ad libitum Hair Samples. Mid-rib hair samples were obtained from each calf on feed on February 17 and 18, 1%5 and February 8, 1966 (average age: 310 and 283 d; range: 57 and 55 d, respectively). An electric clipper with the center teeth removed was used to isolate a 5.1-cm2 area of hair on the left mid-rib of each animal by cutting two pexpendicular swaths in the coat. Each sample then was removed by a second electric clipper with all teeth intact. Raw samples were weighed and all detectable foreign material was removed. Each sample was placed in the center of a folded, 15-cm-square, #lo mesh metal screen and washed in a sequence of wash/rinse solutions typical of wool scouring procedures (A. H. Woolliscroft, Lethbridge Research Station, HAIR COAT CHARACTERISTICS AND GROWTH 499 personal communication). Samples were reweighed ( clean weight), and a subsample of approximately 2096 of the cleaned hair was removed, weighed ( used weight) and subdivided by visual evaluation into guard hairs and undercoat. The number and weight of guard and undercoat hairs in the original sample were estimated by multiplying the number and weight of the hairs of each category in the subsample by the ratio of clean weight to used weight. Length and diameter of undercoat and guard hair for each animal were estimated by averaging the measurements from 100 hairs of each type chosen at random from the subsample of cleaned hairs. Percentage of medullation, or the proportion of guard or undercoat hairs with hollow or fluid-filled centers, was estimated by the relative frequency of medullated hairs in the 1OO-hair subsamples when viewed under magnification. An arcsin-square root transformation was applied to the medullation values to stabilize the variances of these proportions (Snedwr and Cochran, 1967). Statistical Analyses. Least squares ANOVA procedures for unbalanced data (Harvey, 1985) were used to analyze hair coat variables and 168d gain. Because calves were allocated to pens by breed, sex and diet, and only one pen was provided for each of the smallest subclasses (e.g., 1965 Angus bulls fed the HED), pen effects were confounded with interactions among the main effects. Hence, a main effects model was used in the analyses. The linear model included year, breed, sire nested within breed, sex, diet and the linear regression of the dependent variables on age of calf at time of sampling as a covariate. A total of 580 animals were sampled, but undercoat medullation was not recorded for three Hereford bulls on the MED in 1965; those three records were deleted from all subsequent analyses. To further examine the underlying relationships among these variables, factor ROC FACLYIR; SAS, 1985) were performed on the hair coat variables using an oblique rotation model. Factor analysis refers to the multivariate procedures used to describe, if possible, the covariance relationships among many variables in terms of a few underlying, but unobservable, random quantities called factors (Johnson and Wichem, 1982, p 401). With this method, initial factors are obtained from the standardized variables (i.e., from the correlation matrix) by principal component procedures. Because the variables are standardized, the contribution of each variable to the

3 500 GELBERT AND BAILEY TABLE 1. CORRELATIONS PACTOR LOADINGS) BETWEEN FACTORS (FACTOR 1 TO FACTOR 4) AND HAIR COAT VAIUABLES, AND THE PROPORTION OF TOTAL VARJATION IN EACH HAIR COAT VARIABLE EXPLAINED BY THE FOUR FACTORS Hair coat variable -length Undercoat length Guard medullation Undercoat medullation Guard weight Guard number Undercoat weight Undercoat number Guard diameter Undercoat diameter Correlation with factor ' P P '.oJ ' ".Ol ' I2.88' -.OS -.I ' -22.a * %actor loading (correlation with this hair coat variable) was used in factor interpretation. Roportion of variance explained underlying factors is not influenced by its scale of measurement. Because the original factor loadings, the correlations between the underlying factors and the original variables, may not be readily interpretable, a transformation matrix often is used to rotate the loadiigs such that each variable has a high loading on a single factor and small to moderate loadings on the other factors (Johnson and Wichern, 1982). Rotation using an orthogonal transformation matrix maintains the independence of the factors, whereas an oblique (or nonorthogonal) rotation does not assume or require independence. The number of factors retained for consideration was determined by the number of eigenvalues of the correlation matrix with values greater than 1 (Johnson and Wichern, 1982). The proportion of the total variation explained by the n factors chosen is eqd to the sum of the first n eigenvalues divided by the number of variables in the analysis, 10 in this study. It is clearly desirable to have a few factors explain a large proportion of the total variation. Because the s uare of a factor loading yields a partial R 4, all loadings less than.45 (i.e., those variables with partial R2 c.20) were not considered in the interpretation of a factor (Table 1). Standardized factor scores, with zero mean and unit variance, were calculated for each animal and analyzed with the same linear model as the original hair coat variables to determine whether factors could be used in place of the original variables. Clean weight was not included in the factor analyses because it was the sum of guard hair and undercoat weights. Preliminary analyses included raw weight to determine whether any substances removed during cleaning contri- buted to the variation in hair coat characteristics. Only the relative order of the factors changed; therefore, raw weight also was excluded from the final factor analyses. Paternal half-sib heritabilities, genetic and phenotypic correlations, and their respective standard errors were estimated for 168-d gain, the hair coat variables and factors using the methods described by Harvey (1985). Results and Discussion Considerable research has examined the influence of the hair coat on heat tolerance of cattle (Findlay, 1958; Dowling, 1959a,b, 1%0; Schleger and Turner, 1960; Turner and Schleger, 1%0), but little work has focused on hair coat effects on cold tolerance (Yeates and Southwtt, 1958). Research in western Canada has examined the effects of cold tolerance on performance (Webster, 1970; Webster et al., 1970; Milligan and Christison, 1974), but little information exists regarding the effects of hair coat characteristics on winter performance of cattle in cold climates. Dowling (1959a) indicated that the primary differences between winter and summer hair coats of cattle were that the insulating coat found in winter had longer but fewer medullated fibers relative to the noninsulating summer coat. Medullated hairs are shorter and stiffer than nonmedullated hairs and may enhance air movement and heat dissipation (Dowling, 1959a). Webster et al. (1970) found that the rate of hair growth was the same for heifers raised either indoors in a 20'C room or outdoors at -28'C with or without shelter. However, the heifers raised outdoors had twice the total hair cover (mg hair/cm2 of skin) of the heifers raised indoors because of reduced shedding.

4 HAIR COAT CHARACTERISTICS AND GROWTH Hair Coat Variables. Analysis of variance results for clean weight did not differ from those of guard hair and undercoat weights, and least squares means (LSM) for clean weight were the sums of the respective LSM for guard and undercoat weights; therefore, results for clean weight are not discussed. Least squares ANOVA results for the remaining 11 hair coat variables and 168-d gain are presented in Table 2, and LSM and SE in Table 3. The mean lengths of guard (55.9 mm) and undercoat hairs (33.3 mm) in the present study were greater than the mean length of all hairs (25 mm) from Hereford, Shorthorn, and Africander-cross cattle in late fall in Australia (Turner and Schleger, 1970). However, their overall mean hair diameter of 33 and 40 pm at 11 and 24 mo was similar to the mean undercoat hair diameter of 32.8 pm but less than the mean guard hair diameter of 75.3 p in our study. Factor Analyses. Four underlying factors explained 71% of the variation in guard hair and undercoat weight, length, diameter, number and percentage of medullation. The eigenvalues and cumulative proportion of the variance explained by Factors 1 through 4 prior to rotation were 2.553,26%; 1.783,43%; %; and 1.194, 71%. respectively. Examination of the factor loadings in Table 1 provided the following interpretation to the factors: Factor 1: long, medullated vs short, less medullated coat; Factor 2: dense vs sparse guard hair coat; Factor 3: dense vs sparse undercoat and Factor 4: thick vs thin hairs in both coats (i.e., coarse vs fie coat). The procedures generally were effective in creating a pattern of loadings such that the four factors explained a large proportion of the variation of each hair coat variable, and each variable had a large loading on only one factor (Table 1). Medullation of guard and undercoat hairs were the exceptions, possibly due to their scale of measurement or transformation. These four factors explained less than 40% of the variation in guard hair medullation, and loadings were moderate for two of the factors (Table 1). Essentially the same factor interpretations would have resulted from an orthogonal rather than an oblique rotation. The interfactor correlations from the oblique rotation ranged between -.12 and.16, indicating that although correlations between the factors were allowed, the resulting factors were correlated only slightly.

5 502 GILBERT AND BAILEY Year effects were significant for all four factors, 168d gain and all hair coat characteristics except length of undercoat and the number and percentage of medullation of guard hairs (Table 2). The LSM for all variables except number of guard hairs were less in 1%5 than in 1%6 Uable 3). Comparison of average monthly meteorological data for the gain-test periods from November 1, 1944 to February 28, 1%5 versus the same period in 1965 to 1966 indicate little difference in mean minimum temperature (-17.4 vs -15.7'C); lowest temperature (-31.7 vs C); total bright hours of sun, a measure of the amount of direct sunlight per month (84.3 vs 81.8 h) or total wind per month (13,703 vs 13,189 km). Total precipitation during the 1964 to 1965 period was nearly twice that of 1%5 to 1%6 (25.1 vs 13.2 mm). Personal recollection by a resident of the substation during the period of this study (J. de Wit, personal communication) indicated that the winter of 1964 to 1965 was much more severe than the following winter (e.g., more drifting snow and harsher weather conditions than what emerged from the monthly statistics). From the measurements available, we could not determine whether weather differences between the 2 yr had an impact on the results of this study. Compared with the HED, the MED increased (P <.02) all measures of hair weight per unit of surface area, undercoat number and guard hair medullation. Undercoat density (Factor 3) also was increased by MED, whereas guard hair density (Factor 2) was not (Tables 2 and 3). Yeates (1958) found that shedding of winter coat was delayed in Shorthorn heifers fed a diet that restricted gain to slightly above maintenance (.06 kg/d gain) compared with control heifers (.45 kg/d gain). Differences in dietary energy intake and resulting gain in our study were much less. Although our measurements were taken during midwinter, with no follow-up measures of the degree of shedding, the cattle fed the MED had heavier coats with more undercoat hairs than did the cattle fed the HED (Table 3). Whether these differences were due to retention of hairs in the coat with reduced energy intake and ADG could not be determined from our data. Bulls exceeded heifers in the diameter of all hairs, raw weight and length of guard hairs, but bulls had fewer guard hairs and less medullation of undercoat hairs. Heifers had finer, denser guard hair coats (Factors 2 and 4) than bulls. but undercoat density (Factor 3) did not differ. Older calves tended to have fewer guard hairs but more undercoat hairs. Coarseness of the coat (Factor 4), medullation of undercoats and diameter of all hairs increased and length of undercoat decreased with age (Tables 2 and 3). Breed differences were found in Factors 1 and 4 (length and medullation; coarseness) and all hair coat variables except the number of hairs in the undercoat and weights of both types (Table 2). Angus cattle tended to have shorter, less medullated coats (Factor l), shorter, larger-diameter undercoat hairs, and guard hairs with less medullation than Hereford cattle (Table 3). Although the Angus raw samples were lighter than those of the Herefords, no differences were found after the samples were cleaned (e.g., weights of undercoat and guard hairs, Tables 2 and 3), possibly because more foreign material and oils in the Hereford coats were removed during the cleaning process. Sire-within-breed differences existed for weight, length and diameter of both types of hairs and all four factors (Table 2). Genetic Parameter Estimates. Paternal halfsib heritability estimates, genetic and phenotypic correlations, and their respective SE are shown in Table 4. The heritability of and genetic correlations with undercoat medullation could not be computed due to a negative sire variance component estimate. Although the SE were large for many of the estimates, cew relationships did emerge. The number of both types of hairs and their amount of guard hair medullation were lowly heritable (h2 5.07), whereas all measures of hair coat weight were moderately heritable (.24 I h2 I.30). Guard hair length and undercoat diameter were highly heritable, whereas undercoat length and guard hair diameter had very low heritability estimates (Table 4). Heritabilities of overall coat length, guard hair density, and coarseness (Factors 1,2 and 4) were moderate, but that of undercoat density (Factor 3) was low (Table 4). Schleger and Tumer (1960) reported a heritability of.46 for weight of hair coat per unit area in Hereford and Shorthorn cattle, higher than our raw weight estimate of.30 f.13. They estimated heritability by regression of offspring on dam within sire so differences between the estimates included differences in the expectations of the two methods. No other comparable heritability

6 HAIR COAT CHARACTERISTICS AND GROWTH 503 c rim -?e I. P-0, oom

7 504 GILBERT AND BAILEY q. 0 E :

8 HAIR COAT CHARACTWISTICS AND GROWTH 505 estimates are available in the literature. The phenotypic correlations were high and positive between the number of hairs and the weight per unit area of both types of hairs (Table 4). The only other phenotypic correlations greater than.40 in magnitude were between the factors and their respective components. The respective genetic Correlations were similar in magnitude to the phenotypic correlations. In addition, other high positive genetic correlations (2.59) were found between Factors 2,3 and 4 and hair coat variables that were not components of those factors. In most instances, the hair coat variable had strong genetic correlations with one or more of the variables used in the interpretation of the factor. For example, the genetic correlation between guard hair diameter and Factor 3, undercoat density, was slightly higher than the correlations with undercoat hair weight and number, the major components of Factor 3 (Tables 1 and 4). Because the factors were linear combinations of all the hair coat variables, the heritability of a factor could be different from the heritability of one or more of its components. Turner and Schleger (1%0) devised a subjective scoring system for coat type that had higher heritabilities than individual hair coat characteristics in British and Zebucross cattle. Relationship with Gain. All sources of variation were significant for 168d gain (P <.02) except breed (Table 2). As expected, bulls had greater ADG than did heifers, and animals on the HED gained more than those fed the MED (Table 3). All the phenotypic correlations between hair coat characteristics and 168-d gain were low (Table 4, last row), whereas the genetic correlation between 168-d gain and guard hair number was strongly positive and between gain and undercoat weight was moderately negative (Table 4, last column). Our heritability estimate of.17 f.10 for 1686 gain is considerably lower than the.44 average of 36 paternal half-sib estimates summarized by Woldehawariat et al. (1977). When those estimates were weighted by the number of offspring per sire, the weighted average heritability was.32. The average number of offspring per sire in the 36 studies they summarized was 9, compared with 22.8 in our study. Little indirect selection response (Falconer, 1%0) would be expected from selection for a hair coat character or factor to improve postweaning gain, disregarding the complexity of hair coat measurement. With such selection, the product of the genetic correlation and the square root of the heritability of the secondary (e.g., hair coat) character must be greater than the square root of the heritability of the primary character (e.g., gain), assuming equal selection intensities for both traits. For example, selection favoring guard hair number or against undercoat weight each would result in a postweaning gain response of.24 compared, with.41 units from direct selection for 168d gain (calculated from values in Table 4). No advantage would be realized by using any of the four factors as indirect selection criteria to improve postweaning gain. Schleger and Tumer (1%0) reported that simple correlations were significant between gain and percenta e of medullation (.54), coat weight per &m5 area (-.37), hair thickness (.56), hair length (-.47), coat depth (-.63) and subjective coat score (-.72) in Hereford steers and heifers in Australia. Direct comparison of the results of these two studies is not appropriate. Although lower in magnitude, several simple correlations between postwean- i n g g a h a d h a i r c o a t ~ ~ s i g - nificant (P <.05) in our data, most notably guard hair length (r =.14) and undercoat weight (r = -.18). However, after adjustment for important sources of variation, the only statistically significant residual correlations with 168d gain in our study were guard hair number (r = -.15, P =.0003) and Factor 2, our measure of guard hair density (r = -. 11, P =.01). In a more detailed experiment using similar cattle, Turner and Schleger (1970) found that hair diameter and medullation were not useful predictors of body weight gain (P >.05). In that second study, Turner and Schleger (1970) obtained repeated hair samples from the same site on each animal during a 13-mo sampling period to determine both the number of old and new hairs since the previous clipping and the rate of hair growth. Only the number of new hairs was a useful predictor of gain because of a standardized partial regression coefficient different from zero. In both of their studies, growth was measured for a 13-mo period following weaning, and hair coat measurements were taken when the cattle were 2 yr of age, compared with only 4 mo of postweaning gain with hair coats sampled at 9 to 11 mo in our study. Considerable variation occd among the hair coat characteristics measured in this study. Heritabilities of many of the hair coat charac-

9 506 GILBERT AND BAILEY teristics were sufficient to expect a response to selection for those traits; however, none of the characteristics or factors used to consolidate them exhibited any useful relationship with growth rate. Improvement in 168d gain in response to selection for one or more hair coat characters would be less effective than direct selection for gain. Factor analysis was successful in reducing the number of variables from 10 to 4. Although the factor heritabfities and genetic and phenotypic correlations were consistent with the hair coat characteristics used in the respective factor interpretations, no additional benefit would be realized from use of the factors in place of the original variables unless the specific hair coat characteristics associated with a particular factor had economic value. Brody (1948) suggested several ways in which warm-blooded animals adapt to cold weather, these included increased feed consumption, to take advantage of the associated heat increment, and the development of insulating hair coats and subcutaneous fat layers. Webster et al. (1970) observed a 26 and 21% increase in total hay consumption of calves fed outdoors at -2832, with and without shelter, compared with controls fed indoors at 20'C. Individual feed intake measurements were not available for the cattle in the present study, so no hypotheses regarding energetic efficiency could be tested. Differences in internal insulation, such as subcutaneuus fat, may have compensated for differences in hair coat insulation such that growth rate differences in these cattle would not be related to hair coat differences alone. lmplicatlons Considerable variation was observed in the hair coat characteristics of young Hereford and Angus cattle, but differences were not closely associated with postweaning gain. Heritabilities of many of the hair coat characters were of sufficient magnitude to suggest that selection would be effective, but the corresponding genetic correlations with postweaning gain were too low to warrant indirect selection to improve gain. Winter feedlot performance of Hereford and Angus cattle using feeding practices typical of western Canada was not closely associated with hair coat variation. Literature Cited Brwdy, S Enviromnental physiology with special reference to domestic animals. I. Physiological back- grounds. Univ. Missouri Agric. Exp. Sta. Res. Bull Brown, C. J., J. E. Brown and W. T. Butts Evaluating nlationships among immature measures of size, shape, andperformance of beef bulls. I. Principal components aa measnns of size and shape in young Hereford and Angus bulls. J. Anim. Sci. 36:lOlO. Dowling, D. F. 1959a The medullation characteristic of the hair coat as a factor in heat tolerance of cattle. Aust. J. Agric. Res Dowlii D. F. 1959b. The significance of the coat in heat tolerance of cattle. Aust. J. Agric. Res Dowling, D. F llte significance of the coat in heat toleram? of cattle: II. Effect of soh radiation of body temperature. Anst. J. Agric. Res. 11:871. Falconer, D. S Intmduction to Quantitative Genetics. The Ronald Press Co., New Yo&. pindlay, J. D Physiological reactions of cattle to climatic stress. Roc. Nut. Soc Harvey, W. R User's Guide for LSMLMW. The Ohio State University, Columbus (mimeo). Johnson, R A. and D. W. Wichan Applied Multivariate Statistical Analysis. Rcntice Hall, Englewood Cliffs, NJ. Makarechian, M. A. Farid and R. T. Berg Separation of the effects of age, body weight and gain on pregnancy rates of beef heifers by principal component analysis. Can. I. Anim. Sci hlilligan, 1. D. and G. I. christison Effects of severe winter conditions on performance of feedlot steers. Can. J. Anim. Sci Pdas,H.F.andS.B. Slen Haircoatcharactensticsof bison. domestic x bison hybrids, cattalo, and certain domestic breeds of beef cattle. Can. J. Anim. Sci. 44: 48. SAS SAS User's Guide: Statistics. SAS Inst., Inc., Gary. NC. Schleger, A. V. and H. G. Tumtr Analysis of coat characters of cattle. Aust. J. Agric. Res. 11:875. Smoliak, S., and H. F. Peters climatic effects on foraging performance of beef cows on winter range. Caa J. Agric. Sci Sncdecor, G.W. and W. G. Cochraa Statistical Methods (6th Ed.). The Iowa State Univ. Press, Ames. Tmner, H. 0. and A. V. Schleger The significance of coat type in cattle. Anst. J. Agric. Res. 11:645. Tumer, H. G. and A. V. Schleger An aualysis of growth processes in cattle mats and their relations to coat type and body weight gain Aust. J. Biol. Sci. 23: 201. Webster, A.J.F Direct effects of cold weather on the energetic efficiency of beef production in different regions of Canada. Can J. Anim. Sci Webster, AJ.F., J. Chlumecky and B. A. Young Effect of cold environments on the energy exchanges of young beef cattle. Can. J. Anim. Sci Woldehawariat, G., M. A. Talamautes, R. R. Petty, Jr. and T. C. Camwight A summary of genetic and environmental statistics for growth and conformation charactem of young becf cattle. The Texas Agric. Exp. Sta. Tech. Rep. No College Station. Yeates, N.T.M Observations on the role of nutrition in coat shedding in cattle. J. Agric. Sci. (Camb.) Ycata, N.T.M. and W. H. Southcott Coat type in relation to cold adaptation in cattle. Proc. Aust. SOC. Anim. Rod. 2:102.