EVALUATION OF PHENOTYPIC AND GENETIC TRENDS IN WEANING WEIGHT IN ANGUS AND HEREFORD POPULATIONS IN VIRGINIA 1 K. Nadarajah 2, D. R. Notter, T. J. Marlowe and A. L. Eller, Jr. Virginia Polytechnic Institute and State University 3 Blacksburg 24061 ABSTRACT Weaning weight records of 27,774 Angus calves in 13 herds and 14,738 Hereford calves in 11 herds born during 1953 through 1983 in Virginia were analyzed using regression techniques and maximum likelihood (ML) procedures to estimate phenotypic and genetic trends for adjusted weaning weight (AWWT), weaning weight ratio (WWR) and deviations of AWWT from the mean AWWT of the contemporary group (DEV). Phenotypic trends for AWWT in the Angus and Here- fords were .96 +- .02 and .82 (cid:127) .03 kg/yr, respectively. In the Angus breed, estimates of one-half of the sire genetic trend obtained from the ML procedure for WWR and DEV were .40 (cid:127) .04 ratio units/ yr and .72 (cid:127) .07 kg/yr, respectively; corresponding values for Herefords were .25 -+ .06 ratio units/yr and .45 +- .12 kg/yr. Estimates of one-half of the dam trends for the respective traits were .32 +- .02 ratio units/yr and .55 +- .04 kg/yr for Angus and .21 +- .03 ratio units/yr and .30 +- .07 kg/yr for Herefords. Estimates of sire and dam genetic trends from the regression analyses were slightly higher than estimates from the ML procedure, but adjustments to eliminate bias due to non-random mating and culling from the regression analyses increased the similarlity of the results from the two procedures. Average annual genetic trends over the entire study period from the ML procedure for AWWT were 1.27 kg/yr for Angus and .75 kg/yr for Herefords. Genetic trends were not linear over the entire period. Total genetic trends in AWWT for Angus and Hereford, respec- tively, were .30 and -.61 kg/yr before 1971 and 2.18 and 1.98 kg/yr after 1970. (Key Words: Beef Cattle, Selection, Genetic Trend, Weaning Weight.) Introduction components vary depending on the experimen- Evaluation of genetic progress for economic tal situations (i.e., controlled laboratory ex- traits is an essential part of successful planning periments vs field experiments), but in all of future breeding schemes, and allows docu- approaches, regression procedures assume a mentation of progress from past selection. Such central role. an evaluation requires separation of the genetic Past efforts to estimate genetic trend in beef and environmental portions of the total pheno- cattle have been confined mostly to small, single- typic trend. The phenotypic trend in perfor- herd populations (Brinks et al., 1961, 1965; mance can be estimated straightforwardly by Nelms and Stratton, 1967; Bailey et al., 1971). regression of actual performance records on However, in recent years more emphasis has time. However, unbiased estimates of either been given to performance and progeny testing genetic or environmental trends are difficult to and extensive use of artificial insemination in obtain because of potential confounding of North American beef cattle and estimation of changes in environment with changes in geno- genetic progress from herds involved in regional type. Available procedures for appraising these and national programs is occurring (Kennedy and Henderson, 1977; Schaeffer et al., 1981; Crow and Howell, 1983; Zollinger and Nielsen, 1984). Nevertheless, there is a need for more investigations of genetic progress in beef cattle populations under varying environmental condi- The authors express their appreciation to the Vir- tions. Therefore, the objective of this study was ginia Beef Cattle Improvement Assoc., the Amer. Angus Assoc. and the Amer. Polled Hereford Assoc. to investigate the nature and magnitude of gen- for providing the data used in this study. etic changes in a selected sample of Angus and 2Present address: Dept. of Anita. Sci., Univ. of Hereford herds that have participated in the Vir- Guelph, Guelph, Ontario, Canada N1G 2W1. ginia Beef Cattle Improvement Association 3 Dept. of Anita. Sci. Received June 9, 1986. (BCIA) performance testing program for 20 yr Accepted January 4, 1987. or longer. 1349 J. Anim. Sci. 1987.64:1349-1361 1350 NADARAJAH ET AL. Materials and Methods Because birth dates were not available on all Source and Description of the Data. Data sires, the year in which a sire was introduced from 13 Angus and 11 Hereford herds from for the first time into a herd was defined as sire across the state of Virginia were used for this year (SYR). Thus, estimates of genetic trend study. All herds had been in the program within the herds will include both effects of continuously for 20 to 31 yr (1953 through within-herd selection and introduction of older 1983). Weaning weight records for the herds sires (possibly by AI) from outside the herd. were obtained from the Virginia BCIA, the Most sires were assumed to be about 2 yr of age American Angus Association and the American at the time of first use. The dam's birth year Polled Hereford Association. There were was incremented by two to match the sire year 27,774 Angus and 14,738 Hereford records time scale and denoted as dam year (DYR). available after records were eliminated for Thus, a genetic group in this study was a calves outside 125 and 275 d of age at wean- HERD-SYR-DYR combination. There were ing, without sire identification or outside the 3,138 HERD-SYR-DYR combinations, 1,045 designated calving seasons (table 1). sires and 7,669 dams represented in the Angus All herds had a primary breeding season breed and 2,019 HERD-SYR-DYR combina- corresponding to weaning in late summer or tions, 501 sires and 4,293 dams represented in early fall in all years. A few herds also had a the Hereford breed. Most sires were used for at secondary breeding season corresponding to least 2 yr. The average number of progeny per spring weaning in some years; eight herds (four sire was 29 in each breed. per breed) also had a tertiary breeding season A subset of the data from each breed that corresponding to weaning in late fall. Data from included only repeat matings was used for all three seasons were used in the analysis. The estimating repeatibilities. There were 10,085 primary season included almost two-thirds of such records in the Angus data and 6,342 the weaning records in each breed and the records in the Hereford data. tertiary season constituted less than 6% of the Statistical Methods. Environmental effects records of each breed. The feeding and manage- on calf weaning weight were investigated using ment was typical of purebred beef cattle a model that included discrete effects of herd, operations in Virginia. In many herds, at least weaning year, age of dam and sex of calf and some calves were creep-fed in most years. the continuous effect of calf age at weaning. Relatively more Herefords were creep-fed Age of dam was calculated as the difference be- (43%) than were Angus (18%). tween birth dates of calf and dam and cate- TABLE .1 NUMBER OF WEANING WEIGHT RECORDS, NUMBER OF YEARS AND WEANING WEIGHT MEANS AND STANDARD DEVIATIONS (SD) IN EACH HERD FOR ANGUS AND HEREFORD Angus Hereford Mean, SD, Mean, ,DS tterd No. records kg kg Herd No. records kg kg A 728 (31) a 180.8 33.6 N 1,902 (30) 208.5 45.4 B 1,958 (30) 186.4 30.8 O 2,201 (29) 191.8 41.2 C 3,138 (29) 162.7 36.4 P 624 (28) 184.8 37.6 D 967 (29) 214.5 42.9 Q 851 (25) 206.4 34.0 E 7,955 (29) 181.4 36.9 R 1,899 (21) 204.0 48.9 F 1,999 (28) 191.8 43.7 S 1,475 (21) 184.3 5.43 G 969 (26) 186.1 42.9 T 947 (23) 219.5 1.35 H 1,257 (27) 184.6 34.4 U 580 (20) 172.8 44.4 I 1,838 (26) 200.8 34.9 V 992 (20) 195.9 44.9 J 2,348 (26) 207.3 1.83 W 1,365 (20) 5.281 38.6 K 1,722 (24) 188.9 36.8 X 1,902 (20) 219.9 42.6 L 1,582 (22) 177.3 35.0 M 1,313 (20) 186.5 5.93 Total 27,774 185.8 37.2 14,738 199.4 42.9 avalues in parentheses refer to number of years for which records were available. GENETIC TRENDS NI BEEF CATTLE 1351 gorized into 2, 3, 4, 5 to 10 or > 10 yr groups. weights to allow preliminary inspection of Actual weaning weights (WWT) were adjusted phenotypic trends. Subsequent genetic analyses for effects of age of darn and sex and age of were conducted on a within-sex basis . This calf (to 205 d) before further analysis. approach may result in some bias if different Differences among dam age groups in wean- proportions of male progeny are castrated ing weight (table 2) were smaller than those (selected) for different sires. normally assumed in industry performance- For calves that did not have birth weights testing programs (BIF, 1981). They were like- recorded, breed average birth weights of 31.8, wise somewhat smaller than those reported for 29.5, 34.0 and 31.8 kg for Angus males, Angus Angus and Hereford cows in experimental herds female, Hereford males and Hereford females, (Smith et al., 1976; Ochoa et al., 1981). Culling respectively, were assigned for computing 205-d of inferior cows can result in overestimation of adjusted weights. Possible phenotypic trends in performance levels for older cows and concom- calf birth weights were not considered. A itant over-estimation of adjustment factors weaning weight ratio (WWR) was computed for (Henderson et al., 1959). However, because each calf as the ratio of adjusted 205-d weight adjustment factors derived from table 2 were to the average of the contemporaries' adjusted already smaller than most published estimates, 205-d weights. The contemporary group was they were used in preference to published fac- composed of calves of the same herd, weaning tors to derive additive adjustment factors for year, season, sex and management group age of dam. The same age-of-dam adjustments (creep-fed or not creep-fed) within each breed. were used for both creep-fed and non-creep-fed The adjusted weaning weight (AWWT), WWR, calves. Ochoa et at. (1981) reported that creep- and deviation of AWWT from the contempo- feeding of bull calves resulted in only minor rary group mean (DEV) were used as indepen- changes in age-of-dam effects on weaning dent variables in subsequent analyses. weight. Heritabilities (h 2) and repeatibilities (t) for The difference in weaning weight between traits of interest were also estimated from the bulls and heifers was larger than most published adjusted data. To estimate the sire variance estimates from experimental herds (Koch et al., component, a nested model containing effects 1974; Ochoa et al., 1981) and the difference of sires within weaning year and contemporary between steers and heifers was correspondingly group was used for within-herd analyses and a smaller (Sagebiel et al., 1974). This result may model including sires within herd, weaning year reflect selection of inferior males for castration. and contemporary group was used for the Additive adjustment factors for sex derived among-herd analyses within each breed. Effects from table 2 were used to adjust weaning of sires were considered random with zero TABLE 2. LEAST-SQUARES SNAEM DNA STANDARD ERRORS FOR MAD-FO-EGA DNA SEX-OF-CALF EFFECTS NO WEANING WEIGHT (KG) DNA REGRESSION COEFFICIENTS FOR GNINAEW WEIGHT NO AGE OF CALF AT GNINAEW (KG/D) FOR THE OWT SDEERB Breed Effect Angus Hereford egA of dam 2 yr 177.8 (cid:127) 75. 187.8 (cid:127) .69 3 yr 187.0 (cid:127) 192.2 + .76 4 yr 192.4 (cid:127) .62 203.9 (cid:127) 77. 5 o~ 01 yr 194.8 (cid:127) .49 207.5 +- 15. >10 yr 188.0 (cid:127) 16. 205.3 + 1.13 xeS of calf Bull 204.0 (cid:127) .54 222.2 (cid:127) 85. refieH 3.571 (cid:127) .49 3.881 (cid:127) 54. Steer 184.8 -+.53 191.2 (cid:127) .72 Regression of calf ega .67 (cid:127) .005 17. (cid:127) .008 1352 NADARAJAH ET AL. mean and variance equal to one-quarter of the for dams does not change systematically over additive genetic variance. A nested analysis was time, if all females are produced from within used in preference to an analysis in which sires the herd and if the relative age structure of and years were cross-classified because of the sires and dams remains constant, the difference large number of missing sire-year subclasses and S3/( - 3D) provides an unbiased estimate of the the high level of confounding between sires and total maternal trend. Tests of significance of years. Resulting heritability estimates therefore estimates of genetic trend were based upon the are intra-year estimates and resulting sire standard errors of the corresponding regression variance components would include any sire x coefficients. These standard errors do not year interaction variance. Standard errors of include the contributions of genetic drift to heritability estimates were obtained using the variation in response (Hill, 1972) and are conservative estimation procedure of Dickerson therefore conservative. (1969). Estimates of repeatability were ob- Cicogna et al. (1982) suggested procedures tained from regression of second observation on to adjust the sire genetic trend for assortative first observation for each trait in repeat-mating mating and culling of cows such that: data. Estimates of heritability and repeatability were subsequently used to allow mixed model 1/2AGs = (-3S + 1/2AD)/ maximum-likelihood (ML) absorption of ran- (1 + AA). 11 dom effects of sire, dam and mating (Hender- son, 1973). The factor AD is obtained by deviating dams' Two main approaches were taken to obtain first-calf WWR or DEV from the mean first-calf estimates of phenotypic, environmental and WWR or DEV of all females contemporary to genetic trends. The first approach used a series the dam, then calculating the intra-sire regres- of regression techniques essentially like those sion of that deviation on weaning year, and developed by Smith (1962) and used for beef finally multiplying the regression coefficient by cattle by Zollinger and Nielsen (1984). For the h 2 to account for the imperfect heritability of regression approach, the genetic trend attrib- the first calving record. Thus: utable to sires was estimated from the within- sire regression of WWR or DEV on weaning AD = h2~(WWR1 _ WWRMc).WYR/sire year (3S = 3WWR.WYR/sire). This regression coefficient measures the extent to which where WWR1 is the first-calving record and progeny of a given sire lose ground relative to WWRMc is the mean first-calving record of the their contemporaries over time, and therefore dam's contemporaries. The factor AD thereby has an expectation of a negative one-half of the measures intra-sire trend in dam's additive overall genetic trend associated with sires genetic merit relative to all possible mates (ZoUinger and Nielsen, 1984). Thus 3S = within a contemporary group and allows adjustment for assortative mating and(or) --1/2AG S. The genetic trend attributable of dams was culling within the contemporary group. Com- estimated from the intra-dam regression of the parable adjustment of AG D is not required because records are deviated from the paternal deviation of WWR or DEV from the paternal half-sib mean. half-sib (PHS) mean on weaning year 3D = The factor AA reflects changes over time in ~(WWR -- WWRpHs).WYR/dam. 3D thereby measures the extent to which progeny of a the mean age of mates of a given sire. It is given female lose ground over time in relation calculated as the difference between the intra- sire and overall regressions of dam age (A) on to their paternal half-sib contemporaries. Unbiased estimation of 3D is dependent upon weaning year such that: unbiased estimates of adjustment factors for age of dam in order to uncouple dam genetic AA =/3A.WYR/sire --/3A.WY R. trends from cow age effects (Henderson et al., The second major approach for estimation 1959). 3~ D has as its expectation a negative of genetic trend utilized maximum likelihood one-half of the overall transmitted effect of the procedures for a mixed model: dam plus the negative of the total maternal trend (both genetic and permanent environmen- Yijklmn =/a + Hi + SYRj + tal). Thus, flD = --1/2AGD -- .M3Z If there is no DYRk + Sijl + Dikm + assortative mating, if the selection differential WYRn + Eijklmn, GENETIC TRENDS IN BEEF CATTLE 1353 where Yijklm is the observed AWWT, WWR or In general, h 2 for WWT has been in the range DEV; of. 10 to .60, with an average of approximately .30 (Woldehawariat et al., 1977). The estimates a/ is a constant common of all obser- obtained in this study are lower than the vations; estimates of .72 + .33 and .82 + .12 for WWT H i is the fixed effect of the i th herd; reported by Francoise et al. (1973) using SYRj is the fixed effect of the jth sire year; similar BCIA records from Hawaii for Angus DYR k is the fixed effect of the k th dam year; and Hereford cattle but higher than the esti- Sij 1 is the random effect of the lth sire; mates of .30 and .31 reported by Nelsen and Dik m is the random effect of the mth dam; Kress (1979) for Angus and Hereford cattle WYRn is the fixed effect of the nth weaning from Montana Beef Cattle Performance Associ- year, and ation records. The estimate of .51 + .06 for Eijklmn is the residual error. Hereford AWWT in this study is in good agree- ment with the paternal half-sib estimates of .50 Estimates of total phenotypic trend and of + .10 reported by Vesely and Robison (1971) genetic trends attributable to sires and dams for Hereford WWT. Schalles and Marlowe were obtained using the following procedure: (1971) also reported h 2 of .57 +- .09 for pre- 1) Ignore effects of SYR, DYR, S and D and weaning average daily gain for Angus cattle estimate constants for WYR to provide an using data from a herd that was included in the estimate of phenotypic trend in AWWT. present study. Average heritability estimates for 2) Absorb effects of H, SYR and S. Ignore DEV and WWR from both breeds gave a mean effects of DYR and D. Constants for WYR for value of .475; thus for subsequent calculations DEV and WWR estimate the negative of the in the mixed model ML procedures an intra- genetic effects of SYR adjusted for culling of class correlation of .12 was assumed for sires. sires. Repeatability estimates for WWR in repeat 3) Absorb effects of H, DYR and D. Ignore matings were .34 and .36, respectively, for effects of SYR and S. Constants for WYR for Angus and Hereford. Petty and Cartwright WWR and DEV estimate the negative effect of (1966) reported a weighted average repeat- DYR adjusted for culling of dams. ability of weaning weight of .44 based on 16 4) Regress WYR constants on WYR for each estimates found in the literature. Variance com- of the three models to obtain estimates of ponents for sires, dams unadjusted for sires and phenotypic, sire-transmitted genetic and dam- matings (sire and dam) have expectations of transmitted plus maternal trends. Estimate 2 2 2 2 2 1/4OG, 1/4o G + o M and 1/2o G + o M , respec- environmental trend and dam maternal trend 2 tively, where GO is the additive genetic variance by difference. Note that this procedure assumes that dams are assigned to sires at random. All and MO 2 is the maternal genetic plus permanent calculations were made using procedures maternal environmental variance. Thus, in these outlined by Harvey (1982). data using estimates of h 2 of .48 and of repeat- ability in repeat matings of.35, sires, dams and Results and Discussion matings would be expected to account for an Means, Variance Components and Genetic average of 12, 23 and 35%, respectively, of the Parameters. Phenotypic means and standard total phenotypic variance in WWT traits. deviations for actual weaning weight (WWT) for Within-Herd Sire and Dam Genetic Trends. individual herds are shown in table 1. Average Estimates from the regression analysis of WWT of Angus calves was 185.8 kg, which is one-half of the genetic trends attributable to 13.6 kg lower than the average WWT of Hereford sires and dams are presented for each herd in calves (199.4 kg). Heritability estimates for tables 3 and 4. In the Angus breed, estimates of AWWT, DEV and WWR were .63 + .08, .58 (cid:127) sire trends were positive in all herds except one, .08 and .53 + .14, respectively, for Angus and ranging from --.27 +- .14 to 1.25 (cid:127) .26 ratio .51 + .06, .43 + .06 and .37 + .11 for Herefords. units/yr for WWR and from -.70 +- .27 to 2.88 These estimates are higher than average values (cid:127) .53 kg/yr for DEV and were significant in for weaning traits reported in the literature. In nine of 13 herds. Sire trends were generally part, the relatively high estimates of herit- larger for herds that entered the program later. ability obtained from the current data may The estimates of dam trends for Angus were reflect inclusion of sire (cid:141) year interaction also positive in all herds except one and ranged variance in the sire variance component. from -.01 (cid:127) .08 to .63 - .08 ratio units/yr for 1354- NADARAJAH ET AL. 1+ +l 1+ 1+ 1+ 1+ +l +i 1+ 1+ 1+ 1+ 1+ I > ,.. ~7 1+ 1+ ~ 1+ +t +l 1+ +t 1+ ~ ~ 1+ +i ~Z z= Z ~m I ~:< ;>. < ... % ,,,z, 1+ ~ ~ ~ 1+ N ~ 1+ 1+ +l 1+ 1+ +l > 31 t I > 1.T., '?- i I g IIu U II il .Q ,-4 V ~.., .x. t .X. t 4l..It, GENETIC TRENDS IN BEEF CATTLE 1355 ._; . . . . . ~ (cid:12)9 - -,< ~Z ~g Z a.r Z~ : ~ ~. .~:~. z~ 4< < . . ~ . . . ~ . . ~ % . , ~ . . . . . . . I II II II .0 V~ 1356 NADARAJAH ET AL. WWR and from -.23 +- .17 to 1.75 + .26 kg/yr for DEV. Eleven of these estimates were significant. Zollinger and Nielsen (1984) used a similar regression approach on records from 15 herds l l l l of Angus cattle enrolled in the American Angus Association program. Their estimates of genetic trend for WWR ranged from .01 + .23 to 1.30 +- .24 ratio units/yr for sires and from .06 + .06 to .68 +- .11 ratio units/yr for dams. z In the Hereford breed, most herds exhibited positive sire trends but the estimates were 0 significant in only two herds. Negative estimates e~ were obtained for WWR and DEV in a few herds. The estimates were large in absolute (cid:12)9 . . . . +1 +1 +1 +1 +t magnitude but not significant due to their large < standard errors. All herds showed positive dam I I I I trends and eight of 11 of the estimates were significant for both traits. Dam trends ranged o~ from .05 + .11 to .93 +- .15 ratio units/yr for WWR and from .19 +-- .26 to 2.15 + .29 kg/yr o for DEV. This result suggests substantial genetic change in maternal ability may have occurred ,.. )L ,-. in the Hereford breed during this period. Again, herds that entered the program later tended to N have larger genetic trends. Estimates of Genetic Trends from Regres- 2" sion Procedures. Overall estimates of pheno- typic and sire and dam genetic trends for WWR zm and DEV from the regression procedure are given me~ for each breed in table 5. Sire and dam genetic trend estimates were positive (P<.001) in both breeds. However, the sire contribution in the Angus breed was larger than the dam contribu- tion (P<.10 for WWR; P<.40 for DEV), whereas, 0 the dam trend was larger in the Hereford breed (P<.10 for WWR; P<.05 for DEV). Estimates ~e of environmental trend in AWWT were -.58 (cid:127) ua< N TTTT .10 for Angus and -.48 + .14 for Hereford. The estimate of sire trend of .44 ratio units/yr in WWR for Angus in this study was lower than the average estimate of .51 ratio units/yr computed from 15 intra-herd estimates Z by Zollinger and Nielsen (1984), but the dam 1-., trend was in good agreement with their average r,1 estimate of .34 ratio units/yr. The simple M < regression of sire proof constants on year of birth of the calf in the study of Schaeffer et al. (1981) showed genetic gain of .67 kg/yr in Angus and .21 kg/yr in Hereford. Thus, the results in this study indicate that the rate of .O o genetic change for weaning performance has v been moderate and in the favorable direction o for both breeds. GENETIC TRENDS IN BEEF CATTLE 1357 TABLE 6. GENETIC TRENDS (G) COMPUTED FOR WWR (RATIO UNITS/YR) AND DEV (KG/YR) IN ANGUS AND HEREFORD BREEDS AND THE ADJUSTMENT FACTORS USED FOR NON-RANDOM MATINGS AND GENETIC MERIT OF THE DAMS a Breed Traits 1/2AG S AD AA G b Angus WWR .44 --.02 .21 * * .72 DEV .82 -.08 .21"* 1.36 Hereford WWR .28 .08 .14" * .68 DEV .51 .06 .14" * 1.26 asee text for definitions of AGs, AD and AA. bG -- (1/2AG S + 1/2AD)/(1 + AA) + 1/2AG D. **P<.01. There is often a tendency for older bulls to effects of trend in dam's age and for culling be mated to older cows. These older cows as a using equation 1. The total genetic trend group are expected to be genetically inferior to computed after adjusting the sire trend for younger cows if selection is occurring within non-random mating and changes in the average the herd, but this expectation may not be genetic merit of the dams is shown in table 6. realized if culling within the older cows tends The within-herd regression of dam age on time to produce genetic trend within remaining was -.023 -+ .002 yr/yr for Angus and -.027 +- females. Thus, the estimate of sire genetic .002 yr/yr for Hereford. Corresponding regres- trend obtained in this study was adjusted for sions calculated within herds and sires for the 20 ..-.. 61 (cid:12)9 DROFEREH o SUGNA Z 21 0 Jb 3:( 4 I.- :Z 0 -d o z z w -8 Q w -12 I- -.j -16 Q -20 0 ~ ,54 .50 60 63 66 69 72 75 78 18 84 YEAR Figure 1. Phenotypic constants for adjusted weaning weight for Angus and Hereford cattle. 1358 NADARAJAH ET AL. 21 9 I/z SIRE TREND (cid:12)9 ~z DAM TREND 6 3 > 0 Ld D -3 -6 -9 -12 I I I I I l I I t t t t t i t i I t I | l l l | l l 55 60 65 70 75 80 WEANING YEAR Figure 2. Genetic trends for weaning weight in Angus cattle. respective breeds were .189 -+ .009 yr/yr and these biases were .72 ratio units/yr for WWR .114 + .011 yr/yr. Thus, AA, the adjustment for and 1.36 kg/yr for DEV in Angus. Correspond- non-random mating with regard to age of ing figures for Herefords were .68 ratio units/yr mates, indicated that for each year a bull and 1.26 kg/yr. Thus estimates of genetic trend remained in service he was likely to be mated to tend to be inflated by non-random mating with cows averaging .21 yr older in the Angus breed regard to cow age. Adjustment for this bias and .14 yr older in the Hereford breed than his reduced the difference between estimates of mates in the previous year. Thus, the relatively overall trend for the two breeds. poorer progeny performance observed as a bull Estimates of environmental trend were re- ages is in part attributable to increases in the duced in absolute magnitude by adjustment for age of his mates relative to the mates of younger non-random mating to -.40 for Angus and bulls. -.44 for Hereford. Henderson et al. (1959) The adjustment factor AD accounts for have shown that failure to account for effects changes in the genetic superiority of mates as a of culling and non-random mating can lead to bull ages expressed relative to all possible mates serious under-estimation of environmental in the herd. The estimates of AD for WWR and trends. Adjustment for non-random mating DEV were --.02 ratio units/yr and -.07 kg/yr, reduced this tendency, but final estimates of respectively, for Angus and .08 ratio units/yr environmental trend remained large and nega- and .06 kg/yr for Hereford. Thus assortative tive. mating within dam age groups was not observed. Estimates of Genetic Trends from Maximum Total genetic trends estimated after eliminating Likelihood Procedures. Phenotypic constants
Description: