- Email: [email protected]
Genetic Change in Milk, Fat, Days Open, and Body Weight After Calving Based on Three Methods of Sire Selection1
GENETICS AND BREEDING Genetic Change in Milk, Fat, Days Open, and Body Weight After Calving Based on Three Methods of Sire Selection1 J. M. Abdallah and B. T. McDaniel Department of Animal Science, North Carolina State University, Raleigh 27695-7621
YOUNG = young sires selected on pedigree for 3.7% FCM.
ABSTRACT Three Holstein lines, were compared, based on different methods of sire selection, for genetic change in 3.7% FCM, fat yield, days open, and predicted body weight after calving. The three lines were 1) evaluated sires selected only for 3.7% FCM (milk line), 2) evaluated sires selected on an index that included 3.7% FCM and type traits (index line), and 3) young bulls selected on pedigree for 3.7% FCM (young line). Cows from these lines were born in 1971 through 1993 in five experimental herds owned by the State Farm Division of North Carolina Department of Agriculture. Breeding values of cows in each line computed with a repeatability model were averaged by and regressed on birth year to estimate genetic change. Genetic gains in 3.7% FCM were 81 kg/yr for the milk line, 61 kg/yr for the line selected on index, and 68 kg/yr for the young sire line. Estimates of genetic gain in fat yield were 2.99, 2.16, and 2.54 kg/ yr in the three lines, respectively. Genetic gains in 3.7% FCM and fat yield in the milk line were significantly different from the index and young sire lines, but the index and young sire lines were not significantly different. Estimates of genetic change in days open were 0.71, 0.57, and 0.63 d/yr in the milk, index, and young sire lines, respectively. These estimates were not significantly different. Average breeding values for body weight decreased for births from 1971 to 1981 then rapidly increased for later births in all lines. (Key words: genetic change, index, evaluated sires, young bulls) Abbreviation key: BWC= predicted body weight after calving, DO = days open, INDEX = evaluated sires selected on index for 3.7% FCM and type traits, MEFAT = 2X, 305-d, mature equivalent fat yield, MILK = evaluated sires selected only for 3.7% FCM,
Received August 9, 1999. Accepted January 17, 2000. Corresponding author: J. M. Abdallah; e-mail: jmabdall@unity. ncsu.edu. 1 A contribution from Regional Research Projects S-251 and S-284. 2000 J Dairy Sci 83:1359–1363
INTRODUCTION The main goal for dairy cattle selection programs is to maximize genetic progress and profitability of production. The key to genetic progress is the use of the best sires (10). Milk traits are sex-limited, thus selection of sires is based on pedigree information or progeny testing (2). In a population with positive genetic trends, younger animals should be on average superior to older animals, and, therefore, rapid genetic gain may be achieved by selecting and sampling young bulls (9). AI organizations choose young bulls based on pedigree, then test them by progeny performance to identify favorable mendelian sampling before extensive use in AI. Many dairy farmers do not participate in young sire sampling programs (13). It seems that dairy producers feel confident in using older evaluated bulls but remain convinced that increased use of young bulls will result in reduction of genetic gain (5). Loyd and Hargrove (4) found that daughters of progeny tested sires had higher production of milk, fat and protein than daughters of young sampling sires. Vinson and Freeman (12) reported that the mean progeny performance of bulls returned to service exceeds the mean progeny performance of all bulls sampled by 146 and 5.0 kg for regressed deviated milk and milk fat yield. Others (5, 6) have found little or no differences in production between contemporary daughters of young sires and evaluated sires. Study of genetic differences among categories of service sires is needed for educational programs and decision making by the industry and dairy producers (10). The objective of this study was to compare the genetic change of cows for 3.7% FCM, fat yield, days open (DO), and predicted body weight after calving (BWC) by using three methods of sire selection. MATERIALS AND METHODS In 1970 a selection experiment was started in five experimental dairy herds owned by the State Farm Division of North Carolina Department of Agriculture.
1359
1360
ABDALLAH AND MCDANIEL
Sires were chosen from among active AI bulls by one of three methods. These methods were 1. Selection among progeny-tested bulls only for 3.7% FCM (MILK line). 2. Selection among progeny-tested bulls based on a final index that included 3.7% FCM, udder index, and feet and legs index (INDEX line). Udder index, feet and legs index, and the final index were computed as follows from standardized sire summaries: Udder index = [4 × udder support + 3 × teat placement + 2 × udder depth + fore attachment + 0.5 × rear udder width + 0.5 × rear udder height]/11. Feet and legs index = [3 × foot angle - absolute value of rear legs]/3. Final index = [3 × (udder index) + (feet and legs index) + 8 × (3.7% FCM)]/12. 3. Selection among young bulls based on their pedigree indexes for 3.7% FCM (YOUNG line). The pedigree index of a young bull was calculated as I = 0.5 (PTAs + PTAd), where PTAs and PTAd are the predicted genetic merits of the sire and the dam of the bull for 3.7% FCM. In the late summer of each year, 6 to 12 young bulls were selected from approximately 150 bulls to be sampled. Young bulls were chosen from three cooperating AI organizations with a few exceptions. Preference was given to the highest-ranking bulls that would have their first semen available between September 1 and December 31. No more than three sons of any one sire were used in a single year. In later years preference was given to young bulls that were not closely related to those used in previous years. Approximately 90 units of semen was obtained from each bull with smaller amounts in the year when more young bulls were chosen. Each year, 3 to 10 bulls were chosen from each of the MILK and INDEX lines from all active AI bulls in the US with repeatabilities of 60% or more. The number chosen depended on several criteria. These included the reliability of PTA, how much the top bulls differed in their progeny tests, the cost and availability of semen, and the kinships among the bulls in a line. The rule was to purchase 150 units from each bull. For the highestranking bulls with limited supply or those with high semen costs, as few as 50 units were purchased. Cows were bred to bulls from their own line, and mating was managed such that inbreeding in any calf did not exceed 6.25%. Cows that were not pregnant after four inseminations with evaluated sire semen were bred with young sire semen. Traits of interest were 3.7% FCM, mature equivalent fat yield (MEFAT), DO, and BWC. Yield traits were Journal of Dairy Science Vol. 83, No. 6, 2000
Table 1. Number of animals and records, means, and standard deviations for FCM, fat yield, days open and predicted body weight after calving by sire line. Sire Line 1
2
Milk
Index3
Young4
Sires(n) Cows (n) Records (n)
117 1542 3760
142 1588 3958
233 1534 3806
Mean, kg SD
9170 1794
9074 1723
8884 1791
Mean, kg SD
337 70
336 67
330 70
Mean, d SD
153 83
152 84
150 83
Mean, kg SD
592 85
594 84
596 88
Trait
3.7% FCM MEFAT DO BWC
1 MEFAT = 2X, 305-d, mature equivalent fat yield; DO = days open; and BWC = predicted body weight after calving. 2 Evaluated sires selected for 3.7% FCM only. 3 Evaluated sires selected by index for 3.7% FCM and type traits. 4 Young bulls selected by pedigree for 3.7% FCM.
on 2X, 305-day, mature equivalent basis. Days open (DO), as measured from calving to conception, were recorded to the nearest 10 d. If DO were greater than 310 d, it was set to 310 d. If a cow was sold or removed for other reasons without being pregnant, number of DO was calculated from calving day to the day it was sold or removed. The BWC of cows calving before 1988 were predicted from a heart girth-weight table developed by Kendrick and Parker (3). After 1988, mostly scale weights were used. Heart girths and actual body weights were measured at or before the first test day after calving. Record number, means, and standard deviations of traits are in Table 1. Breeding values of cows for each trait were obtained in a previous study (1) by using MTDFREML programs and single-trait repeatability model that included relationships and genetic groups. Breeding values were averaged by birth year of cows within sire line. The annual genetic change in a trait was estimated as the linear regression of average breeding values on year of birth. The differences in annual genetic changes among lines were tested using the CONTRAST statement in general linear model procedure (PROC GLM) of SAS software (11). RESULTS AND DISCUSSION Average breeding values for 3.7% FCM of cows born in 1971 through 1993 in the MILK, INDEX, and
1361
SIRE SELECTION METHODS AND GENETIC CHANGE
Figure 1. Average breeding values of 3.7% FCM by sire line. ♦, MILK = evaluated sires selected for 3.7% FCM; •, INDEX = evaluated sires selected by index for 3.7% FCM and type traits; and ▲, YOUNG = young bulls selected by pedigree for 3.7% FCM.
YOUNG lines are in Figure 1. Breeding values increased in all three lines. Average breeding values of cows in the MILK line were higher than in the INDEX and YOUNG lines over most of the period. Only small differences existed in breeding values of YOUNG and INDEX lines. Annual genetic gains were 80.9, 60.9, and 67.7 kg/yr for MILK, INDEX, and YOUNG lines, respectively (Table 2). The differences were highly significant (P < 0.01) between the MILK and each of the INDEX and YOUNG lines. Genetic changes in INDEX and YOUNG lines did not differ significantly (P > 0.10). Average breeding values for MEFAT are in Figure 2. Cows in the MILK line had higher average genetic merit than cows in the INDEX and YOUNG lines in 1978
Figure 2. Average breeding values of fat yield by sire line. ♦, MILK = evaluated sires selected for 3.7% FCM; •, INDEX = evaluated sires selected by index for 3.7% FCM and type traits; and ▲, YOUNG = young bulls selected by pedigree for 3.7% FCM.
through 1988. The INDEX line was inferior to the other two lines after 1987. Annual genetic gain in fat yield (Table 2) was 2.99 kg/yr for the MILK line, 2.16 kg/yr for the INDEX line, and 2.54 kg/yr for the YOUNG line. The MILK line was different from INDEX and YOUNG lines (P < 0.05), but the INDEX and YOUNG lines were not different (P > 0.10). Breeding values for DO (Figure 3) slowly increased for births before 1989 but increased rapidly thereafter for all lines. Yearly genetic change in DO did not differ between lines. Estimates of annual genetic change were 0.71, 0.57, and 0.63 d/yr in the MILK, INDEX, and YOUNG lines, respectively (Table 2). Higher breeding
Table 2. Estimates of annual genetic change in FCM, fat yield, days open, and predicted body weight after calving by sire line. Sire Line 1
Index2
Milk
4
Trait
3.7% FCM, kg MEFAT, kg DO, d BWC, kg 1971–1981 1981–1993
Annual genetic change b
SE
Annual genetic change a
Young3
SE
Annual genetic change a
SE
80.9 2.99b 0.71
3.8 0.13 0.10
60.9 2.16a 0.57
5.0 0.20 0.07
67.7 2.54a 0.63
7.2 0.29 0.20
–3.58a,b 3.46b
0.31 0.13
–2.92a 3.27b
0.30 0.34
–4.30b 2.54a
0.45 0.16
a,b
Values in a row with different superscripts are significantly different (P<0.05). Evaluated sires selected for 3.7% FCM. 2 Evaluated sires selected by index for 3.7% FCM and type traits. 3 Young bulls selected by pedigree for 3.7% FCM. 4 MEFAT = 2X, 305-d, mature equivalent fat yield; DO = days open; and BWC = predicted body weight after calving. 1
Journal of Dairy Science Vol. 83, No. 6, 2000
1362
ABDALLAH AND MCDANIEL
Figure 3. Average breeding values of days open by sire line. ♦, MILK = evaluated sires selected for 3.7% FCM; •, INDEX = evaluated sires selected by index for 3.7% FCM and type traits; and ▲, YOUNG = young bulls selected by pedigree for 3.7% FCM.
values for DO means more days from calving to conception and, thus, lower reproductive efficiency of cows. Averages of breeding values for BWC are in Figure 4. Breeding values rapidly decreased in the three lines for births from 1971 through 1981 but increased for cows born in later years. From 1971 to 1981, breeding values in the YOUNG line decreased more than in the INDEX and MILK lines but increased less (P < 0.05) for births from 1981 to 1993 (Table 2). Annual genetic changes in Table 2 were approximated by linear regression. All traits showed a slight curvilinear relationship to year of birth. But the linear regression accounted for most of the sums of squares,
and the significant differences among lines did not change when quadratic and cubic effects of birth year were included (results not shown). Working with a subset of the data used in this study, McDaniel and Bell (6) found no significant differences among the three lines in genetic gain of milk yield. Powell and Norman (10) reported that average evaluations of young bulls were similar to those of active AI bulls, but with differential usage considered, active bulls averaged 108 kg more for predicted difference of milk. Loyd and Hargrove (4) found that daughters of progeny tested sires outproduced daughters of sampling sires by 146, 52, and 174 kg of milk yield in first, second, and third generations and by 1.2, 0.8, and 1.1 kg for fat yield. McDaniel et al. (7) compared the three lines and found that the MILK line had the highest income over feed costs followed by the YOUNG line. They (7) also found that cows in the MILK line may have had slightly more days to first service (81 d for the MILK line vs. 79 d for the INDEX and the YOUNG lines). McMahon et al. (8) reported that the average profitability of young sires was equal to average profit from evaluated bulls when young sire semen cost $5/ unit and was $36 more profitable than average evaluated sires when young sire semen cost was nil. CONCLUSIONS Cows in the MILK line attained the highest genetic gain in milk and fat yields but tended to have more DO compared to cows in the INDEX and the YOUNG lines. Cows sired by young bulls were competitive in production with cows sired by evaluated bulls selected by index for milk and type traits. Although more risk is associated with using young sires, use of the highest on pedigree merit is expected to result in substantial genetic improvement. The use of young bulls becomes more advantageous when the lower price of semen from these bulls compared to semen from evaluated bulls is considered. Sires selected on an index that included milk yield and type traits had progeny with lower production than sires selected only for milk yield. Putting emphasis on type traits could hinder genetic gain in yield traits. In all three lines, DO increased as breeding values for 3.7% FCM increased. All lines showed significant decreases in breeding values for body weight from 1971 to 1981 but large increases from 1981 to 1993. REFERENCES
Figure 4. Average breeding values of predicted body weight after calving by sire line. ♦, MILK = evaluated sires selected for 3.7% FCM; •, INDEX = evaluated sires selected by index for 3.7% FCM and type traits; and ▲, YOUNG = young bulls selected by pedigree for 3.7% FCM. Journal of Dairy Science Vol. 83, No. 6, 2000
1 Abdallah, J. M. 1998. Genetic parameters and genetic trends in North Carolina experimental dairy herds. M.S. Thesis, North Carolina State Univ., Raleigh. 2 Freeman, A. E. 1975. Choosing and sampling young bulls: theory, background, and general problems. J. Dairy Sci. 58:1063–1070.
SIRE SELECTION METHODS AND GENETIC CHANGE 3 Kendrick, J. F., and J. B. Parker. 1936. Estimating the weights of dairy cows from heart girth measurements. BDIW 695. United States Department of Agriculture, Washington, D.C. 4 Loyd, B. M., and G. L. Hargrove. 1988. First lactation comparisons between daughters of summarized sires and sampling sires in artificial insemination. J. Dairy Sci. 71(Suppl. 1):235. (Abstr.). 5 McCraw, R. L., K. R. Butcher, and B. T. McDaniel. 1980. Progeny tested sires compared with pedigree selected young sires. J. Dairy Sci. 63:1342–1350. 6 McDaniel, B. T., and W. E. Bell. 1989. Genetic gain in milk yield from three systems of sire selection. J. Dairy Sci. 72(Suppl. 1):68. (Abstr.). 7 McDaniel, B. T., R. E. Collins, L. E. Pitzer, R. H. Lee, W. E. Bell, C. H. Brown, P. L. Williams, J. C. Wilk, and C. V. Tart, Jr. 1995. Comparison of three sire selection systems in Holsteins. ANS Report No. 247. Dep. Animal Science, North Carolina State Univ., Raleigh.
1363
8 McMahon, R. T., R. W. Blake, C. R. Shumway, and M. A. Tomaszewski. 1985. Selection of young and proven Holstein artificial insemination sires to maximize profits from milk. J. Dairy Sci. 68:2303–2308. 9 Nizamani, A. H., and P. J. Berger. 1996. Estimates of genetic trend for yield traits of the registered Jersey population. J. Dairy Sci. 79:487–494. 10 Powell, R. L., and H. D. Norman. 1989. Genetic differences among categories of service sires. J. Dairy Sci. 72:1847–1853. 11 SAS威. Release 6.11 for Windows 95. 1995. SAS Inst., Inc., Cary, NC. 12 Vinson, W. E., and A. E. Freeman. 1972. Selection of Holstein bulls for future use in artificial insemination. J. Dairy Sci. 55:1621–1630. 13 White, J. M. 1975. Choosing and sampling dairy sires: future needs for better sampling methods. J. Dairy Sci. 58:1086–1093.
Journal of Dairy Science Vol. 83, No. 6, 2000
YOUNG = young sires selected on pedigree for 3.7% FCM.
ABSTRACT Three Holstein lines, were compared, based on different methods of sire selection, for genetic change in 3.7% FCM, fat yield, days open, and predicted body weight after calving. The three lines were 1) evaluated sires selected only for 3.7% FCM (milk line), 2) evaluated sires selected on an index that included 3.7% FCM and type traits (index line), and 3) young bulls selected on pedigree for 3.7% FCM (young line). Cows from these lines were born in 1971 through 1993 in five experimental herds owned by the State Farm Division of North Carolina Department of Agriculture. Breeding values of cows in each line computed with a repeatability model were averaged by and regressed on birth year to estimate genetic change. Genetic gains in 3.7% FCM were 81 kg/yr for the milk line, 61 kg/yr for the line selected on index, and 68 kg/yr for the young sire line. Estimates of genetic gain in fat yield were 2.99, 2.16, and 2.54 kg/ yr in the three lines, respectively. Genetic gains in 3.7% FCM and fat yield in the milk line were significantly different from the index and young sire lines, but the index and young sire lines were not significantly different. Estimates of genetic change in days open were 0.71, 0.57, and 0.63 d/yr in the milk, index, and young sire lines, respectively. These estimates were not significantly different. Average breeding values for body weight decreased for births from 1971 to 1981 then rapidly increased for later births in all lines. (Key words: genetic change, index, evaluated sires, young bulls) Abbreviation key: BWC= predicted body weight after calving, DO = days open, INDEX = evaluated sires selected on index for 3.7% FCM and type traits, MEFAT = 2X, 305-d, mature equivalent fat yield, MILK = evaluated sires selected only for 3.7% FCM,
Received August 9, 1999. Accepted January 17, 2000. Corresponding author: J. M. Abdallah; e-mail: jmabdall@unity. ncsu.edu. 1 A contribution from Regional Research Projects S-251 and S-284. 2000 J Dairy Sci 83:1359–1363
INTRODUCTION The main goal for dairy cattle selection programs is to maximize genetic progress and profitability of production. The key to genetic progress is the use of the best sires (10). Milk traits are sex-limited, thus selection of sires is based on pedigree information or progeny testing (2). In a population with positive genetic trends, younger animals should be on average superior to older animals, and, therefore, rapid genetic gain may be achieved by selecting and sampling young bulls (9). AI organizations choose young bulls based on pedigree, then test them by progeny performance to identify favorable mendelian sampling before extensive use in AI. Many dairy farmers do not participate in young sire sampling programs (13). It seems that dairy producers feel confident in using older evaluated bulls but remain convinced that increased use of young bulls will result in reduction of genetic gain (5). Loyd and Hargrove (4) found that daughters of progeny tested sires had higher production of milk, fat and protein than daughters of young sampling sires. Vinson and Freeman (12) reported that the mean progeny performance of bulls returned to service exceeds the mean progeny performance of all bulls sampled by 146 and 5.0 kg for regressed deviated milk and milk fat yield. Others (5, 6) have found little or no differences in production between contemporary daughters of young sires and evaluated sires. Study of genetic differences among categories of service sires is needed for educational programs and decision making by the industry and dairy producers (10). The objective of this study was to compare the genetic change of cows for 3.7% FCM, fat yield, days open (DO), and predicted body weight after calving (BWC) by using three methods of sire selection. MATERIALS AND METHODS In 1970 a selection experiment was started in five experimental dairy herds owned by the State Farm Division of North Carolina Department of Agriculture.
1359
1360
ABDALLAH AND MCDANIEL
Sires were chosen from among active AI bulls by one of three methods. These methods were 1. Selection among progeny-tested bulls only for 3.7% FCM (MILK line). 2. Selection among progeny-tested bulls based on a final index that included 3.7% FCM, udder index, and feet and legs index (INDEX line). Udder index, feet and legs index, and the final index were computed as follows from standardized sire summaries: Udder index = [4 × udder support + 3 × teat placement + 2 × udder depth + fore attachment + 0.5 × rear udder width + 0.5 × rear udder height]/11. Feet and legs index = [3 × foot angle - absolute value of rear legs]/3. Final index = [3 × (udder index) + (feet and legs index) + 8 × (3.7% FCM)]/12. 3. Selection among young bulls based on their pedigree indexes for 3.7% FCM (YOUNG line). The pedigree index of a young bull was calculated as I = 0.5 (PTAs + PTAd), where PTAs and PTAd are the predicted genetic merits of the sire and the dam of the bull for 3.7% FCM. In the late summer of each year, 6 to 12 young bulls were selected from approximately 150 bulls to be sampled. Young bulls were chosen from three cooperating AI organizations with a few exceptions. Preference was given to the highest-ranking bulls that would have their first semen available between September 1 and December 31. No more than three sons of any one sire were used in a single year. In later years preference was given to young bulls that were not closely related to those used in previous years. Approximately 90 units of semen was obtained from each bull with smaller amounts in the year when more young bulls were chosen. Each year, 3 to 10 bulls were chosen from each of the MILK and INDEX lines from all active AI bulls in the US with repeatabilities of 60% or more. The number chosen depended on several criteria. These included the reliability of PTA, how much the top bulls differed in their progeny tests, the cost and availability of semen, and the kinships among the bulls in a line. The rule was to purchase 150 units from each bull. For the highestranking bulls with limited supply or those with high semen costs, as few as 50 units were purchased. Cows were bred to bulls from their own line, and mating was managed such that inbreeding in any calf did not exceed 6.25%. Cows that were not pregnant after four inseminations with evaluated sire semen were bred with young sire semen. Traits of interest were 3.7% FCM, mature equivalent fat yield (MEFAT), DO, and BWC. Yield traits were Journal of Dairy Science Vol. 83, No. 6, 2000
Table 1. Number of animals and records, means, and standard deviations for FCM, fat yield, days open and predicted body weight after calving by sire line. Sire Line 1
2
Milk
Index3
Young4
Sires(n) Cows (n) Records (n)
117 1542 3760
142 1588 3958
233 1534 3806
Mean, kg SD
9170 1794
9074 1723
8884 1791
Mean, kg SD
337 70
336 67
330 70
Mean, d SD
153 83
152 84
150 83
Mean, kg SD
592 85
594 84
596 88
Trait
3.7% FCM MEFAT DO BWC
1 MEFAT = 2X, 305-d, mature equivalent fat yield; DO = days open; and BWC = predicted body weight after calving. 2 Evaluated sires selected for 3.7% FCM only. 3 Evaluated sires selected by index for 3.7% FCM and type traits. 4 Young bulls selected by pedigree for 3.7% FCM.
on 2X, 305-day, mature equivalent basis. Days open (DO), as measured from calving to conception, were recorded to the nearest 10 d. If DO were greater than 310 d, it was set to 310 d. If a cow was sold or removed for other reasons without being pregnant, number of DO was calculated from calving day to the day it was sold or removed. The BWC of cows calving before 1988 were predicted from a heart girth-weight table developed by Kendrick and Parker (3). After 1988, mostly scale weights were used. Heart girths and actual body weights were measured at or before the first test day after calving. Record number, means, and standard deviations of traits are in Table 1. Breeding values of cows for each trait were obtained in a previous study (1) by using MTDFREML programs and single-trait repeatability model that included relationships and genetic groups. Breeding values were averaged by birth year of cows within sire line. The annual genetic change in a trait was estimated as the linear regression of average breeding values on year of birth. The differences in annual genetic changes among lines were tested using the CONTRAST statement in general linear model procedure (PROC GLM) of SAS software (11). RESULTS AND DISCUSSION Average breeding values for 3.7% FCM of cows born in 1971 through 1993 in the MILK, INDEX, and
1361
SIRE SELECTION METHODS AND GENETIC CHANGE
Figure 1. Average breeding values of 3.7% FCM by sire line. ♦, MILK = evaluated sires selected for 3.7% FCM; •, INDEX = evaluated sires selected by index for 3.7% FCM and type traits; and ▲, YOUNG = young bulls selected by pedigree for 3.7% FCM.
YOUNG lines are in Figure 1. Breeding values increased in all three lines. Average breeding values of cows in the MILK line were higher than in the INDEX and YOUNG lines over most of the period. Only small differences existed in breeding values of YOUNG and INDEX lines. Annual genetic gains were 80.9, 60.9, and 67.7 kg/yr for MILK, INDEX, and YOUNG lines, respectively (Table 2). The differences were highly significant (P < 0.01) between the MILK and each of the INDEX and YOUNG lines. Genetic changes in INDEX and YOUNG lines did not differ significantly (P > 0.10). Average breeding values for MEFAT are in Figure 2. Cows in the MILK line had higher average genetic merit than cows in the INDEX and YOUNG lines in 1978
Figure 2. Average breeding values of fat yield by sire line. ♦, MILK = evaluated sires selected for 3.7% FCM; •, INDEX = evaluated sires selected by index for 3.7% FCM and type traits; and ▲, YOUNG = young bulls selected by pedigree for 3.7% FCM.
through 1988. The INDEX line was inferior to the other two lines after 1987. Annual genetic gain in fat yield (Table 2) was 2.99 kg/yr for the MILK line, 2.16 kg/yr for the INDEX line, and 2.54 kg/yr for the YOUNG line. The MILK line was different from INDEX and YOUNG lines (P < 0.05), but the INDEX and YOUNG lines were not different (P > 0.10). Breeding values for DO (Figure 3) slowly increased for births before 1989 but increased rapidly thereafter for all lines. Yearly genetic change in DO did not differ between lines. Estimates of annual genetic change were 0.71, 0.57, and 0.63 d/yr in the MILK, INDEX, and YOUNG lines, respectively (Table 2). Higher breeding
Table 2. Estimates of annual genetic change in FCM, fat yield, days open, and predicted body weight after calving by sire line. Sire Line 1
Index2
Milk
4
Trait
3.7% FCM, kg MEFAT, kg DO, d BWC, kg 1971–1981 1981–1993
Annual genetic change b
SE
Annual genetic change a
Young3
SE
Annual genetic change a
SE
80.9 2.99b 0.71
3.8 0.13 0.10
60.9 2.16a 0.57
5.0 0.20 0.07
67.7 2.54a 0.63
7.2 0.29 0.20
–3.58a,b 3.46b
0.31 0.13
–2.92a 3.27b
0.30 0.34
–4.30b 2.54a
0.45 0.16
a,b
Values in a row with different superscripts are significantly different (P<0.05). Evaluated sires selected for 3.7% FCM. 2 Evaluated sires selected by index for 3.7% FCM and type traits. 3 Young bulls selected by pedigree for 3.7% FCM. 4 MEFAT = 2X, 305-d, mature equivalent fat yield; DO = days open; and BWC = predicted body weight after calving. 1
Journal of Dairy Science Vol. 83, No. 6, 2000
1362
ABDALLAH AND MCDANIEL
Figure 3. Average breeding values of days open by sire line. ♦, MILK = evaluated sires selected for 3.7% FCM; •, INDEX = evaluated sires selected by index for 3.7% FCM and type traits; and ▲, YOUNG = young bulls selected by pedigree for 3.7% FCM.
values for DO means more days from calving to conception and, thus, lower reproductive efficiency of cows. Averages of breeding values for BWC are in Figure 4. Breeding values rapidly decreased in the three lines for births from 1971 through 1981 but increased for cows born in later years. From 1971 to 1981, breeding values in the YOUNG line decreased more than in the INDEX and MILK lines but increased less (P < 0.05) for births from 1981 to 1993 (Table 2). Annual genetic changes in Table 2 were approximated by linear regression. All traits showed a slight curvilinear relationship to year of birth. But the linear regression accounted for most of the sums of squares,
and the significant differences among lines did not change when quadratic and cubic effects of birth year were included (results not shown). Working with a subset of the data used in this study, McDaniel and Bell (6) found no significant differences among the three lines in genetic gain of milk yield. Powell and Norman (10) reported that average evaluations of young bulls were similar to those of active AI bulls, but with differential usage considered, active bulls averaged 108 kg more for predicted difference of milk. Loyd and Hargrove (4) found that daughters of progeny tested sires outproduced daughters of sampling sires by 146, 52, and 174 kg of milk yield in first, second, and third generations and by 1.2, 0.8, and 1.1 kg for fat yield. McDaniel et al. (7) compared the three lines and found that the MILK line had the highest income over feed costs followed by the YOUNG line. They (7) also found that cows in the MILK line may have had slightly more days to first service (81 d for the MILK line vs. 79 d for the INDEX and the YOUNG lines). McMahon et al. (8) reported that the average profitability of young sires was equal to average profit from evaluated bulls when young sire semen cost $5/ unit and was $36 more profitable than average evaluated sires when young sire semen cost was nil. CONCLUSIONS Cows in the MILK line attained the highest genetic gain in milk and fat yields but tended to have more DO compared to cows in the INDEX and the YOUNG lines. Cows sired by young bulls were competitive in production with cows sired by evaluated bulls selected by index for milk and type traits. Although more risk is associated with using young sires, use of the highest on pedigree merit is expected to result in substantial genetic improvement. The use of young bulls becomes more advantageous when the lower price of semen from these bulls compared to semen from evaluated bulls is considered. Sires selected on an index that included milk yield and type traits had progeny with lower production than sires selected only for milk yield. Putting emphasis on type traits could hinder genetic gain in yield traits. In all three lines, DO increased as breeding values for 3.7% FCM increased. All lines showed significant decreases in breeding values for body weight from 1971 to 1981 but large increases from 1981 to 1993. REFERENCES
Figure 4. Average breeding values of predicted body weight after calving by sire line. ♦, MILK = evaluated sires selected for 3.7% FCM; •, INDEX = evaluated sires selected by index for 3.7% FCM and type traits; and ▲, YOUNG = young bulls selected by pedigree for 3.7% FCM. Journal of Dairy Science Vol. 83, No. 6, 2000
1 Abdallah, J. M. 1998. Genetic parameters and genetic trends in North Carolina experimental dairy herds. M.S. Thesis, North Carolina State Univ., Raleigh. 2 Freeman, A. E. 1975. Choosing and sampling young bulls: theory, background, and general problems. J. Dairy Sci. 58:1063–1070.
SIRE SELECTION METHODS AND GENETIC CHANGE 3 Kendrick, J. F., and J. B. Parker. 1936. Estimating the weights of dairy cows from heart girth measurements. BDIW 695. United States Department of Agriculture, Washington, D.C. 4 Loyd, B. M., and G. L. Hargrove. 1988. First lactation comparisons between daughters of summarized sires and sampling sires in artificial insemination. J. Dairy Sci. 71(Suppl. 1):235. (Abstr.). 5 McCraw, R. L., K. R. Butcher, and B. T. McDaniel. 1980. Progeny tested sires compared with pedigree selected young sires. J. Dairy Sci. 63:1342–1350. 6 McDaniel, B. T., and W. E. Bell. 1989. Genetic gain in milk yield from three systems of sire selection. J. Dairy Sci. 72(Suppl. 1):68. (Abstr.). 7 McDaniel, B. T., R. E. Collins, L. E. Pitzer, R. H. Lee, W. E. Bell, C. H. Brown, P. L. Williams, J. C. Wilk, and C. V. Tart, Jr. 1995. Comparison of three sire selection systems in Holsteins. ANS Report No. 247. Dep. Animal Science, North Carolina State Univ., Raleigh.
1363
8 McMahon, R. T., R. W. Blake, C. R. Shumway, and M. A. Tomaszewski. 1985. Selection of young and proven Holstein artificial insemination sires to maximize profits from milk. J. Dairy Sci. 68:2303–2308. 9 Nizamani, A. H., and P. J. Berger. 1996. Estimates of genetic trend for yield traits of the registered Jersey population. J. Dairy Sci. 79:487–494. 10 Powell, R. L., and H. D. Norman. 1989. Genetic differences among categories of service sires. J. Dairy Sci. 72:1847–1853. 11 SAS威. Release 6.11 for Windows 95. 1995. SAS Inst., Inc., Cary, NC. 12 Vinson, W. E., and A. E. Freeman. 1972. Selection of Holstein bulls for future use in artificial insemination. J. Dairy Sci. 55:1621–1630. 13 White, J. M. 1975. Choosing and sampling dairy sires: future needs for better sampling methods. J. Dairy Sci. 58:1086–1093.
Journal of Dairy Science Vol. 83, No. 6, 2000