Young Sires for Herd Improvement

By Bennet Cassell, Extension Dairy Scientist, Genetics and Management, Virginia Tech.
calendar icon 22 September 2006
clock icon 9 minute read
Table of Contents

Research on young sires versus proven bulls
How well does pedigree selection work?
Should you use young sires?


Farmers can choose between herd bulls, AI proven sires, and young sires in progeny testing programs when breeding cows in their herds. Herd bulls as a group are significantly poorer genetically than AI young sires, though an occasional full brother to an AI young sire may be available for natural service. Regardless of which choice is made, daughters of the three groups of bulls find their way into herds and may compete head to head for the resources available. This guideline is about the choice between AI proven bulls and AI young sires. Which group offers the best opportunity for genetic progress? Figure 1 shows the impact of genetic trend on the choice between AI proven and young sires. Our purpose is to make genetic improvement within a herd rather than within a breed. Figure 1 shows all the young sires being born in 1990 and all the proven bulls being born in 1985. In reality, AI proven bulls would have been born over several years, but none more recent than 5 or 6 years earlier, because of the time required for sampling. AI young sires available through semen distribution channels would be as young as about 15 months and as old as 5 to 6 years of age. The trend line increases over time as genetic progress is made for whatever trait has been selected for improvement. The trend line is set by the average merit of bulls born in any year. Bulls much better and worse than average are born all the time. The average bull born in 1990 is marked by the arrow for "Current PT bulls" where "PT" means progeny test. Such bulls are better than the average of all progeny tested bulls born 5 years earlier by the increment labeled "genetic trend."

Of all the bulls progeny tested from the 1985 birth year, only a select few are returned to service as proven sires. This group might be as good, on average, as the point indicated by "Active AI bulls." If so, the average young sire born in 1990 would be genetically equal to the average proven bull born 5 years earlier and returned to service following sampling. On the other hand, a farmer could be more selective in use of proven bulls and only use a selected group with average merit indicated by "Top bulls." In this case, the AI proven bulls would be genetically superior to AI young sires born 5 years later. Again, it is important to remember that the decision to use a sire through AI is independent of his age. Daughters of average AI proven bulls, top AI bulls, and young sires would be competing against each other in the herd at the same time.

Figure 1. Relationship of genetic merit of young sires to proven bulls when genetic trend is strong and positive.

Figure 1 illustrates the problem faced by dairy producers in trying to choose between proven bulls and young sires. Among the questions to be answered are:

  • How good are AI young sires compared to proven bulls?
  • Can AI young sires be used in place of proven bulls if only the best young sires are used?
  • If a farmer is careful to use only the best proven bulls, are AI young sires too expensive to use even with a lower semen price?
  • Just how well can genetic ability of AI young sires be estimated from pedigree information?

    Research on young sires versus proven bulls

    At the time a farmer must make a decision about using young sires, only pedigree information is available to evaluate each youngster. Pedigree data has suffered from a bad reputation through the years, in some cases with good reason. In the past, some pedigree information was not very reliable. Bull mothers were given preferential treatment and their genetic evaluations overstated their merit. This caused the pedigree evaluations of their sons to be too high and ultimate progeny test information was frequently disappointing to producers who milked the daughters. While the incentive to bias a bull mother's evaluation (a high price for a son in AI) may still exist, some things have changed for the better. We use better methods for genetic evaluations than were used several years ago. Many bull studs are choosing "bull mothers" from first calf heifers before preferential treatment can bias evaluations very much. The bottom line is that AI young sires are generally evaluated more accurately for production then they were several years ago.

    The critical question to be answered to know how much to use young sires is "How good are they compared to proven AI sires available at the same time?" A criterion for comparing the two groups is needed such as type, first lactation milk yield, longevity, and so forth. A recently completed study at Virginia Tech compared young bulls to proven AI sires on the basis of PTA for MFP$. The study examined young bulls that were sampled in each of the past 10 years, the active AI sires that were available during the same year, and where the young bulls RANKED compared to the proven bulls. Comparisons were based on the ultimate progeny tests of the young sires rather than on the pedigree information available at the time they were sampled. An average proven bull would have a rank of 50. He would be better than 50% of the active AI bulls available. If the average merit of young sires ranked them at the 50th percentile, farmers would make as much genetic progress using young sires as by using average active AI bulls.

    Table 1 shows that young sires ranked between the 55th and 65th percentile during the early 1980s. However, their rank has declined in recent years. This does not mean that young sires declined in merit, but that the proven bull population improved more rapidly due to larger sampling programs. The number of young bulls sampled each year went up. At the same time, the number of proven bulls on the market stayed the same. The proven bulls available in the latter 1980s were better because they survived more stringent culling to reach active AI status.

    Table 1. Number of Holstein young sires sampled, number of proven bulls available per year, and relative genetic merit of young sires during the 1980s
    Year Number of young Number of proven
    1980 734 405
    1981 868 461
    1982 856 380
    1983 922 377
    1984 905 357
    1985 910 378
    1986 881 328
    1987 1031 367
    1988 1117 363
    1989 1111 367

    How well does pedigree selection work?

    Farmers want an opportunity to select the bulls they use, even when young sires are involved. How successful can producers be at picking future progeny test graduates from pedigree information available at the time young sires are used? The study mentioned earlier used the same records on which Table 1 is based to see how useful Parent Average or PA was in selecting the best of the young sires in sampling. Parent average is the average PTA of the sire and the dam of each young bull for MFP$. Only young bulls sampled from 1987 to 1990 were used so that PA's would contain the same information as farmers had to go on when that group of youngsters was sampled. Bulls born earlier had milking daughters when the first animal model genetic evaluations were run (in July 1989) and that progeny test information affected PA! Young bulls were compared to other young bulls sampled at the same time, simulating the kinds of choices farmers would make when selecting young bulls for use in a herd. Young sires were divided into 10 groups based on PA for MFP$ and the percentage of bulls in each group which were returned to active AI service or which had high TPI values was determined.

    Results are in Table 2.

    Table 2. Graduation rates and percent with high TPI values for young sires ranked on their Parent Average (PA) MFP$
    Percentile rank for PA MFP$
    Percent returned to active AI service Percent above the 85th percentile of TPI
    90 28 23
    80 26 28
    70 21 20
    60 18 15
    50 17 14
    40 15 16
    30 13 15
    20 7 7
    10 7 9
    0 6 3

    Young bulls in the group labeled "90th percentile" for PA MFP$ would have been in the top 10% of all youngsters sampled for the birth years 1987 through 1990. Bulls in the bottom group had the lowest 10% of PA's for MFP$. Table 2 shows that pedigree selection works. Four times as many bulls in the top group were returned to service as graduated from any of the bottom 3 groups. Higher ranks for PA MFP$ also produced many more of the top TPI bulls than bulls whose pedigree information ranked them low compared to other bulls sampled at the same time. All AI young sires are not created equal. Additional selection on pedigree information from among those young bulls in stud sampling programs appears worthwhile.


    Risk is a factor in young sire use. Pedigree information is not as accurate as progeny test information. It indicates what genes a calf MIGHT have inherited from its parents, but does not reveal whether an individual calf inherited a better or worse than average sample of genes from its parents. About half of all young bulls will inherit a better than average sample of genes while the remainder will be less fortunate. The problem is that progeny test information is required to tell which young bull received the best sample of genes. The best way to avoid the risk of using too many young sires with good pedigrees but poor gene samples is to limit semen purchases to 10 to 20 units of semen on each young bull. Herds using young sires for cleanup could justify more semen per young bull because of reduced fertility among the mates. Fewer pregnancies per 10 unit batches means fewer replacements. Realistically, 10 units of semen will produce either zero, one, or (rarely) two first calf heifers in the milking string in most cases. The recommended upper limit of 20 units per young sire could be raised for large herds willing to accept more risk in order to reduce the number of young bulls used in a breeding program.

    Should you use young sires?

    Young sire semen is currently cheaper than semen from proven bulls, even proven bulls ranking well below the 50th percentile for MFP$. For most herds, young sires should be used on problem breeders and during hot weather when fertility is reduced. The current economic incentives to use young bulls make them a good choice for reasons unrelated to genetic merit. However, genetic progress is higher from using top proven bulls than from using average AI young sires. Even average young sires are as good as proven AI bulls at or below the 50th percentile. Farmers who pick AI young sires with the best pedigrees for MFP$ can use young bulls that are well above the 50th percentile for MFP$. The study cited earlier showed that the best 25% of young sires were genetically equal to proven bulls at the 75th to 85th percentile for MFP$. Unfortunately, farmers currently have no good method to rank young sires on PA MFP$. The best option would probably be to ask the semen salesman for pedigree information on young sires currently available and then to purchase small amounts of semen from the best youngsters.


    Selection is the key in deciding whether to use young sires for genetic improvement. Careful selection from available proven AI sires makes young sire use less appealing. Selection of the best pedigrees among young sires sampled by a stud(s) makes young sire use more beneficial for genetic improvement.


    Weigel, D.J., B.G. Cassell, and R.E. Pearson. 1995. Relative genetic merit and effectiveness of selection of young sires for artificial insemination. Journal of Dairy Science 78:2481
    August 1996

  • © 2000 - 2023 - Global Ag Media. All Rights Reserved | No part of this site may be reproduced without permission.