GENETIC IMPROVEMENT OF DAIRY CATTLE
DAIRY HERD IMPROVEMENT
1985
DICKINSON, F.N.
VOLUME: NCDHIP HANDBOOK
The main goal of a dairy producer's genetic improvement program
should be to produce replacement cows with the greatest possible
genetic capability for making a profit. Fulfilling this goal
requires strong, healthy cows that produce high levels of milk of
desirable composition. Also, these cows must be able to stand the
stress of high production through many lactations with a minimum of
special treatment. This goal can be attained by combining the best
of the genetic material presently in the herd for economically
important traits with the best available germplasm from outside.
For most herds, this means 1) using Dairy Herd Improvement
Association (DHIA) records to optimize the impact of voluntary
culling and minimize inbreeding and 2) purchasing genetically
superior semen of artificial insemination (AI) bulls.
The ideal cow genotype varies from herd to herd because traits have
varying economic importance in different herds. Predicted
Difference dollars (PD$) and Cow Index dollars (CI$), PD$ cheese
(also known as cheese yield dollars), or indexes utilizing those
values, probably are the most important methods for ranking bulls
and cows because total income from milk depends on the amount of
milk, the milk fat percentage, and in some markets, the percentage
of protein or solids-not-fat. Therefore, PD$, CI$ and PD$ cheese
should be understood and the importance of PD$ and CI$ percentiles
appreciated fully as tools for ranking bulls and cows genetically.
Understanding Genetic Evaluations
Genetic Bases. All genetic evaluations are expressed as deviations
from a genetic base because genetic evaluation procedures estimate
only differences among bulls and among cows. The purpose of a
genetic base is to provide a reference point so that genetic
evaluations are understandable and useful.
Three genetic bases are possible: 1) fixed, 2) moving and 3)
stepwise. The U.S. dairy industry has adopted a stepwise genetic
base for national genetic evaluations. This stepwise genetic base
was established in 1974 with genetic evaluations labeled PD74 and
CI74. The genetic base for each trait was changed in January 1984
to a 1982 genetic base with genetic evaluations now labeled PD82
and CI82.
Interpreting Genetic Evaluations. Genetic merit of dairy cattle for
yield and physical type traits in the United States is represented
by PD's and CI's. These evaluations should be used to rank bulls
and cows. No matter what the overall magnitude of PD's and CI's
(which is a function of the genetic base), dairy producers cannot
do any better than to use the highest ranking bulls and cows
available. Genetic evaluations under the 1982 genetic base (or any
other genetic base) are not estimates of the amount of genetic
improvement that can be obtained from use of a specific bull or cow
for breeding in a specific herd. Rather, they are estimates of how
much the genetic merit of a specific cow or bull differs from the
genetic base. The amount of genetic improvement that a specific
animal can be expected to contribute to a specific herd depends on
the relation between that animal's evaluation and the average
genetic merit of the cows in the herd.
Estimating a bull's transmitting ability from a few daughters can
be misleading. Most quantitative traits, including yield traits,
vary considerably among a bull's daughters. The best bulls have
some low-producing daughters, and the worst bulls have some
high-producing daughters. Variation in level of daughter yield is
due to the effects of genetic sampling and environment.
Role of Repeatability in Selecting Bulls. Repeatability is a
measure of the reliability of a genetic evaluation. Repeatability
should be used to determine how heavily to use each bull in a
single herd, but it should not be used to decide whether or not to
use a bull. In most cases, bulls with high Repeatabilities can be
used more heavily in a herd with confidence that their PD's are
close to their true transmitting abilities. However, individual
bulls with low Repeatabilities should be used less heavily in each
herd because their PD's may vary more from their true transmitting
abilities. Groups of low-Repeatability bulls can be used with
confidence that their average PD is close to their average true
transmitting ability.
Managing a Breeding Program
Breeding goals for the herd should be established with full
consideration given to their economic impact. These goals can vary
widely from herd to herd, and short-term goals may be different
from long-term goals. However, herd owners should adhere to goals
once they are established. Vacillating from one set of herd
breeding goals to another will decrease a herd's genetic progress
and also result in a loss of potential economic gain.
Relative Emphasis on Various Traits.
The relative emphasis to be placed on different traits depends on
their relative economic importance for fulfilling herd breeding
goals, their degree of genetic control (heritability), their
genetic relationships (genetic correlations) and the accuracy of
information available. Approximate heritabilities of traits for
which genetic evaluations are available for large numbers of
animals are in Table 1. Phenotypic and genetic correlations of some
traits of common interest with first-lactation milk yield are in
Table 2. For almost all herds, yield traits should receive the
greatest emphasis because research has shown that yield per cow
almost always has the highest relation to herd profitability.
Component percentages almost never should receive Primary selection
emphasis in spite of their high heritabilities. Instead, selection
for components should be directed at increasing yields of fat,
protein or solids-not-fat. Component percentages have negative
genetic correlations with yield traits and smaller genetic and
phenotypic variances. Thus, little economic progress would be made
through direct selection for component percentages. Traits for
which relatively little genetic information is available probably
should receive little emphasis because selection is likely to be
ineffective even though a dairy producer may consider those traits
important.
The emphasis to be placed on type traits depends on the dairy
producer's breeding goals. Dairy producers who expect to sell a
significant amount of breeding stock on a regular basis usually
place more emphasis on type than they would if income were
exclusively from the sale of milk. Relative progress for yield and
type expected from different amounts of emphasis on each is shown
in Table 3. For most herds and most breeds, the optimum ratio is
about 3:1 or more in favor of yield over type. Even though primary
selection emphasis should be on yield traits, a bull's strengths
for specific uniform functional type traits (UFTT's) can be taken
advantage of through selective matings. Information on a bull's
genetic evaluations for UFTT's can be obtained from breed
associations and from AI organizations for AI bulls.
TABLE 1. Approximate heritabilities of traits for
which genetic evaluations may be available.(1)
Approximate
Trait heritability
Yield
Mature-equivalent milk 0.30
Mature-equivalent fat 0.25
Milk (deviation from herdmates) 0.25
Fat (deviation from herdmates) 0.25
Protein 0.25
Solids-not-fat (SNF) 0.25
Fat percentage 0.50
Protein percentage 0.50
SNF percentage 0.50
Linear type traits
Final type score 0.30
Stature 0.40
Chest and body (strength) 0.20
Dairy character 0.20
Foot angle 0.10
Rear legs (side view) 0.15
Rear legs (rear view) 0.10
Pelvic angle (rump side view) 0.20
Rump width 0.25
Fore udder attachment 0.20
Rear udder height 0.15
Rear udder width 0.15
Udder depth 0.25
Suspensory ligament 0.15
Teat placement (rear view) 0.20
Miscellaneous
Calving ease 0.05
(1) Source: Bath, D.L., et al. 1985. Dairy Cattle.
Principals, Practices, Problems, Profits. pp.
94-95.
Selection pressure does not need to be exerted for calving ease.
After a group of bulls has been selected for the herd's breeding
program, those bulls with desirable evaluations for calving ease
can be mated to heifers that have never calved to minimize calving
problems.
Genetic-Economic Indexes. Genetic-economic indexes (PD$, CI$ and
PD$ cheese) are released in the United States as a standard part of
the genetic evaluations of bulls and cows. They express
transmitting abilities as dollar values based on the average
national price for milk volume and components. These indexes rank
bulls and cows reasonably well for most parts of the country.
Percentile rankings based on PD$, CI$ and PD$ cheese are included
in genetic evaluations for bulls and cows. These percentiles
emphasize the importance of PD$, CI$ and PD$ cheese as tools to
rank bulls and cows and helped users of genetic evaluations adjust
to the genetic base change in January 1984. The three percentile
rankings have different definitions. Percentile rankings for PD$
are based on the active AI bulls of each breed; that is, a bull
with a percentile ranking of 80 percent has a higher PD$ than 80
percent of the active AI bulls for that breed. Percentile rankings
for CI$ are based on the cows of each breed that pass the
first-stage screening for the Elite Cow Index (see Fact Sheet H-4).
Percentile rankings for PD$ cheese are based on each breed's active
Al bulls that have protein Summaries.
Other indexes have been developed to rank bulls on a dollar
net-return basis. These indexes can be calculated for individual
bulls from PD$, semen cost and other information, and account for
a variety of economic factors. Two of the early indexes rank bulls
on dollars of net return per unit of semen: 1) dollars of net
return = (.3PD$ + $2.50) - (dollar cost per unit of semen), which
was developed in Wisconsin and 2) dollars of net return = PD$
-6(dollar cost per unit of semen), which was developed in New York.
TABLE 2. Phenotypic and genetic correlations of some
traits of common interest with first-lactation milk
yield. (1)
Phenotypic Genetic
Trait correlation correlation
Yield
Fat yield (first lactation) 0.85 0.70
Solids-not-fat(SNF) yield 0.85 0.90
Protein yield 0.85 0.90
Fat percentage (1st lact) -0.35 -0.35
SNF percentage -0.30 -0.25
Protein percentage -0.35 -0.30
Lifetime milk yield 0.35 0.80
Linear type traits
Final score 0.29 0.00
Stature 0.11 -0.01
Strength 0.12 0.07
Dairy character 0.50 0.68
Foot angle 0.00 -0.24
Rear legs(side view) 0.02 0.14
Pelvic angle 0.04 0.19
Thurl width 0.10 -0.11
Fore udder attachment -0.09 -0.47
Rear udder height 0.12 -0.13
Rear udder width 0.16 0.09
Udder depth -0.27 -0.64
Suspensory ligament 0.14 0.12
Front teat placement 0.02 -0.12
Miscellaneous
Length of productive life 0.25 0.75
Mastitis -0.05 0.10
Breeding problems 0.05 0.00
Milking speed 0.05 0.00
Excitability 0.00 0.05
Feeding speed 0.15 0.45
1 Source: Bath, D.L., et al. 1985. Dairy Cattle:
Principles, Practices, Problems, Profits. p. 95.
Indexes have been developed that rank dairy cattle on a combination
of PD$ (or CI$) and final score for type. These indexes are called
Production-Type (or Cow Performance) Indexes for all breeds except
Holstein, for which the index is called a Total Performance Index.
Production-type indexes for bulls of each breed are available in
breed association publications.
A computerized bull selection system called Net Present Value (PV$)
was developed in Texas to provide profit-maximizing rankings of
individual bulls tailored to each herd's genetic improvement goals.
(See Fact Sheet H-7.) The PV$ for each bull represents an optimal
weighting of genetic, reproductive and economic information to rank
bulls by net profit of their daughters in current dollars. This
system allows dairy producers to exercise selection policies with
their goals for milk income and type. Milk income can be the only
selection goal, or milk income and type can be weighted according
to the dairy producer's preference. Genetic improvement goals for
Holsteins can be based on planning horizons of one generation or an
infinite number of generations.
A computerized bull selection procedure called MAXBULL was
developed in Virginia. (See Fact Sheet H-S.) The MAXBULL system is
different from Net Present Value in that it permits dairy producers
to establish breeding goals for important traits for the average of
bulls to be used in the breeding program while also setting minimum
and maximum limits as appropriate for individual traits of critical
importance. In other words, MAXBULL maximizes the overall merit of
the group of bulls selected to meet herd breeding goals rather than
selecting bulls on how they meet a set of criteria individually.
The principle on which MAXBULL is operated is complementary sire
selection. Complementary sire selection does not eliminate a bull
that is strong for many traits just because he is weak for one
trait; rather it tends to minimize the impact of his weakness by
selecting other bulls for the group that are strong for that trait.
This approach tends to maximize the average genetic merit of the
group of bulls selected at the average semen price that the dairy
producer is willing to pay.
Dairy producers are well advised to make use of the best
genetic-economic indexes available to them to maximize economic
returns from their genetic improvement program.
Sources of Germplasm. The most reliable source of superior
germplasm for a herd's breeding program is bulls that have been
evaluated accurately for a large number of traits including yield,
type conformation and calving ease. Generally, this information
will be obtained from bulls available through AI: 1) those with
Sire Summaries and 2) those young sires that are being progeny
tested. Dairy producers can be confident that most bulls with high
Repeatabilities have evaluations that are close to their
transmitting abilities. However, dairy producers should take
advantage of the superior germplasm of progeny-test bulls. Because
of the competition among breeders and among AI organizations to
produce genetically superior bulls, most young bulls in
progeny-test programs have high pedigree merit (that is, high
expected transmitting ability) for economically important traits.
Research has shown that, for at least a decade, most young bulls in
progeny-test programs have had higher transmitting abilities than
active AI bulls on the average at any given point in time.1
Therefore, virtually all dairy producers should consider using
groups of young progeny-test bulls for some portion of their herd.
Dairy producers can take advantage of young progeny-test bulls in
active AI service through the computerized sire selection programs
(MAXBULL and Net Present Value) by setting low minimum
Repeatabilities. Semen from young progeny-test bulls frequently can
be obtained at a low cost.
TABLE 3. Expected relative progress for Predicted
Differences (PD's) for dollars (PD$) and type (PDT)
from varying amounts of selection emphasis for the
top 10 percent of Jersey sires.(1)
Emphasis Progress for
(PD$:PDT) PD$ PDT
0:1 +20 +.69
1:1 +67 +.50
2:1 +80 +.33
3:1 +84 +.23
6:1 +87 +.12
9:1 +87 +.08
1:0 +88 -.01
(1) Source: Norman, H.D., et al. 1979. J. Dairy Sci.
62:1914.
Dairy producers can take advantage of another source of potentially
superior germplasm through participation in bull-proving
syndicates. Bulls being proven by syndicates usually have
estimated transmitting abilities that are competitive with bulls
being proven by the major AI organizations.
The use of herd bulls or pasture bulls should be limited to
providing a mechanism for getting cows or heifers pregnant under
the most difficult management conditions. Such bulls should be used
only as a last resort when it is impossible to manage an effective
genetic improvement program. The unknown genetic merit of those
bulls dictates that they never should be used heavily in a herd if
an alternative breeding procedure is available. If a dairy producer
must resort to the use of herd or pasture bulls, those bulls
selected should have the highest possible pedigree merit for yield
and type traits important to the producer. Each bull should be used
to breed only a small percentage of the herd.
Embryo transfer and various forms of genetic engineering have great
potential for genetic improvement of dairy cattle. In the past,
much of the potential of embryo transfer for genetic improvement of
yield has not been realized because of heavy selection emphasis on
nonyield traits. However, much greater attention now is being
directed towards cows that rank high on the Elite Cow Index list,
especially when donor cows are sought for the international embryo
market. High cost still is a significant obstacle to widespread use
of embryo transfer for genetic improvement, but as techniques and
reliability improve, costs will undoubtedly decrease. A significant
breakthrough in genetic engineering, such as a practical method of
sex control, should increase the impact of embryo transfer as a
tool for genetic improvement.
Planning the Herd's Genetic Improvement Program. Dairy producers
can maximize the profit from their herd's genetic improvement
program by a two-step planning procedure. The first and by far most
important step is selecting the group of bulls that will be used in
the herd. The second step, which should yield added benefits, is
the planning of individual matings between each bull and cow using
DHIA records and other information.
Bulls should be selected for use in the herd with the goal of
improving the herd as a whole, not for improving individual cows.
This can be done most effectively by selecting a group of bulls
according to a set of average minimum and maximum standards for the
group (complementary sire selection) rather than by selecting
individual bulls according to a set of minimum or maximum
standards. Such selection is more successful than individual bull
selection because if bulls are evaluated against a set of standards
individually, a bull is eliminated if he falls outside the
acceptable range for any single standard. The best bull for one
trait might be eliminated because he is just outside the acceptable
range for another trait. However, when a group of bulls is selected
against a set of average standards for the group (as in
complementary sire selection), bulls just outside the acceptable
range for one or even a few traits can be selected for breeding if
they are sufficiently superior for other traits as long as the
average minimum or maximum standards are met for the group of bulls
as a whole. Use of this group selection concept based on average
minimum or maximum merit almost certainly will result in selection
of a group of bulls that will have higher genetic and economic
merit for most traits than will selection of individual bulls.
For example, Table 4 shows data for three hypothetical bulls for PD
milk, PD fat and PD type. Naturally, not all bulls can be selected
for breeding. At first glance, bull B might he eliminated easily on
the basis of low PD's for fat and type, and bull C might be
eliminated on the basis of low PD for milk. Bull A might be
selected as a reasonable compromise for all three traits. However,
the average merits of bulls B and C are + 1,100 pounds for PD milk,
+.05 for PD fat percentage and +.25 for PD type, all higher than
bull A's merit for those traits. Therefore, equal use of bulls B
and C as a result of complementary sire selection would lead to
greater genetic improvement than using just bull A.
TABLE 4. Example Predicted Differences (PD's) to
demonstrate advantages of selecting groups of bulls
by average standards.
PD PD fat PD
Bull milk percentage type
A +800 0 0
B +2,000 -.15 -.75
C +200 +.25 +1.25
After a group of bulls has been selected for breeding to the herd,
choosing a specific bull to mate to each cow for the purpose of
corrective mating can yield an added bonus to the herd's genetic
improvement program. Records from DHIA programs, type information
from breed association programs and results of corrective mating
evaluations should all be helpful. Making planned or corrective
matings in a herd should boost the herd's rate of genetic progress
somewhat, but the group of bulls selected for breeding in each herd
largely will determine the herd's genetic progress.
Dairy producers should make new bull selections and semen purchases
each time new Sire Summary information is available (every 6
months). Many dairy producers who keep a semen tank could increase
future genetic progress in their herds by culling the genetically
inferior semen and replacing it with genetically superior semen.
Ideally, dairy producers should buy semen only when they can
specify the bulls they want in the amounts they want and at prices
that will make the investment in their genetic improvement programs
a profitable one.
Comparing Bulls From Different Countries
Every country has a different genetic base for each trait in each
breed. Bases can be determined by a variety of different methods
and can refer to different points in time. For example, the next
change in the stepwise genetic base in the United States will be an
estimate of the amount of change in transmitting ability that will
have occurred since the PD82 base was established. However, other
counties may establish genetic bases according to the merit of a
particular group of bulls sampled in AI without regard for the
genetic contribution those bulls made to the cow population.
Interpreting changes in bases is made more difficult because some
countries measure changes in terms of transmitting ability and
others in terms of breeding value. Transmitting ability is one-half
breeding value and is the method for expressing genetic merit (PD
and CI) in the United States.
A basic formula for converting genetic evaluations between
countries is:
E1 EGM =