1 The USDA-DHIA Buck Evaluations are national genetic
evaluations for
dairy bucks. These evaluations rank bucks in order of
genetic merit for
milk yield for each breed. This information is of
great importance to
breeders because one of the most important factors
affecting genetic
progress is the genetic superiority of animals
selected to be parents.
To select genetically superior animals as parents, a
breeder must have
accurate information on the genetic merit of bucks and
does. This
information should help breeders make informed
decisions about their
breeding programs, resulting in significant genetic
improvement for
yield in their herds.
2 The USDA-DHIA Buck Evaluations represent an
important first step
toward eventually providing dairy goat breeders with a
comprehensive
array of genetic information. They are the result of
several years of
research at the Animal Improvement Programs
Laboratory, USDA, in
Beltsville, Maryland, with cooperation from
researchers at the
University of California at Davis, the University of
Illinois at Urbana,
and the University of Maryland at College Park.
Information, advise,
and support from the American Dairy Goat Association
(ADGA) has been
invaluable. However, these evaluations are not the
final answer, by
any means. Research will continue to improve the
scope, usefulness, and
accuracy of genetic evaluations for dairy goats.
3 Buck evaluations are computed so that they are best
linear unbiased
predictions of genetic value. Procedures that result
in evaluations
with these statistical properties are used widely for
dairy cattle. For
dairy goats, all lactation records for each doe are
used and
relationships among bucks are incorporated. Lactation
data are received
from all dairy record processing centers in the
National Cooperative
Dairy Herd Improvement Program (NCDHIP), and pedigree
data are received
from ADGA.
4 Procedures to edit incoming lactation records
include checks on
yield and kidding information. Yield for only the
first 305 days of
lactation is considered. Records with more than 305
days in milk are
excluded because a standard lactation length of 305
days is imposed.
(Records with more than 305 days in milk usually have
a corresponding
305-day record reported.) Terminated records have to
have at least 15
days in milk and records in progress at least 80 days
in milk. Actual
milk yield per day has to be at least 2 lb but no more
than 30 lb at
the beginning of the lactation, with the upper limit
declining to 20
lb as number of days in milk increases. Fat percentage
has to be at
least 2.0but no more than 7.5 These restrictions were
established
to remove unusual or highly variable records. Records
from does with a
kidding date before 1976 or a birth date before 1973
are excluded
because of lack of data and less reliable
identification information in
earlier years.
5 Lactation records also are excluded if they have the
following
identification conditions: (1) sire not registered,
(2) sire
identification missing, and (3) breed other than
Alpine, LaMancha,
Nubian, Saanen, and Toggenburg. These records are
excluded because
little interest has been shown in evaluating bucks
that are not
registered or that do not belong to a major dairy goat
breed. Does do
not have to be registered for their records to be used
in computing
genetic evaluations.
6 For January 1983 evaluations, 58,562 records
representing 43,913
does sired by 11,670 bucks in five breeds were
available for
computations after editing. Distribution of records,
does, and bucks
by breed are in Table 1. The does were located in
3,781 herds, but
numbers of herds by breed is not reported because of
the large number
of multibreed herds.
7 Average yields for milk and fat as of January 1,
1983, are in Table
2 by breed. Actual yields are based on lactations with
275 to 305 days
in milk. This reflects the actual yield of a doe
successfully
completing the major portion of a 305-day lactation.
Standardized yields
are estimates of what a doe would produce if she had
been 36 months old
at kidding, had kidded in late winter, and had 305
days in milk or a
completed record of less than 305 days. They are based
on the entire
data set used for buck evaluations and are calculated
by averaging the
averages for each doe's records rather than by
averaging all records.
Standardized yields are slightly higher than actual
yields for all
breeds but Nubian. The value for Nubians may be lower
because their
shorter average lactation length is considered when
standardizing
yield. Therefore, Nubian records included when
determining actual
yield were more selected than for other breeds.
8 All lactation records with fewer than 305 days in
milk that are
used in computing genetic evaluations are projected to
305 days, not
just lactation records coded incomplete. This
procedure was used for
the first time in computing the USDA-DHIA January 1983
Sire Summaries
and Cow Indexes for dairy bulls and cows. Research
with cow lactation
records showed that this procedure should increase
accuracy of genetic
evaluations by removing environmental variation caused
by different
lactation lengths. Thus, dairy goat breeders were able
to benefit from
the most recent research find ings on genetic
evaluation of another
species.
9 Predicted Difference (PD) is the measure usually
used in the United
States to express genetic transmitting ability
(one-half breeding
value) for a trait of male dairy animals. Each animal
has a separate
transmitting ability for each trait. For example, a
buck can have an
especially high PD for milk yield and only a medium or
even a low PD
for fat yield.
10 All PD's relate to a genetic base for each trait
and are a way to
rank bucks. A genetic base can be established in many
different ways,
and the way in which it is established affects the
magnitude of
genetic evaluations. For each dairy goat breed,
genetic bases for milk
and fat evaluations were obtained by setting to zero
the average PD of
sires of all does with lactation records used in the
January 1983
evaluations. Each individual PD was weighted by the
number of daughters
for that buck. Future evaluations will be comparable
with the January
1983 evaluations because the genetic base will be the
same.
11 Goat breeders should not expect PD's to estimate
the exact amount
of improvement or decline for a trait that will result
from using a
particular buck in their herds. This amount is
dependent primarily on
the genetic merit of bucks that were used previously.
The PD should be
thought of as the amount by which the buck's average
off spring would
be superior or inferior to offspring of a buck with PD
=0. Individual
offspring may be better or worse than expected.
However, on the
average, the higher a buck's PD or percentile ranking,
the better his
offspring.
12 An example of the first USDA-DHIA Buck Evaluations
and a detailed
explanation of the information in them follow.
Registration number: Buck's registration number in the
ADGA or American
Goat Society herdbook. The registration number is
preceded by two
digits required for data processing in NCDHIP.
Name: First 30 characters of the buck's name from the
ADGA herdbook.
Herds: Number of herds represented by lactation
records from NCDHIP
that passed edits.
Daus: Number of daughters with lactation records from
NCDHIP that
passed edits.
Lacts: Number of daughter lactation records that
passed edits.
Rpt: Repeatability (or reliability) of a buck's
genetic evaluation
expressed as a percentage. A high Repeatability means
that a breeder can
have confidence that a buck's PD's are accurate
estimates of his true
(genetic) transmitting ability for the traits shown.
With low
Repeatabilities, PD's may vary quite a bit from true
transmitting
ability. For a buck's PD to be included in USDA-DHIA
Buck Evaluations,
the PD must have a Repeatability of at least 15 For
these
evaluations, Repeatability is based only on amount of
daughter
information. Information from a buck's other relatives
contributes to
his evaluation but is not reflected in his
Repeatability at present.
PD milk (lb): Predicted Difference for milk yield in
pounds. The PD is
a buck's estimated transmitting ability (or one-half
breeding value).
The PD milk (lb) is expressed (as are all genetic
evaluations) as a
deviation from a genetic base.
tile milk: Percentile ranking for PD milk (lb) based
on evaluations of
all bucks with Repeatabilities 15
PD fat (lb): Predicted Difference for fat yield in
pounds, again the
estimated transmitting ability expressed as a
deviation from a genetic
base.
tile fat: Percentile ranking for PD fat (lb) (see tile
milk).
PD fat (): Predicted Difference for fat percentage
(see PD milk (lb).
The PD fat () is computed from PD milk (lb) and PD fat
(lb).
