Changes to evaluation system (December 2015)
Genetic variance for
Jersey type traits
By
Jan Wright
The
genetic SD of several conformation traits will decrease by about 5% in Dec 2015
for Jerseys because an array overflow had affected the iterative herd variance
adjustments for that breed only. The program bug had caused almost no change in
PTA rankings, only the small changes in SD. Formulas for net merit
will not be affected. New and previous SD for each
trait are compared below.
|
Changes
in Jersey genetic standard deviation for December 2015 |
||
|
Trait:
|
New
Genetic SD |
Previous
Genetic SD |
|
Final
score |
0.91
|
0.91
|
|
Stature
|
1.28
|
1.40
|
|
Strength
|
0.87
|
0.94
|
|
Dairy
Form |
0.93
|
0.98
|
|
Foot
angle |
0.70
|
0.76
|
|
Rear
leg (side view) |
0.63
|
0.68
|
|
Rump
angle |
1.05
|
1.11
|
|
Rump
width |
0.76
|
0.82
|
|
Fore
udder attachment |
1.09
|
1.16
|
|
Rear
udder height |
1.06
|
1.13
|
|
Rear
udder width |
0.98
|
1.04
|
|
Udder
depth |
1.43
|
1.51
|
|
Udder
cleft |
0.77
|
0.83
|
|
Front
teat placement |
1.04
|
1.12
|
|
Teat
length |
0.93
|
1.02
|
Unknown parent group
definitions
By
Paul VanRaden, Jay Megonigal, Curt Van Tassell, and
Mel Tooker
The
automated system to assign unknown parent groups was revised to improve
stability and convergence with data updates. The new group definitions are
applied to yield traits, productive life (PL), somatic cell score (SCS),
daughter pregnancy rate (DPR), heifer conception rate (HCR), and cow conception
rate (CCR). Since Dec 2014 the animal model includes pedigrees for young
animals, but too few cows with data were included in the most recent groups, so
those groups will be combined. Genomic PTAs should change little because most
parent averages were recomputed as needed, but parent averages for young,
non-genotyped animals with missing sires or dams are the most affected.
Additional breeds from
Interbull
By
Paul VanRaden and Gary Fok
Bulls
with breed codes MO (Montbeliard), NO (Normande), SM (Simmental), and Fleckvieh
(FL) will be on the Simmental scale in the Interbull multitrait across-country evaluation (MACE) instead of the
Holstein scale. As a result, about 7,000 foreign MO, NO, SM, and FL bulls will
be converted by MACE onto the U.S. Holstein base. PTAs for bulls of those
breeds will include the expected heterosis when mated
to Holstein cows and be comparable to Holstein evaluations, as previously. The
MACE system does not exchange fertility PTAs for those breeds, and conformation
traits on U.S. scale are not available, so Holstein birth year averages are
substituted for the missing traits. Reliability improves for the foreign bulls
with U.S. daughters because of more pedigree information and additional
ancestor PTAs from MACE for the additional breeds.
Editing changes for
sire conception rate
By
Duane Norman
An edit
previously used in the calculation of sire conception rate (SCR) was if the
expected calving date (last breeding date plus 280 days) was 21 or more days
greater than the actual subsequent calving date, the breeding was deleted
because it was assumed the date was recorded wrong or the service was to a cow
already pregnant. However, the average heat cycle is 21 days but individual
heats can vary by a few days. According
to Ray Nebel (personal communication), “the industry uses 18 to 24 as the
normal estrous interval.”
Therefore CDCB is changing this edit from 21 to 17 days to allow for the
variation expected in the heat cycle.
The gestation period previously used in editing was 280 days; this is
being modified to use the gestation period appropriate for the specific breed
addressed, i.e., changed to the breed average in the literature (AY, 282; BS,
288; GU, 286; HO, 279; JE, 280; MS, 281; and WW, 280 days). These breed
averages were obtained from national data from 1999 through 2006 (Norman et
al., J. Dairy Sci. 92(5):2259–2269, 2009).
Another
edit previously programmed in SCR was “A herd needed a conception rate between
10 and 90% over the entire 4 year period or the inseminations from the herd
were not used” (Kuhn and Hutchison, J. Dairy Sci. 91(6):2481–2269, 2008 and
Kuhn et al., J. Dairy Sci. 91(7):2823–2835, 2008). This range seems extreme so
we are lowering this on the top to 75% to make it more in line with other
procedures being used.
The
number of age categories for SCR will be reduced in the Jersey breed to make
them resemble more closely a smooth biological curve.