Íslenskar landbúnaðarrannsóknir - 01.03.1977, Síða 27

SOURCES OF VARIATION IN WEIGHTS OF LAMBS 25
type of birth and rearing x sex groups are
shown in table 11.
After correction one will have similar
means and standard deviations in the sub-
groups. The additive constants for type af
birth and rearing are shown to give more
similar standard deviations in the subgroups
than the multiplicative one and it is there-
fore concluded that the additive correction
gives the best correction for the effect of
type of birth and rearing. The greatest varia-
tion is found in the 2X1 group of lambs
but these lambs will naturally have the
greatest variation because it is not known
when the partner is lost.
There was found to be some variation in
the sex difference between districts. The sex
difference was shown to be greater in districts
with polled sheep than in districts with
horned sheep.
GENETIC PARAMETERS
HERITABILITY
The heritability was estimated by half-sib
correlation. The variance components were
estimated from a hierarchial analysis of
variance as follows:
The material was restricted to sire groups
with 10 or more progeny within each flock.
Then one has eliminated most of the cross
classification in the data. It is not common
that rams have progeny in more than one
flock except for rams used in artificial inse-
mination but they have very few progeny in
any one floclc. A total of 42888 lambs were
included in the analysis for all three traits.
Sampling error of the heritability was esti-
matmated with the formula by Dickerson
(1959) modified by Becker (1967).
The heritability estimates are given in
table 12. There are only a few older estimates
of heritability of these traits in Icelandic
sheep. Hallgrímsson (1971) estimated her-
iability for weaning weight 0.44 for singles
and 0.18 for twins and for the same trait
JÓNMUNDSSON (1971) found 0.29 for ewe
lambs and 0.19 for ram lambs. Thorsteins-
son and Björnsson (1971) have estimated
heritability of carcass weight of twin ram
lambs as 0.30. Many estimates of heritability
of weaning weight of lambs are found in
the literature and they show very wide varia-
tion (Bowman 1966). In Norway Eikje
(1974a) has estimated it 0.14 on a huge
Sources of variation
E (MS)
Between flocks
ae + ^Gd + k50I + ksab
Between sires within flock
2 2 2
a + k2o, + k3o
e d s
Between dams within siré
2 2
o + kjo,
e d
2
O
e
Within dams