Íslenskar landbúnaðarrannsóknir - 01.09.1976, Side 57

STUDIES ON AUTUMN WEIGHT OF ICELANDIC LAMBS 55
Source of variation DF MS E(MS)
Between sires S-1 MS 2 2 « + ka
s s e s
Within sires 2(n. -1) MS 2 O
i=r e
In the above, s is number of sires, n.
number progeny within the .th sire, 2 is
1 0
e
the variance component within sires, 2 the
O
S
variance component between sires and k the
number of progeny per sire, calculated as
follows:
i
The heritability estimate is then obtained
as follow:
Here it is assumed that the relationship
between half-sibs is 0.25. In case of data of
the kind described here, the average relation-
ship, r can be calculated according to the
D
following formula given by Osborne
(1957):
r n (d+l)4-2 ,
8 4 (nd -r 1) ,
where d is number of dams per sire and n is
number of progeny per dam.
The sampling variance of the heritability
estimate has been calculated from the follo-
wing formula given by Robertson (1959):
where T is total number of animals in the
analysis.
b. Maternal effect on autumn wetght
of lambs.
Gjedrem (1967) has described a method
for estimating the heritability of maternal
effect on autumn weight of lambs. The ma-
terial is grouped on the ram’s sire and the
variance components calculated in the fol-
lowing way:
°SD = 4 + í 'S = 0ovMS, where
2 __ the genetic component for maternal
M effect,
2 __ the sire component for additive
S genetic effect, estimated in the
between sire analysis of variance,
and
Cov = the genetic covariance between
MS maternal effect and additive genetic
effect.
No estimate of Cov MS is available in the
literature for sheep, but weaning weight of
beef cattle, Kock and Clark (1955) and
Hohenboken and Brinks (1971) found a
negative genetic correlation between direct
and maternal effect.