Íslenskar landbúnaðarrannsóknir - 01.03.1977, Qupperneq 29

SOURCES OF VARIATION IN WEIGHTS OF LAMBS 27
flock was 1.26. This indicates that there is
a certain degree of assortative mating with
regard to relationship. The reason for this is
probably that the farmer tries to avoid inbre-
eding and when using old rams that also have
daughters in the flock one will get such a
relationship structure. The structure is not so
marked that it should affect the heritability
to any great extent.
CORRELATIONS
The genetic and phenotypic correlations are
estimated from covariance components from
a covariance analysis by the same structure
as variance analysis for estimating herita-
bilities. The sampling error of the genetic
correlations were estimated by the formula of
Robertson (1959).
The correlations are shown in table 13.
HallgrÍmsson (1971) estimated the gene-
tic correlation between weaning weight and
carcass weight as 0.8—0.9 with tendency
towards higher correlations for ram lambs
than for ewe lambs as found here. The reason
for this may be that there is often a greater
time-lag between weighing and slaughter
with ewe lambs than ram lambs. This can
also be a selection bias or a real sex difference.
The genetic correlation between weaning
weight and dressing percentage is estimated
negative. Gjedrem (1970) has estimated this
correlation as —0.18 and Bradford and
Spurlock (1972) as —0.57. This correla-
tion may be a measurement of a negative cor-
relation between devélopment and absolute
weight. This correlation will affect the selec-
tion pratice because the economic character
is carcass weight.
For testing the evenmal effect of selection
on the correlation one has calculated
the covariance between weaning weight of
ewe lambs and carcass weight and dressing
percentage of ram lambs for sires with five