Íslenskar landbúnaðarrannsóknir - 01.03.1970, Page 72

70 ÍSLENZKAR LANDBÚNAÐARRANNSÓKNIR
TABLE 32
Least squares estimates of effects of genotype of 2-year old ewes
on number of lambs per ewe lambing
Genotype of ewe No. o£ ewes Estimated deviations
Among all White vs. nonwhite Among white Among nonwhite
Ai— 1935 - 0.078 - 0.004
AiA2 7 -0.043 0.036
A1A4 4 -0.107 0.002
A1A5 115 -0.109 - 0.034
Wliite 2061 -0.070
A2A4 26 - 0.005 0.132
a2- 22 0.076 -0.083
AoAg 84 0.064 0.001
AsAg 15 0.136 0.020
A4A5 31 0.000 -0.083
A5A5 244 0.065 0.014
Nonwhite 422 0.070
Sum 2483
Overall mean (unadjusted) . . S.E. (within subclasses) 1.224 0.364 1.224 0.368 1.194 0.349 1.370 0.465
ing. Ewes not lambing were exclnded from
this analysis.
The comparisons were first made among
10 genotypes of ewes and then the geno-
types were split into two groups, white and
nonwhite, which were compared, and fin-
ally comparisons were made among white
and among nonwhite genotypes respective-
ly. The least squares estimates from these
analyses are shown in table 32 and the
analyses ot' variance in table 33.
The sums of squares for main effects
given in tables 33, 35 and 37 refer to the
adjusted sums of squares. The between
farm-years sums of squares have thus been
adjusted for genotypes ancl the between
genotypes surns of squares adjusted for
farm-years.
The error sums of squares have been
obtained as the within subclass sums of
squares.
The rnain conclusions to be drawn from
tables 32 and 33 are that there is a sub-
stantial and significant effect of colour
genotype of ewe on number of lambs per
ewe lambing. This effect can be ascribed
completely to the presence or absence of
the A^-allele. There is no indication of a
heterosis effect, as known white hetero-
zygotes do not cliffer from the A^-group
which must include a considerable nurnber
of homozygous white ewes. There is no de-
tectable difference between nonwhite geno-
types either, wlien compared to each other.
This is of particular interest in connection
with the grey ewes which would be ex-