Íslenskar landbúnaðarrannsóknir - 01.03.1970, Qupperneq 89

COLOUR INHERITANCE IN ICELANDIC SHEEP 87
of p and q given in table 48. A comparison
of the expected frequencies with the ob-
served frequencies in tabe 46 gives ^2i =
0.058 (0.90 > P > 0.80).
The significant difference between p and
q shows that the distribution of the homo-
zygotes on the two score classes differs signi-
ficantly from that of the heterozygotes. The
nonsignificant shows furthermore that
the estimates of p and q suffice to explain
the observed difference between mating
groups with respect to the score for tan
colour.
Of the homozygotes, 98.6 per cent fall
into score class a, and only 1.4 per cent
into score class b, while only 75.6 per cent
of the heterozygotes fall into score class a
and 24.4 per cent into score class b.
This result shows an interesting aspect
of the action of the Aj-allele. In its hetero-
zygous state it inhibits the formation of
black and brown pigment, while the homo-
zygous condition reduces the production of
tan pigment. This finding is comparable
to the results obtained by Dry (1924), that
the blue complexion among white Wensley-
dale sheep was far more common among
heterozygotes than among homozygotes.
I. RELATION BETWENN SCORE
FOR GREY COLOUR AND HOMO-
ZYGOSITY FOR THE A2-ALLELE
Score for grey or greybrown colour was
available for 344 lambs out of parents with
known genotypes at the A-locus in the
present study. The scores 1—7 were used
for the tlifferent intensity of colour of grey
lambs, as described earlier, p. 13, score 1
being used for the lightest grey and score
7 for the darkest grey at birth. In the fol-
lowing, only 3 score classes will be used,
class a containing scores 1 and 2, class b
for scores 3, 4 and 5 and class c for scores
6 and 7, the main reason for adding to-
gether adjacent classes being low numbers
in several of the original score classes. In
the following, grey and greybrown lambs
liave been taken together and will be re-
ferred to as grey.
The lambs with score for grey colour can
be divided into 4 groups according to the
type of mating from which they came. In
mating group 1 all grey lambs were homo-
zygous grey, while in groups 2, 3 and 4 the
expected ratio between homozygous and
heterozygous grey lambs was 1:1, 1:2, ancl
0:1, respectively.
The distribution of the progeny from the
4 mating groups on score classes when
compared under the assumption that mat-
ing group had no effect on colour score
distribution, resulted in ^26 = 151.194
(P < 0.001). This value of the x2 shows
without any doubt that there is a very
marked difference between the mating
groups with respect to distribution of grev
progeny on score classes. The difference is
obvious without the test, as seen from the
observed frequencies in table 49.
For the estimation of the difference be-
tween homozygotes and heterozygoes with
respect to score for grey colour, 4 para-
meters were postulated as a measure of the
proportion of homozygotes and hetero-
zygotes in the 3 core classes. Parameters pt
and p2 gave the proportion of homozygotes
in score classes a and b respectively and
1 — px — p2 tlie proportion of liomozygotes
in score class c, wliile qx, q2 and 1— q3 —q2
referred to the corresponding proportions
for heterozygotes. Table 50 shows the ex-
pected frequencies of progeny in each score
class for each mating group as a function
of the parameters.
The maximum likelihood estimates of
the 4 parameters were obtained in the
same way as described for the score for tan
colour, p. 85. Table 51 shows the estimates
of the parameters together with their stand-
ard errors.
As seen from table 51, the homozygous