Íslenskar landbúnaðarrannsóknir - 01.03.1977, Síða 38

36 ÍSLENZKAR LANDBÚNABARRANNSÓKNIR
Table 5 . Increase in the multiple correlation coeffisient by including
weight of ewe and date of roating in the model in addition to
the age of the ewe.
The number of lambs born is more strongly
affected by ewe body weight in January than
in October. There are several reasons for this.
The January weight is taken just after the
mating season and therefore measures the
condition of the ewe at mating which is
known to influence the number of lambs
born. The autumn weight of the ewe is also
affected by the number of lambs reared the
year before. This environmental effect is
negatively correlated with number of lambs
born, but is masked in January.
The effect of the date of mating on the
number of lambs born is small. The linear
effect alone is not significant (P>0.05)
Mating takes place during a short period in
late December and early January. The ewes
that are mated outside this period are the
ones which are inseminated before the normal
mating season starts and ewes that remrn
to service. Flushing is commonly practised
by farmers in Iceland as a means of in-
creasing prolificacy, but has not been started
when the insemination takes place. Many
are also of the opinion that the prolificacy
of the ewes is negatively affected by insemina-
tion and evidence of this is found in these
data (Jónmundsson, 1975b). This will give
a curvelinear effect of the date of mating on
the number of lambs at birth.
Johansson and Hansson (1943) found
in Sweden a curvelinear effect of the date of
Table 6 . Regression coeffisients, linear (b) and second degree (bi and b2),
of number of lambs at birth on the weight of the ewe and on the
date of the mating.
Curvelinear regression Linear regression
t>l t>2 t)
Body weight in October Body weight in January Date of mating 0,04901±0,00589 -0,00032+0,00005 0,01132±0,00066 0,07493±0,00599 -0,00051±0,00005 0,01452±0,00068 0,02682±0,00383 -0,00023±0,00003 -0,00014±0,00054