Genetic Influences on the Behavior of Chickens Associated with Welfare and Productivity - Genetics and the Behavior of Domestic Animals - page 329

individuals that otherwise would have been low in social status over the
entire test period if they had been kept in stable organizations.
Mating Behavior of Males
Two bidirectional selection studies for frequency of matings by males
( Tindell and Arze, 1965; Wood-Gush, 1960 ) indicated that large differences
could be produced readily. However, a long-term bidirectional study, in
which the foundation population was maintained as a control, by Siegel and
his colleagues ( Dunnington and Siegel, 1983 ) indicated complex results.
It became evident early in the study that results were not symmetrical.
Although the initial response to selection in the high line was near zero in
the first few generations, responses then began and accumulated until at least
20 generations had elapsed. Realized heritability was 0.18 when calculated
over 23 generations. In the low line very different results were obtained; a
realized heritability of 0.32 was obtained through generation 11 ( Siegel,
1972 ), but thereafter a very large additional response appeared to be associ-
ated with little change in effective selection differentials ( Dunnington and
Siegel, 1983 ). Some males failed to mate at all during the test situation. The
percentage of non-maters decreased in the high line, was unchanged in the
control, and increased dramatically in the low line. During generations 21,
22, and 23 the incidence was 84, 89, and 60%, respectively in the low line.
A surprising lack of correlated responses occurred in Siegel's long-term
selection study and an important genotype by rearing interaction was found.
Although very large differences were present among the high, control, and
low lines for the trait of selection ( Dunnington and Siegel, 1983 ), fertility in
natural matings did not differ when males were present for extended periods
of time or when artificial insemination was practiced ( Bernon and Siegel,
1981 ). Females of the lines in generations 16 through 23 were tested for mat-
ing behavior, and when analyses were carried out within generations no con-
sistent line differences were found ( Dunnington and Siegel, 1983 ).
The question of whether males of the high-, control-, and low-mating fre-
quency lines would respond similarly if reared with females as compared to
separate sex rearing after 6 weeks of age was answered by Cook and Siegel
(1974) . A large genotype by rearing environment interaction was found.
Rearing males of the unselected and low lines in heterosexual flocks
increased the frequency of maters by about 20% whereas essentially all
males of the high line mated regardless of how they were raised.
Appetite
From a controlled five-generation study of single-trait selection on body-
weight gain, food consumption, and food conversion ratio ( Pym and
Nicholls, 1979 ) it was learned that all three selected lines responded with