Biology Reference
In-Depth Information
each generation. Selected animals from each hatch served as the breeding
population for the next generation. The sire of every chick was identified
(one male per cage) and within dam families were identified on the basis
of egg color, shape, and size. Morphological characteristics of eggs are very
constant within females but very different between females. Furthermore, each
chick was individually banded at hatching. Wing bands were used to identify
the animals throughout their lives. In the SR lines, the measure of SR behavior
(measured in arbitrary units) is the difference between the distance run on the
treadmill minus the “far” time. It was shown in a preliminary experiment that
the distance run and the far time were equal when the animals were tested in
the treadmill apparatus without stimulus animals, resulting in a score of 0.
In order to ensure that selection for TI did not influence SR behavior or
vice versa, all animals were tested for both TI and SR behavior and the scores
for each trait were subjected to regression analysis. In the lines selected for
high levels of social reinstatement behavior the SR score was calculated on
the regression graph with the duration of TI. The selection criterion was the
deviation of each individual data point from the regression line. In the tonic
immobility lines (high fear), the measure of TI was the time taken by the birds
to right themselves. This value was regressed on the chick's score for social
reinstatement behavior. Again, the selection criterion was the deviation of
each data point from the regression line. This transformation reduced the
risk of co-selection for the two traits due to correlation between social rein-
statement and tonic immobility. Breeding values were calculated taking into
account not only the individual values but also those of full siblings ( Mills and
Faure, 1991 ). The two control lines were bred from one son chosen at random
from each father and one daughter from each mother. Lines were selected as
above until the 16th generation. However, from generation 17 onward, “skip a
generation” selection was employed. In the nonselected generation, breeding
animals were chosen at random and used for the control lines.
SELECTION RESULTS
Selective breeding for high and low social reinstatement and high and low
tonic immobility produced significant changes in behavior. In the first gener-
ation, the mean (
SD) value for distance run on the treadmill was 14.4 meters
6
(
8.9). After 20 generations of divergent selection for high or low levels of
social reinstatement behavior, values were 59.0 (
6
6
27.3) meters run in the
high SR line, 7.4 (
8.9)
meters run in the low SR line ( Figure 8.2 ). The distribution of values for the
distance run were completely different between the first and 20th generation
of the high SR and low SR lines ( Figure 8.3 ). From the 10th generation of
selection onward, the standard deviations of the high SR and low SR lines
ceased to overlap. This would seem to indicate that
6
11.4) meters run in the control SR line and 4.0 (
6
the high social
Search WWH ::




Custom Search