235
Training stimuli
Approach, leave and control measures
Results. Preoperative measures of the latency to approach and leave after the
train-trill in the training and test regimes showed no significant differences.
The high score for controls was due to one bird that otherwise performed
normally (Table I). The latency for approach, postoperatively, increased for
both Iesioned groups with respect to the sham-controls (P < 0.01). The post-
erior group stayed longer by the dish than the sham-control group (P < 0.05).
The normal rhythm of feeding pecks could be interrupted by a bird raising
its head before continuing to feed. The number of such feeding bouts did not
differ between groups. ‘Punishment’ continued to be delivered outside the
1.5 min analysis period around a stimulus. There was no difference in the num-
ber of pre- and postoperative ‘punishments’ received by the three groups.
Discussion. The consistent proportion of feeding bouts suggests that the op-
eration produced no major side effects on hunger or the coordination of the
feeding pattern. No distraction from feeding was observed. With no change in
the number of ‘punishments’ delivered nor of fearfulness shown in the pres-
ence of novelty one may conclude that motivation levels were stable and
there was no gross interference with the retention of the task. The increased
approach latency of the Iesioned groups is discussed with search behaviour
below.
Overt behaviour
After the train-trill there was a rapid onset of feeding that lasted for about
30 s (Fig. 4, sixth 5-s period). The incidence of feeding rapidly decreased over
the next 10 s. During these IOs movement (M) sharply increased. The birds
left the food area. Movement peaked between 45 and 55 s and then approach-
ed prestimulus levels. After feeding the probability of search behaviour (S)
rapidly increased (Fig. 5, sixth to seventh 5-s period). This phase was follow-
ed by an increasing frequency of low tension (8—10) that was in turn super-
ceded by a peak of high tension behavioural components (10—12).
After operation there were two points where the changeover between the
dominant behavioural components was different between the groups (feed—S,
LT—HT). Firstly,the posterior group fed longer than the anterior group (Fig.
4, P < 0.03) and then searched more than the anterior or control groups
(Fig. 5, P< 0.05). This was reflected by the lower movement peak, with res-
pect to prestimulus levels, shown by the posterior group (P < 0.04). Second-
ly, at the end of the behaviour sequence the anterior group showed more low
and less high tension than the posterior and control groups (Fig. 5, P < 0.016).
Behaviour during extinction
There were no significant differences between groups in the length of time
spent by the dish. There were no consistent tendencies across 20 extinction