The Control of Two Antagonistic Locomotor Activities in a Predatory Insect

Aeschna cyanea larvae are ambush hunters which, however, may readily pursue an escaping prey. Target tracking at first consists of lively swimming movements, and later of rapidly decelerating walking and creeping movements. If a dummy prey is removed after having been presented for 40 to 80 seconds, the insect 1) freezes and simultaneously looks in the direction where the prey stimulus has just disappeared. 2) Then it steps backwards repeatedly, and finally 3) it changes its body orientation by a series of (clockwise and/or counterclockwise) turning movements. Experiments in which the duration of the presentation of a moving dummy prey was extended from 5 to 80 seconds showed that the longer the insect spends tracking, the more probable and the more intense its subsequent backing-turning responses, and the sooner these responses occur after the disappearance of the prey. The occurrence and intensity of the backing-turning pattern seems to be more closely related to the insect's preceding tracking time than to the simultaneously covered tracking distance, which depends on the tracking speed. Intense backing-turning responses were not only primed by an extended presentation of the dummy, but also by a series of discontinuous brief presentations. Short tracking spells therefore exert a cumulative after-effect upon backing-turning. When the prey stimulus had been presented for longer than 80 seconds, backing-turning started to compete with fixation and approach reactions towards the prey. Under these circumstances, the insect backed and turned away from the dummy prey, as if the latter were no longer located in its visual field. The variation of both 1) the internal and 2) the external variables which activate predatory behaviour influenced the balance between tracking and backing-turning. i) During the first two days of food deprivation, the insects showed a progressive increase in the intensity of their tracking behaviour as well as in the probability and strength of their subsequent backing-turning responses. At the same time, the onset of the first backing response was delayed. 2) During a prolonged period when two artificial prey stimuli were presented alternately, the two differing in their degree of conspicuousness, more backing-turning responses occurred when the less conspicuous prey stimulus was present. Thus, not only a total disappearance, but also a partial reduction in the intensity of the prey stimulus, favoured the appearance of backing-turning. On the other hand, during a prolonged presentation of the prey stimulus, a sudden increase in its intensity while the insect was initiating the first backing response, was not followed up by the immediate resumption of predatory approach behaviour. During a sequence of brief presentations of the dummy prey, the insects increased their relative amount of tracking after they had displayed backing-turning for the first time. This suggests that the performance of the stereotyped pattern facilitates the subsequent resumption of tracking. The backing-turning pattern appeared only during or after the presentation of a prey stimulus and it was always preceded by approach or tracking behaviour orientated towards the prey. It seems therefore that only predatory locomotion - i.e. approach reactions orientated towards a static prey and target tracking towards a moving prey- prime backing-turning, itself a form of locomotion. A model is proposed according to which the performance of tracking exerts a negative feedback effect upon itself and at the same time progressively lowers the threshold of the mechanism controlling backing-turning. Therefore, after a prolonged pursuit, backing-turning starts to interfere with tracking. As long as the prey stimulus remains present, fixation and approach reactions alternate with backing and turning away from the prey. If the prey, however, is completely removed, positive appetitive behaviour towards the prey can no longer compete with backing-turning, and the stereotyped pattern can appear in its full intensity. On the other hand, tracking itself seems to be facilitated by the previous performance of backing-turning. From a functional point of view, the stereotyped pattern of locomotion may be conceived as a device 1) to stop the larva's unsuccessful attempts to reach a rapidly escaping prey, and 2) to diminish the probability that the insect may re-encounter this prey after its momentary disappearance.