Visual Responses of Mosquitoes Artificially Rendered Flightless

The work of Kennedy (1940) showed that flying mosquitoes (Aëdes aegypti) orientate themselves inaccurately but consistently away from the source of a horizontal beam of artificial light and also that when presented with several moving black stripes they tend to face one or-other of them. This characteristic negative phototaxis of mosquitoes during daytime has either been observed or suggested by numerous previous workers, though not with the same experimental clarity. It is also known that mosquitoes of many species, if not all, tend to leave daytime shelters at dusk and move towards comparatively brightly illuminated exits. There is thus an apparent positive phototaxis in the failing light of dusk. The studies reported in this paper were made in an endeavour to determine more exactly the mechanism of orientation of mosquitoes to light and darkness and to observe experimentally if there was a change of direction of taxis at dusk. A large series of observations made with flying mosquitoes by quite different methods will be reported in another paper.

The basic method in all experiments reported here was to render the mosquito flightless so that its movement towards light or darkness could be easily followed, and its track recorded. The fact that an insect which normally flies is forced to respond to stimuli entirely by pedal locomotion may not be exactly natural, but it ‘enables us to analyse behaviour from a new type of data, always keeping in mind the nature of the experimental conditions. How far the behaviour of walking mosquitoes corresponds to or differs from the natural behaviour of flying ones is a matter for further investigation.

Two species of mosquitoes, Culex (Culex) molestus Forskål and Anopheles maculi-pennis atroparvus van Thiel, were used, cultures of which have been maintained in this laboratory. The specimens of Culex molestus were sugar-fed and most of the specimens used had laid the first and only batch of autogenous eggs. Those of Anopheles maculipennis were in various gonotrophic stages, unfed, freshly fed on blood, semi-gravid and fully-gravid.

* The author desires to place on record his gratitude to Prof. P. A. Buxton, F.R.S., and Dr Kenneth Mellanby, O.B.E., for kindly advice and encouragement in many ways. To Miss Barbara M. Gilchrist special thanks are due for the readiness with which she made available large numbers of Culex molestus in the appropriate physiological condition whenever required for these and other studies.

The mosquitoes were rendered flightless in most of the experiments by removing the wings with clean cuts with needles or fine scissors across the wing expanse at a little distance from the base so that the base was invariably preserved at least up to the region of the humeral cross-vein. In some critical experiments, instead of removing the wings the latter were either stuck together with a minute quantity of shellac placed at their tips or were tied together lightly with a piece of hair. Mosquitoes were lightly etherized to facilitate these operations and experiments were carried out at least an hour later, by which time the animals had completely recovered. Experiments have been carried out both in artificial and in natural light; the details of the experimental conditions are given in individual cases.

No attempt was made to precondition the mosquitoes, but in some of the experiments with artificial light it was noted that specimens which had been kept for an hour or two in darkness seemed to respond more quickly (i.e. appeared to be more sensitive than those kept in daylight); such animals were kept covered with black paper for a short period prior to experimentation. No change in the kind of response resulted from this treatment ; subsequent responses were, however, quicker.

The apparatus used was as follows : A glass stage, 20 cm. square, was set up on a metal frame and an electric lamp, 60 W., was placed about 16 cm. below it. The top of the glass stage was covered with a sheet of white paper so that when viewed from above it appeared fairly evenly illuminated. A glass cylinder, 7 cm. tall and 7 cm. diameter, lined on the inside with thin white cartridge paper, was inverted over the stage. The whole apparatus was so placed in the darkened laboratory that little light from any other source influenced the behaviour. Vertical strips of black matt paper of varying widths were pasted inside the white wall of the cylinder. The ‘wingless’ mosquito was then placed on the stage and the cylinder rapidly inverted over it in such a way that the insect was approximately at the centre of the arena. The roof of the cylinder was faintly marked with blue pencil with four equidistant diametrical lines to facilitate placing the cylinder at certain definite angles in relation to the axis of the insect and the position of the black bands. In a typical response the ‘shock ‘of handling the mosquito was sufficient to keep it quiet on the stage for a few seconds so that the manipulation of the cylinder presented no difficulty.

In the first series of experiments a vertical black stripe exactly covered a quarter (i.e. 90 °) of the inner surface of the cylinder. The ‘wingless ‘mosquito was put on the stage by tapping it off the container in which it was kept or by dropping it from the tip of a pencil or a finger upon which it was previously picked up. In a typical satisfactory trial, the mosquito would start moving towards the wall after about 10 sec. During this interval the cylinder was inverted over it, centred and so adjusted that one edge of the black stripe subtended a definite angle from the axis of the head towards the left or the right eye. Experiments were carried out at angles of o (i.e. directly in front), 45 and 90 ° to one side. Tests were also carried out with the specimen directly facing the centre of the band or facing exactly in the opposite direction.

In a few seconds the mosquito would begin to crawl and from the very beginning it would move towards the black band. The track followed was immediately recorded on a suitably marked paper. On reaching the base of the wall the mosquito would walk up the wall, very often perpendicularly but sometimes diagonally upwards—at the top of the wall movement usually stopped.

Typical responses of a female Anopheles maculipennis are illustrated in Fig. 1. The continuous lines indicate the tracks followed when the right eye was towards the black band and the broken line when the left eye was towards it. When the edge of the dark band was directly in front (A) and when it was 45 ° to one side (B), the insects in almost all cases moved towards the black band. When the edge was at 90 ° to one side (C), the response became a little uncertain and at 135 ° (D), more so. A. maculipennis became more uncertain than Culex molestus at both these angles but the differences are not very significant. Considering both species together, about 50% of the movement was towards the black band at 90 ° and about 25% at 135 °. When the black band was directly in front (E), there was usually no marked curvature, the specimen walking more or less straight towards the central part of the band.

When the band was directly behind (D, 135 °) the specimens walked straight ahead in the majority of cases, and behaved as though they did not detect the band. But in about 25 % of cases, as stated above, movement towards the black band did take place. In such cases this response occurred invariably after the specimen had started its forward movement and it was a sharp turning-back as though the band had suddenly-come into the field of vision. In an all-white cylinder (F) (an experiment used frequently to test whether individuals had any particular bias), the mosquitoes usually went straight ahead.

While the typical responses of both Anopheles maculipennis and Culex molestus females are as described above, there were always a few specimens of each species which did not conform to any fixed rule, moving at random over the arena and not moving consistently towards or away from the black. These atypical cases represent about 10% of all specimens tested.

The total number of specimens tested exceeded a hundred C. molestus and fifty Anopheles maculipennis, though records of the tracks were not kept in some cases once the general nature of the responses had been determined.

Numerous trials were carried out to determine the narrowest black band which the mosquitoes could perceive and respond to. In these trials the band was almost always set at 45 ° towards one side or the other so that the turning response could be clearly noticed. Consistent and clear-cut responses to the bands were obtained when the width of the band was 20 mm. The width was progressively reduced during the trials and it was found that a sensitive specimen could respond to a band 5 mm. in width roughly nine times out of ten; when the width was reduced to 4 mm. there was a marked drop in the number of responses and at lesser widths the responses became quite uncertain. It appeared, therefore, that at a distance of 3 · 5 cm. a 5 mm. wide vertical stripe, subtending an angle of about 8 °, represented approximately the limit of visual acuity of the mosquitoes.

A large number of tests were carried out with both A. maculipennis and Culex molestus in a cylinder with two black vertical bands of about 20 mm. width. The bands were so placed that their two extreme edges were about 90 ° apart from one another on the circular wall. The position of the mosquito was so adjusted that it faced directly between the two bands. Fig. 2 illustrates the tracks of typical responses of Anopheles maculipennis and Culex molestus when the cylinder was adjusted so that the mosquito was at its centre. Almost identical responses were obtained when the distance between the mosquito and the bands was increased by adjusting the position of the mosquito very near the wall opposite to the bands. The response was to one of the dark bands and not to the area between them. The direction of movement of the specimen appeared to be decided almost at the start. There was no case in which the mosquito changed over to another band after having set a course towards one of them, but there were a few cases in which the mosquito normally going to one of the bands went in between the two but on reaching the extreme edge of the arena turned towards one. (Two such responses are illustrated in Fig. 2B.) When similar trials were carried out in a large cylinder of 15 cm. diameter with two black bands of 2 · 5 cm. width, the results were exactly the same and the mosquitoes could detect, and move towards, the black bands from a distance of 12 cm.

A few trials conducted with bands of different widths are illustrated in Fig. 2. In C and D the two bands, 1 · 2 and 3 · 5 cm. wide, were so placed that the centre of the intervening white area was facing the mosquito and in E and F the narrower band was directly in front while the broader band was to one side. In both cases the movement was directly towards one of the bands; sixteen responses were towards the narrow band and twenty-five towards the wider one. In one instance only did the specimen change over to the wider band after nearly reaching the narrow one.

Several specimens of both species were unilaterally blinded by complete occlusion of one eye with a light coat of shellac mixed with indian ink. Such insects, when enclosed in a cylinder with an all-white wall, performed typical circus movements turning towards the blinded side. Fig. 3 A illustrates a typical track of an Anopheles maculipennis blinded in the right eye and Fig. 3B a similar one of Culex molestus blinded in the left eye. Only one track in each case has been drawn to avoid confusion of lines.

When enclosed in a cylinder with a single dark band, 12 mm. wide, the circus movement continued until the band came into the field of vision of the functional eye; the mosquito then turned sharply towards the dark band (Fig. 3C, D).

A large cylinder of 15 cm. inner diameter made it possible to carry out a new type of test. By fixing a paper screen to the wall in a suitable position it was possible to arrange the mosquito so that only one of the bands was visible to it (Fig. 4). Once the mosquito started to move towards the visible band, the second band came into its view as soon as it passed the edge of the screen. A few of the responses are illustrated in Fig. 4. It will be noticed that quite often the sudden appearance of a new band at one side caused an immediate turn towards it.

The following observation was made with the ordinary cylinder having either one band or two equal bands. The mosquito was allowed to start moving towards one band and the cylinder was then set in rotation. So long as the rotation was slow the mosquito continued to face the same band, but if the rotation was rapid it failed to do so.

Tests were made in an all-black cylinder with a white band, 30 mm. wide, to see if movement to black described in the above experiment was due to contrast. If so, the white band on a black background should attract the mosquitoes as readily as the black one on a white background. In not a single case was movement or even a slight curvature towards the white band observed. When two white bands were presented the movement was towards the black intervening area. If the white band was directly in front, the mosquitoes curved away from it.

All the experiments described so far were carried out during daytime but several similar experiments carried out after dusk failed to reveal any change in the type of responses. In all such experiments the mosquitoes had been kept in natural light in the laboratory and therefore if there was any reversal of reaction due to natural causes it might have been expected at dusk. It was thought that the continued negative phototaxis even after dusk might be due to the strong artificial light used and therefore tests were carried out (1) with progressively reduced intensities of artificial light, finally almost reaching the limit of visibility; and (2) with the natural light of the evening.

Except for a few trials with subdued artificial light coming from below, these experiments could not be carried out in a white arena illuminated from below and therefore the experimental conditions were slightly modified. The inner surface of the cylinder was covered with black paper except for a narrow vertical slit of 8 mm. The floor of the arena was of black matt paper and to the top of the cylinder was fixed a long narrow black paper cone, about 25 cm. in height, with an aperture at the end for viewing the arena from above, without allowing any other light to fall on the arena. The electric lamp was placed about 30 cm. to one side of the stage so that it threw a horizontal beam of light at the level of the stage and the light penetrated the arena through the vertical slit in the black paper. A mosquito was put on the stage, the cylinder inverted over it and the light switched on. In this black arena with a horizontal beam of light, the mosquito was barely visible except when directly within the beam. It was impossible to adjust the position of the mosquito in relation to the illuminated slit and therefore whatever position it assumed in relation to the beam was recorded. It was noticed that all but one of over thirty specimens tested moved away from the beam towards the darker part of the arena. Fig. 5 A illustrates typical responses of a specimen of Culex molestus, and Anopheles maculi-pennis was found to behave in exactly the same manner. Even a considerable reduction in the intensity of the light failed to produce a change in the responses. When the experiment was carried out using the natural evening light after dusk, the responses were exactly as in artificial light. The tracks of the single exceptional Culex molestus which showed clear responses towards light are illustrated in Fig. 2B.

Numerous tests with over twenty specimens of each species were made without using the special apparatus described above. The mosquitoes were tested on a white sheet of paper placed on the lab oratory bench near a large window. Individuals of the horizontal beam marked by two were placed on the sheet of paper and the tracks made by them recorded. The tests were first made in the middle of the afternoon in bright light and the same specimens were tested at or after dusk in the same manner, at a time when the free-flying mosquitoes in the room were attempting to escape through the window. Fig. 6 shows the tracks made by a specimen of Anopheles maculipennis, before and after dusk. It is very clear that even in the subdued light of dusk specimens whose wings have been cut off continue to curve away from light. This negative phototaxis was very clear in most of the cases. In a few individuals the responses were indeterminate and the direction of movement did not seem to be related to the direction of the light, but in the majority the responses were consistent and movement was away from the light, even after dusk.

During these experiments, when the mosquitoes were walking away from the window, a strong beam from a flashlight was often turned on the mosquito from several angles. Invariably the mosquito immediately turned away from the beam.

Since it was considered probable that the consistent negative phototaxis of the mosquito with the wings cut off might have been due to some damage to an undetermined sensory centre, a few tests were carried out with specimens in which (1) the wings were stuck together with a small drop of shellac placed at their tips, and (2) with the wings simply tied together with a piece of hair. Both these types-continued to exhibit the movement away from light, before, at and after, dusk.

A variation of the above experiments, to reduce physical handling of the mosquitoes to a minimum, was tried as follows : Eight to twelve specimens of Culex molestus with the wings removed were put into a covered Petri dish and placed on a black sheet of paper on the laboratory table. The Petri dish was covered by a bigger shallow glass dish, half of which was lined by black paper both at the top and sides so that one-half of the Petri dish became completely shaded. The specimens invariably walked towards the darker half of the Petri dish and settled there. After 15 min. the Petri dish was very gently rotated so that the section with mosquitoes came to lie in the full light.

In a few seconds the mosquitoes began to move again towards the shaded half and collected there. This was repeated several times with similar results. The experiments were repeated both before and after dusk without any difference being noted in behaviour.

The important points arising from the above observations and requiring a few words of elaboration are (1) the consistent negative phototaxis without any indications of a reversal of response at or after dusk, and (2) the mechanism or orientation:

The total absence of any reversal of the direction of response to light at dusk was an unexpected and striking discovery. Hundertmark (1938) has suggested, on the basis of certain laboratory experiments with flying mosquitoes, that such a reversal does occur. This matter certainly needs further experimental investigation, but the well-known fact that mosquitoes are ‘attracted’ to the exits of shelters at dusk, irrespective of their physiological state, has been clearly established in recent studies (Eyles & Bishop, 1943, Viswanathan, Ramachandra Rao & Ramarao, 1944).

Possible explanations of this behaviour are: (1) that the phototactic response is reversed in sign ; (2) that the mosquitoes show positive phototaxis in weak light, up to such intensities as occur in the evenings and that no reversal of reaction is involved (Weiss, 1913); (3) that the behaviour is due not to taxis but to kinesis.

Be that as it may, it is established that flying mosquitoes tend to move towards the bright exits of their shelters at dusk and it is therefore of considerable interest to find that mosquitoes rendered flightless always move away from light however weak it may be, even in the natural twilight of the evening. It could not have been due to damage to any undiscovered sense organs located on the wings or connected with the wings because the same type of response occurs in individuals with wings which are lightly tied up together with a piece of hair and are otherwise quite undamaged. Considering also the fact that responses are similar in specimens of varied gonotrophic conditions, it seems highly probable that the movement of flying mosquitoes towards light at dusk is a process which involves the use of wings, and is a phenomenon bound up with the spontaneous flight activity. This aspect of the problem will be dealt with at greater length in another paper in connexion with certain studies on the phases of the diurnal cycle of activity in mosquitoes.

It has to be considered whether this permanent negative phototaxis of ‘wingless ‘specimens could be explained as an ‘escape ‘response of an ‘excited ‘individual or as a shelter-seeking response of a disabled one. All the specimens used in these experiments were, in a way, disabled individuals because none of them could use the wings. Some of them must certainly be regarded as ‘excited ‘because of repeated handling and knocking about, but quite a few, such as those used in the Petri-dish experiments, were practically undisturbed. Among the latter the only handling involved was a very gentle and slow rotation of the dish to bring the originally shaded half into the light and the lighted half into deep shade. After this gentle change, the mosquitoes slowly began to move and then to walk back to the shaded part. These responses, observed in the total absence of physical disturbances of any kind, were exactly similar to the responses of mosquitoes which were thrown about, often with some force, in the arena. This seems to indicate that the movement of walking mosquitoes towards darkness does not depend upon ‘excitation’. In any case the flying mosquitoes, in an excited state, are supposed to fly towards light as an immediate response and not towards darkness (Polezhaev, 1936; Kennedy, 1940). Inability to use the wings seems to lead to inhibition of certain types of activity, a problem of some physiological interest. It may be that the lack of wings, so essential for normal activity, always induces mosquitoes to look for shelter and, therefore,-instinctively to move towards darkness.

As to the mechanism of orientation, it may be stated that the present studies broadly confirm the observations on flying mosquitoes made by Kennedy (1940). It is very difficult, however, to determine whether the movement towards the black bands is tropotactic or telotactic. Fraenkel & Gunn (1940) have made it clear that the same animal may be tropotactic and telotactic at the same time. As few animals can see directly behind them (and the mosquito certainly cannot see directly behind it; for example, a black band, 10 mm. wide, placed directly behind it in the cylinder at a distance of 3 · 5 cm. is practically invisible to it), there can be no negative phototelotaxis. Even when the edge is at 135 °, as has been shown in the experiments, only in a few cases does the movement to black take place. It is not possible to regard the movement to black exhibited by the mosquitoes as telotaxis, unless, of course, we believe in the existence of ‘skototaxis ‘as a definite type of orientation. But on the evidence at present available the behaviour can be explained equally convincingly as negative phototropotaxis. Unfortunately the circus movements performed by unilaterally blinded specimens in uniform light such as described above can be performed telotactically as well as tropotactically and therefore are not likely to be useful in distinguishing between these two types of response. The movement to one of the dark bands when two are presented and not to a point in between them has also been explained by Fraenkel & Gunn (1940) as due to negative tropotaxis. The only semblance of telotaxis is the response of a mosquito (Fig. 2C) already moving towards one of two visible dark bands, deflecting its course towards the other. The conditions are quite different when a new band suddenly comes into view as in Fig. 4. Whether this can be explained as an incipient form of vision is a matter requiring further investigation. But it appears to be almost exactly like the responses found by Kennedy when he observed that Aëdes was attracted to and followed stripes which were moving, abandoning stationary ones. It is possible that to the walking mosquito already moving towards a stationary stripe straight ahead, a new stripe which suddenly looms into view on one side provides an illusion of a backwardly moving band and the mosquito therefore turns towards it. Once it turns and heads towards the new stripe the direction becomes fixed and the old stripe no longer influences it, just as in the case of the experiments with two bands both of which were simultaneously visible from the very beginning.

Turning now to the case of the unilaterally blinded specimen performing circus movements, we recall that when the dark band suddenly comes into the field of the functional eye the animal turns abruptly towards it. How far this type of behaviour can be fitted into the general idea of a negative phototropotaxis is not clear. If tropotaxis is a result of a balance of stimulation by simultaneous comparison, then circus movements should continue even in a beam or gradient ; the total ipsilateral turning effect of the blind eye of a negatively phototactic animal should continue to be greater than the opposing effect produced on the functional eye by a narrow dark band, because the latter can affect only a small proportion of ommatidia of the functional eye and not all of them. Here again the effect of the sudden appearance of the band in the view of the functional eye may result in an illusion of movement leading to orientation towards it.

There is another important point to mention regarding the region of the dark bands towards which the movement takes place. In typical cases when the band was either to the left or the right of the specimen, the movement involved first turning towards the appropriate side and then walking towards the black band. In such a movement the first point of the dark band to be reached was usually the extreme edge. The mosquito sometimes walked a few millimetres from the edge of the black band towards its centre, then climbed the wall and stayed there. But in many cases the movement proceeded further than that and the mosquito climbed the black wall at about the middle section of the band. But since in a few cases the mosquito remained near the edge it was necessary to make further experiments to discover if there was any evidence of photohorotaxis (Kalmus, 1937). Accordingly, many observations were made in the cylinder with the 90 ° black band, placing the band in such a way that the mosquito directly faced its centre (Fig. 1E). It was noticed that the mosquitoes moved towards the middle of the band without showing any tendency to curve towards either edge. The apparent attraction of the edge when the band was placed to the side was simply due to the fact that the mosquitoes had to move along a curved track towards the band and so reached it first near the edge, where they then remained. In the experiments with two bands it was also a matter of interest to observe the position at which the specimens came to rest. There was no indication of their coming to rest near the inner edge of one of the bands: had they done so the idea of a negative phototropotaxis would have received further support. On the whole the observations do suggest that the visual orientation of the mosquito can be regarded as due to a negative phototropotaxis.

Unfortunately lack of time did not permit quantitative estimation of turning responses in relation to the position of a single dark band at various angles in the horizontal plane. It was noticed that when the edge of the dark band was directly in front or about 45 ° to left or right of the longitudinal axis of the head the responses were perfect, that is the movement to the black was nearly always completed. But when the edge lay at about 90 ° to one side the movement to black did not take place quite so often, and when the edge of the band was 135 ° towards one side the movement to black was still less common. In a few trials conducted in a cylinder with a band of only about 10 mm. width, very similar differences in the responses were noted. These suggest that the anterior ommatidia are more sensitive to the black band or perhaps initiate turning movements to black more strongly than lateral ommatidia, a view which is in agreement with Kennedy’s (1940) postulation. Of course, if the band is directly behind them the mosquitoes cannot see it, and therefore are not likely to respond to it. But when it is anywhere from 0 to 110 ° from the front the number of ommatidia stimulated must be more or less the same, because at this plane the visual angle of the eye is very nearly 225 ° all round and 110-115 ° on either side. It might be possible with this new method of study with wingless mosquitoes to make a complete reflex-map of the eye by using a small dark spot of adequate dimensions and exposing it to the mosquito at all possible angles so as to affect particular groups of ommatidia.

The variabilities in behaviour of individual specimens have already been mentioned. Only about one in ten of the specimens failed to show any sort of consistent response; the reason for this is obscure. Fraenkel & Gunn (1940) (pp. 292-3) have stressed that the future behaviour of a particular individual is always unpredictable with a 100% assurance, but what the majority of a given population will do can be foretold reasonably correctly. Exactly the same thing occurs in the mosquitoes so far examined. There must always be unusual individuals in any group which may be regarded as extreme variants of a normal distribution. An analytical study of their behaviour would contribute to an understanding of the behaviour of the normal ones. In the present series of observations, however, the behaviour of the atypical forms has been extremely inconsistent, and they have shown no regular pattern of behaviour.