The Humidity Behaviour of the Mealworm Beetle, Tenebrio Molitor L: III. The Mechanism of the Reaction

The analysis of the mechanism of behaviour reactions of animals to diffuse molecular stimuli—odours, humidity and air temperature (as distinct from “radiant heat”)—has been attempted only on a limited scale. This is no doubt due to the difficulties of controlling and measuring the conditions and of evolving a technique as accurate and precise as that which can be used when light is the stimulus. Such analyses as have been made, in general, attempt to carry over the theories and classifications—especially the classification of Kühn (1919)—that have been developed mainly in relation to light. The modifications made by Gunn (Fraenkel & Gunn, 1940) in Kühn’s classification make it more suitable for the inclusion of reactions to diffuse stimuli.

This system includes two main categories: kineses, or undirected reactions, in which the direction of movement of the animals is not precisely related to the direction of the source of stimulation, and taxes, the directed reactions in which the orientation of the axis of the animal’s body is influenced by the direction of the source. In ortho-kinesis, the average velocity of locomotion or the frequency of activity depends on the intensity of stimulation, and aggregation occurs in the region of lowest velocity; in klino-kinesis (avoiding reactions) the amount of turning during locomotion depends on the intensity of stimulation and upon some other factor like sensory adaptation (Ullyott, 1936; Gunn, et al. 1937). In klino-taxis the animal compares the intensities of stimulation in its vicinity by alternative movements to the left and right (either movements of the whole body or of some receptor region such as its antennae) and the comparison is thus between two intensities which are successive in time. In tropo-taxis the behaviour is similar but orientation is immediate, by simultaneous comparisons of intensities at two symmetrically placed receptors. Telo-taxis seems to be out of the question with diffuse stimuli (Fraenkel & Gunn, 1940).

There have been several reports of directed reactions (taxes) to odours, for example by Barrows (1907) and Flügge (1934) on Drosophila, by Warnke (1931) on Geotrupes, by Murr (1930) on Habrobracon and by Hartung (1935) on Calliphora. Fraenkel’s analysis of these reports (Fraenkel & Gunn, 1940) suggests that a whole series of reactions (ortho-kinesis, klino-kinesis, klino-taxis and tropo-taxis) come into action in turn as the animal approaches the source of stimulation and gets into a steeper and steeper gradient. No directed reaction to the stimulus of humidity has yet been described but ortho-kinesis towards humidity has been established for the woodlouse (Gunn, 1937) and for the locust (Kennedy, 1937).

The material, apparatus and methods used in this work have been described, in the main, in previous papers (Gunn & Kennedy, 1936; Gunn, 1937; Pielou & Gunn, 1940; Pielou, 1940).

Experiments were carried out with the humidity alternative chamber (Gunn & Kennedy, 1936) using the short gradient of humidity from 94 to 100% R.H. in which Tenebrio molitor shows a very intense reaction (Pielou & Gunn, 1940). Here the animals were used one at a time and the paths of the animals in the chamber were followed and recorded on squared paper. The false floor of the chamber was marked off into squares to facilitate this recording. A mechanical device sounding at half-minute intervals was used and time marks were recorded on the tracks; this allowed the rates of movement and times spent on each side of the chamber to be worked out. Twenty-two animals were observed in this way for a total period of slightly over $2512$ hr.

The times spent on the drier side (94—97% R.H.) and on the moister side (97—100% R.H.) were obtained from these track records. The detailed results are shown in Table I and their most striking points shown in Fig. 1. The total time spent on the drier side is, of course, far greater than that spent on the moister side, and this is in keeping with previous experiments conducted in a different manner (Pielou & Gunn, 1940). It is also seen from Table I and Fig. 1 that the animals practically never come to rest on the moister side, they remain continuously active there, only 6 % of their total time on this side being spent motionless. On the drier side, however, over 80 % of their time is spent inactive and, as the total time spent on the drier side is greater than on the moister, the ratio of the absolute times spent in the inactive state on the two sides is even more striking (⁠$1912$ hr. to 4 min.).

This very marked differential activity observed with such a small difference of humidity undoubtedly accounts for the very intense reaction to a great extent. A similar pronounced hygro-kinesis, but in the reverse direction, was found in the woodlouse, Porcellio scaber, by Gunn (1937). On the other hand in the mosquito, Culex fatigans, which reacts strongly to humidity, there is apparently no such effect (Thomson, 1938).

This conspicuous activity difference (ortho-kinesis) cannot, however, be the sole factor in the reaction since it will be noted that the absolute times spent by the animals in the active state is five times higher on the drier side than on the moister (Table I). This may be accounted for either by a difference in the rates of movement of animals when they are active or by some turning mechanism tending to keep the animals on the drier half of the chamber. There is no obvious change of pace of the beetles when they cross from one side of the chamber to the other as in the woodlouse (Gunn, 1937), but in this case a more detailed estimate of the rates of movement can be made from the time marks and their distances apart on the charted tracks. The values for these rates on the two sides of the chamber were obtained, the distances being estimated with a map-measuring instrument. There was found to be no significant difference in the rates of movement of active animals on the drier and moister sides of the chamber. It is therefore clear that the other possible factor, namely some form of turning mechanism, must be playing a considerable part in the reaction. Experiments on this point are described later.

Differential activity of the type described in the above section may also be exhibited when animals are kept at different constant humidities; at one intensity they are continuously active, at another they remain inactive. This is the case in the woodlouse (Gunn, 1937). The behaviour of Tenebrio molitor was examined under these conditions. The apparatus used was the activity chamber, a closed vessel kept at a constant uniform humidity. In order to make the results comparable, the vessels used for this purpose were the same as those used in the previous humidity gradient experiments. Five animals were placed in each chamber and the numbers active and inactive noted after 15 min. The animals were then activated mechanically by pushing them with a pipe cleaner and the process repeated twenty times daily so that each experiment yielded 100 records of activity. The experiments were carried out at various different humidities and repeated several times at each humidity. The results obtained from data on 300 animals in sixty experiments are shown in Fig. 2. It will be seen that the activity increases sharply as the relative humidity approaches 100%. This is in keeping with the experiments previously described by Pielou & Gunn (1940) on the reactions of the beetles to differences of humidity in which the sensitivity increased sharply as the highest relative humidity approached 100 %. It will be seen, however, that when these insects are kept at constant humidities they do not exhibit such a marked difference of activity as is shown in the alternative chamber where there is a gradient of humidity. A further more detailed series of experiments was carried out to examine this point.

The animals were placed in activity chambers as in the above series but the numbers active and inactive were noted at 3, 6, 9, 12 and 15 min. following each activation. The results are shown in Fig. 3 which are obtained from twenty experiments involving 100 animals and yielding 10,000 activity records. It will be seen that in constant humidity the activity tends to fall to a basal level; the animals are more active in wet air than dry air. But the extremely marked difference of activity seen in the alternative chamber (Fig. 1 and Table I) is much reduced in constant humidities. At the drier side of the alternative chamber the relative humidity is slightly over 90 % and here 80 % of the animals’ time is spent inactive; in a constant relative humidity of 90 % approximately 90 % of their time is spent inactive (Fig. 3). There is no marked divergence here. But in the alternative chamber on the wet side (100 % R.H.) only 6 % of the animals’ time is spent inactive; at a constant relative humidity of 100 % however, 75 % of the time is spent inactive. There appears therefore to be a marked reduction of the activity of animals at 100% R.H. if they are kept at this humidity even for a short time. This marked reduction is no doubt due to sensory adaptation, for in the alternative chamber they are constantly exposed to changing humidities, for a few centimetres movement by the animals is sufficient to carry them right across the chamber, with no chance to adapt.

The track records reveal at once a reaction which intensifies the differential activity mechanism. On the drier side, and only on the drier side, the animals are frequently “virtually inactive” although they are moving about. This is because in this type of movement they turn very frequently and sharply and remain in a small confined area (Fig. 4B); they do not move far, spend much time poking their heads through the holes in the zinc platform and, if they are near the edge of the chamber, trying to climb up the walls. Rather more than one third of the total time spent in active movement on the drier side is occupied in this type of movement (Table II). This “virtual inactivity” tends to confine the animals on the drier side just as does the real inactivity so preponderant here.

It is clear from the data already described that the bulk of the intense reaction is accounted for by a marked differential activity or ortho-kinesis; very frequently the tracks show evidence of no other mechanism (Fig. 4A). The “virtual inactivity” also accounts for part of the reaction (Fig. 4B). Besides these, however, records are also obtained (Fig. 4C, D) which seem to show that some other reaction is at work. These tracks are characterized by frequent turning movements which, in general, tend to keep the animal, when actively walking, on the drier side of the chamber. Such movements must explain why the time spent by the animals in the active state, even after “virtual inactivity” is excluded, is three times greater on the drier side than on the moister (Table II). These turning movements appear, in general, quite random and undoubtedly they mainly represent an undirected reaction towards humidity, not of the sharply defined and stereotyped kind seen in Paramecium (Jennings, 1904) but of the more meandering type analysed by Ullyott (1936) in the flatworm Dendrocoelum lacteum. These reactions are of the type usually called “phobotaxis” (Kühn, 1919) but are better designated as “klino-kinesis” (Gunn et al. 1937). In this group of reactions the random movements of the animal as a whole bring it into a successive series of different intensities of stimulation; there is not necessarily independent movement on the part of the receptors. In a separate series of experiments, the movements of the antennae (on which the humidity receptors are situated (Pielou, 1940)) were observed with a lens and it is clear that this type of bold random movement is accompanied by very little motion of the antennae. This therefore provides additional evidence for supposing that this type of reaction is undirected.

Apart from these undoubted random and undirected turning movements, there are some indications of a small directed element in the reaction. Sometimes actively moving animals slow up, even stop momentarily, and during this hesitant phase make very varied trial movements with the antennae; this is followed by a fairly precise movement towards the dry side (Fig. 4E; trial movements at T). This appears to correspond to the type of behaviour reaction in which the path of the animal is directed with the aid of trial movements of the receptor region of the body into regions of different intensity of stimulation. Here the antennae are the receptor regions and presumably their Johnston’s organs provide the animal with information about the direction in which the antennae are pointing and so can result in “a reflex pursuit of the antennae” (Wigglesworth & Gillett, 1934). This is the type of orientating mechanism now known as klino-taxis (Fraenkel & Gunn, 1940). Such extreme cases as those shown in Fig. 4E are however, rare; only fifteen clear-cut cases were seen in a total of over 25 hr. observation. An example of this type of behaviour is seen in the reaction of Rhodnius towards temperature (Wigglesworth & Gillett, 1934).

Further evidence on the existence of a slight directed element in the reaction was obtained from a detailed analysis of the directions of the turning movements made by the animals. The tracks particularly examined were those in a central zone 2·5 cm. wide and parallel with the mid-line, where the gradient is steepest. Records were picked out in which the paths of the animals were, for some time, parallel or very nearly parallel to the mid-line; that is, the movements of the animals were approximately at right angles to the direction of the gradient of humidity. If there is no directed element in the reaction any turn from this position should be at random, without bias to left or to right. In many cases there was no turn at all, the animals continuing on, uninterrupted, to the edge of the chamber. Frequently, however, at some point the animal turned to one side or the other. Using 15⁰ as the limit of error in estimating these turns and counting turns to the dry side as positive and those to the moister side as negative, the average angle of turning was + 36·0°. This indicates a significant directed reaction towards the drier side; if the directed reaction was maximal the angle would be +90·0°. The average +36·0° was obtained from thirty-four records of such movements in the narrow central zone only, where the turning movements are most intense.

The corresponding data were also abstracted for such turning movements in a similar but wider zone, also central and parallel with the mid-line, 10 cm. in width. From this larger area 124 such records were obtained and here the average resultant angle of turning was + 29·0°.

All analysis dealing with turning movements was made from these records in which the animals were walking freely in the chamber. The animals spend a good deal of time walking at a normal rate but close to the edge of the chamber, but the ratio of the times spent on the two sides is the same as when the animals are moving freely in the open space of the chamber (Table II). Moreover, the turning movements are also shown here (Fig. 4 F). Presumably the same mechanisms are at work in both cases but the analysis can only be carried out with those track records in which the animals were free from the edge of the chamber.

There is no evidence of a true tropo-tactic element in the reaction; unilateral amputation of an antenna does not result in any tendency to turn to one side (circus movement) either in a gradient of humidity or in constant humidity. There is none of the marked circling of the sort shown by many unilaterally blinded animals when reacting to light, while statistical analysis of the records shows that there was not even a slight tendency to turn to one side. Circus movements in response to odours seem always to be of the latter kind, with a somewhat uncertain meaning, but even these were not found in Tenebrio.

1. The mechanism of the reaction of Tenebrio molitor towards humidity has been analysed into its component behaviour elements. An estimate has been made of the quantitative importance of each type of behaviour.

2. In a gradient of humidity from 94 to 100 % R.H. the beetle rarely comes to rest in the moister region but 80 % of its time in the drier region is spent motionless.

   In different uniform humidities this differential activity is still evident but very much reduced.

3. The reaction is intensified somewhat by “virtual inactivity”—restricted movement confined to a small area—which is shown only in the drier part of the gradient.

4. Animals approaching regions of high humidity show turning movements which have both an undirected component (klino-kinesis) and a directed component involving movements of the antennae (klino-taxis).

5. Tropo-taxis does not occur. There are no circus movements after unilateral amputation of the antennae.

The work described here and in the previous two papers was carried out during the tenure by one of us (D.P.P.) of a grant from the Department of Scientific and Industrial Research.