ABSTRACT
In a long temperature gradient apparatus, Ptinus tectus aggregates around two distinct temperatures, 4 and 24° C.
In a circular temperature alternative chamber there is no marked avoidance of low temperatures if the gradient is steep, but characteristic avoiding reactions occur if the gradient is not very steep.
When the stationary animals are subjected to rising or falling temperatures in the long gradient, there is a well-marked evacuation temperature at 30–33° C., but none at lower temperatures.
A temperature preference around 24° C. is correlated with favourable temperature for growth and other activities. The aggregation around 4° C. is the result of a failure of co-ordinated behaviour in a particular kind of gradient, and P. tectus is practically immobilized by cold in that temperature region.
INTRODUCTION
Previous papers in this series (Bentley, Gunn & Ewer, 1941; Gunn & Hopf, 1942; Ewer & Ewer, 1942) have provided information about the relations between Ptinus tectus and environmental factors, particularly temperature; this paper is concerned with the temperature region in which this species tends to aggregate when placed in a temperature gradient. Determination of this preferred temperature presents certain difficulties which have been recorded by Deal (1941) in connexion with several species of beetles but, as will be seen, on the whole the results fall into place with information already obtained here about P. tectus.
APPARATUS AND METHODS
The beetles used in these experiments were descendants of a stock supplied a year earlier by the Biological Field Station of the Imperial College at Slough. They were bred in l. jars in 300 g. of the usual wholemeal flour and yeast mixture. The animals were always allowed to drink before an experiment.
The temperature gradient apparatus mainly used was the one described by Gunn (1934, 1935), and a second kind of gradient apparatus is dealt with on p. 137. The former has been silver-plated, for use with aquatic organisms, and the cooling was done by putting ice or solid carbon dioxide in a box fixed to one end, instead of by circulating cold alcohol around that end. By means of solid carbon dioxide a temperature of − 6° C. could be maintained at the cold end and the insects could not reach the end corners; there was no difficulty about keeping the hot end of the gradient hot enough to keep the insects away. Interpretation of the results was therefore not complicated because of reactions to the ends of the apparatus. In order to eliminate the tendency of the insects to collect round the thermometers (Deal, 1941), the holes for them were in the double glass roof, instead of in the side walls. Instead of the insects being allowed to walk on the metal floor, when the smallest irregularities lead to aggregation in response to the extra contact stimulation, a false floor of fine bolting silk was slung from the sides of the apparatus, providing a trough-shaped run for the animals. In order to remove smells left by the animals, which seem to be an important circumstance causing aggregation in places where aggregation has occurred before, the bolting silk was washed and ironed after each experiment. This material was more satisfactory than smooth metal, for when the beetles tumbled over on to their backs they more quickly regained the feet on it than they did on metal.
The apparatus was covered with a black light-tight plywood cover and a curtain which could be lifted for inspecting the animals in the gradient. Inside the cover there were two alternative sets of lamps, one (neon) giving an intensity of about i me. and the other of about 10 me. around the animals.
The humidity of the air current passed along the gradient could be controlled and measured as previously described (Gunn, 1934). Using the standard range of temperature (− 4 to +38° C.), moist air passed in at the cold end would deposit moisture in the entry pipe and so leave the air nearly dry, while if put in at the warm end it would condense in the gradient on reaching its dew-point. If, therefore, condensation in the gradient were to be avoided, the behaviour of the animals could not be tested in moist air, and the only comparison possible was between dry air and nearly dry air. On the other hand, it was possible to investigate the effect of a somewhat higher humidity without much interference with behaviour due to liquid water, with air entering at the warm end, because the condensation appears to have taken place mostly on the metal of the apparatus, while the animals themselves were on bolting silk and could not touch the metal.
It has been shown by Bentley et al. (1941) that the frequency of locomotory activity of P. tectus varies with a diurnal rhythm, under suitable conditions, and that this rhythm can be controlled by controlling the times of day during which the animals are in the light. In the temperature preference experiments three groups of animals were used: rhythmless animals, which had been kept at 25° C. in unvarying light long enough to abolish the rhythm; day-phase animals, which were similarly kept, but in artificial light of about 25 me. in the day (06.00−18.00 hr.) and in darkness at night; and night-phase animals, which were kept in darkness during the daytime and in the artificial light at night (18.00−06.00 hr.). Since all the experiments were done in the daytime, day-phase animals were in a rather inactive condition, rhythmless animals were more active, and night-phase animals, with their normal rhythm reversed, were the most active.
The effects of the three variables—light intensity in the experiment, air humidity, and phase of activity—were studied in the twelve possible combinations. Two such sets of experiments were carried out.
The twenty animals used in each experiment were put into the gradient through a thermometer hole at 26−27° C. ; this point was chosen because it was not the preferred temperature and not a damaging temperature, but a temperature at which the animals were active. If, instead, animals were initially scattered along the gradient, those put into the colder parts tended to stay there, and records of their positions were not representative of their locomotory behaviour. The positions of all the animals were recorded every 10 min., usually for 5 hr., thus providing 600 observations in each experiment. The direction of the air current was reversed hourly.
RESULTS
The results of all twenty-four experiments added together, including some 14,800 individual readings, are shown in Fig. 1. In Fig. 2 are shown the results of six of these experiments, selected to illustrate particular points.
The average preferred temperature from the data shown in Fig. 1 is 12·1° C. This average temperature is obviously not representative of the behaviour of the species, for fewer animals were recorded at about that temperature than at any other temperature between 1 and 29° C., while considerable numbers were recorded around 4 and 24° C. Fahmy (1931) does not mention the aggregation of Ptinus in the cold region; in a few rough experiments, about 40% of the animals settled at 22−25° C., but nothing was said of the rest. Deal (1939) found only a minor aggregation at 20−25° C., while most of the animals collected at the coldest end of the gradient at about 8° C. ; the animals were active and not immobilized by the cold and Deal concluded that this aggregation was due to a true temperature preference. He also found a bi-modal distribution of the earwig, Forficula auricularia, with a large peak at the cold end and a smaller one at 25–30° C. The bi-modal distribution of Ptinus requires further consideration.
In some of the twenty-four experiments there was little or no aggregation at the high temperature (Figs. 2f, d) and in others little at the low (Fig. 2,b). Thorough treatment of the data failed to reveal any correlation between the occurrence or absence of these two aggregations and the three variables investigated. Thus Figs. 2 a, b (also Figs. 2 d, e) are for the same conditions but do not agree, while Figs. 2 a, e are for very different conditions but are similar. It therefore appears that the relative magnitudes of the two peaks depended on chance.
In view of this conclusion it is worth noticing that in Figs. 2 b, e the aggregation between 20 and 30° C. does not occur over quite the same temperature zone; in Fig. 2 b the peak is at 21–23° C., while in Fig. 2 e it is at 27–28° C. These are selected extreme cases, but clearly a single experiment is not enough to show the precise position of such a peak. This is especially true when testing insects which tend to collect in clumps, for the exact position of the peak depends both on temperature and on the place at which a small clump happens to form in the beginning.
In both of the main aggregation regions the animals become immobilized, but apparently for different reasons. At 24° C. Ptinus can be active quite frequently and it can walk quite quickly; at 4° C. activity is rare (Gunn & Hopf, 1942), walking is very slow, and immobility is forced upon it by the physiological effects of cold. It seemed rather improbable that an insect would have such poor protection against immobilization by cold as this behaviour suggests, and it seemed likely that the failure of behaviour was due to the form of presentation of the stimulus.
Two features of the apparatus seemed likely to be involved, namely, its shape and the steepness of the temperature gradient. The metal apparatus is long and narrow (122 × 4 cm., Gunn, 1934), and the bolting silk was slung so as to form, in cross-section, the shape of a wide ⋃ with the ends turned over to the horizontal. The animals usually walked along the sides of the apparatus with their backs nearly touching the roof. As long as they kept to this line, manoeuvring was limited; but this was certainly not the cause of the low-temperature aggregation, for when the bolting silk was stretched horizontally on a frame, such aggregation still occurred. The limitation on manœuvre in the latter case was not severe, for Ptinus is only 5 mm. long, including antennae, and it had a width of 4 cm: in which to turn round. The possibility nevertheless remains that the shape of the apparatus canalized the movements of the insects. Some experiments were therefore carried out with a different form of apparatus.
Alternative chamber
The other form of apparatus used was of the alternative chamber type constructed by Thomson (1938). It consisted essentially of a circular glass dish (29·5 cm. diameter and 4·5 cm. deep) with a double glass roof, through which thermocouples were inserted. This chamber was almost submerged in a water tank which was divided into two by a vertical wooden partition. Half of the chamber was on one side of the partition and the other half on the other; the water on the two sides was kept at two different constant temperatures.
The air in the chamber was kept dry with sulphuric acid, and there was a perforated zinc false floor as in the humidity alternative chamber (Gunn & Kennedy, 1936). The temperature did riot change sharply over the partition, presumably because of conduction in the zinc, but the isotherms were not sufficiently evenly spaced to allow the temperature at any point to be inferred from a row of seven stationary thermocouples across the chamber; the isotherms could be obtained by rotating the roof carrying the thermocouples into various positions. There were considerable gradients of temperature between the zinc platform and the roof 2·5 cm. above; at about 9 and 29° C. the differences between the temperatures of zinc and roof were about 3° C. at a room temperature of 19° C. This result does not agree with Thomson’s (1938) statement that ‘there was no appreciable vertical gradient of temperature such as necessarily exists in the usual type of apparatus’. In the long apparatus, the vertical differences of temperature are usually less than 0·1° C., though in Herter’s apparatus they may reach 10° C. or more in a height of 10 cm. The situation in the alternative chamber is therefore more complicated than in the long apparatus because of convection currents and because of uncertainty about the precise temperatures to which the animals are reacting; but the alternative chamber has the great advantage that it offers a large arena for manœuvre and it has no corners or ends, in which insects tend to collect.
Results
In this apparatus, with a gradient from 6 to 13° C., there were repeated turns away from the cold part of the chamber at about 10° C., and not one of eight animals tested entered the colder half of the chamber. In a gradient from 8 to 36° C., however, behaviour was much as in the long apparatus, with prolonged periods of inactivity in the cold regions. Thus Ptinus can show reactions against low temperatures if the steepness of gradient is suitable. The surprising fact is that it is not a steep gradient but a gentle one which elicits the successful avoiding reactions.
The successful avoiding reactions observed in the gentle gradient were generally of a well-defined type: the animal slows down, or stops altogether, swings the antennae from side to side, then suddenly turns right round and walks off in the opposite direction, the whole reaction taking about 2 sec.
Turning reactions in the steep gradient were not usually of this type. A preliminary analysis of tracks recorded in this gradient showed that turns amounted to about 250, 900 and 1600°/min. at 10, 20 and 30° C. respectively, varying irregularly but without any indication of particularly high values at any special temperature. Average speeds, were roughly 6, 22 and 38 cm./min. When the amount of turning was referred to distance travelled instead of to time, it was found to vary around 40°/cm. without any consistent trend, the values for all temperatures lying between 20 and 60°/cm. These figures are based on a small number of observations, but the analysis could not be carried further because the work had to be broken off.
Evacuation temperature
Further evidence was obtained on the difference in behaviour towards high and low temperatures. A score or so specimens were induced to clump near one end of the long gradient apparatus by utilizing their photonegative behaviour. This part of the apparatus was then heated or cooled and a record was made of the temperatures and the movements of the insects. Starting at 25° C. the temperature was raised at about 1° C. in 5 min.; the animals began to leave at 30° C. and all had left for cooler parts at 33° C. When the temperature was lowered from 20 or 25° C. at the same rate or faster, the insects were activated, the clumps broke up and reformed, but less than one-third of the animals left before cold immobilization occurred. That is to say, the upper evacuation temperature is well defined at 30–33° C., but there does not seem to be a lower evacuation temperature at all.
DISCUSSION
It seems likely that behaviour evolves, as structure does, in co-ordinated fashion. If, then, it appeared that the behaviour of Ptinus is poorly co-ordinated with its physiology, that would be a matter worthy of closer investigation. A temperature of 24° C., around which Ptinus aggregates, is quite favourable for development and egg laying (Ewer & Ewer, 1942) and for activity (Gunn & Hopf, 1942). On the other hand, 4° C. is probably Eelow the developmental minimum for the species (Ewer & Ewer, 1942) and activity is reduced to a very low level (Gunn & Hopf, 1942). Even a temperature of 8° C., the lowest available in Deal’s apparatus (1939) and an aggregation temperature in his experiments, is at least very near the developmental minimum and may be below it. Only eighteen eggs were laid by 200 adults in 6 days at 7° C. (Ewer & Ewer, 1942). If Ptinus were frequently to choose such a low temperature in nature the species could hardly survive. The results described above indicate, however, that the occurrence of aggregation at low temperatures depends upon the form of the apparatus used, and if the gradient of temperature is gentle enough the low temperatures are avoided.
It has been shown by Ullyott (1936) that when the flatworm, Dendrocoelum, reacts klino-kinetically in a gradient of light, aggregation in the dark region fails to occur if the gradient is not steep enough; sensory adaptation then takes place sufficiently quickly, as the animal moves into brighter light, to prevent effective stimulation of the receptors (Fraenkel & Gunn, 1940). A steep gradient should produce the most effective reaction, culminating in the familiar avoiding reaction where the gradient is steep enough to constitute a boundary. With Ptinus in a temperature gradient, however, it is the steep gradient which fails to. produce a reaction against cold and the gentle gradient which succeeds. This fact is inconsistent with both the older conception of shock reaction and the new theory of klino-kinesis. Analysis of tracks in the alternative chamber (8–36° C.) provided no support for the latter, but, of course, the animals did not then react against cold. When they did react (6–13° C.) the turnings were not random but resulted in a complete reversal of direction. Unfortunately, experiments with gentle gradients in the long apparatus were unfruitful because of the almost immediate aggregation of the animals at one end or the other.
The steepness of the gradient, besides affecting the state of sensory adaptation, is also important in relation to the lag in body temperature and the consequent lag in adjustment of speed of walking. Little can be said about the lag in body temperature as the animal walks along a gradient. An attempt was made to determine the lag, using a thermocouple, but the only ammeter available was rather sluggish, and the conclusion reached was that body temperature follows environmental temperature over a change of 10° C. with at most a few seconds delay. From the speeds of walking given above, it will be seen that Ptinus would normally traverse a gradient 55 cm. long extending from 10 to 30° C. in about min. That is to say, body temperature is not likely to differ much from environmental temperature as the animal walks about. It is hardly likely that the animal rushes into the cold, with its body still warm, but cannot get away before cooling stows it down too much. In any case, this does not correspond with observation; the animal walks more and more slowly as it approaches the cold end. At the same time it is clear enough that temperature lag would be less in a gentle gradient than a steep one.
The metabolic effect, as distinct from the behaviour one, is clearly important. If an animal overshoots the high temperature at which turning back normally occurs, locomotion is progressively accelerated and a successful reaction at a still higher temperature can and does occur. On the other hand, overshooting the low-temperature reaction point leads to immobilization, and it is rarely that an animal gets away again to a higher temperature. That is to say, a single failure to react against cold leads to immobilization; a high proportion of the more active animals, which approach the colder regions several times, will be trapped by cold sooner or later.
Low-temperature trapping of insects has been recorded before by a number of authors (see Fraenkel & Gunn, 1940). Because of this phenomenon, Nieschulz (1934, 1935) calculated a ‘purified’ average preferred temperature which ignored small percentages of flies (Musca domestica, Stomoxys calcitrant and Fannia canicularis) spread over low temperatures. This procedure has the disadvantage of depending on an arbitrary choice of what shall be regarded as a temperature low enough to cause metabolic trapping. Whether an aggregation of Ptinus at 4° C. is said to be due to a temperature preference or not is a matter of definition, but it seems more useful to regard only the aggregation around 24° C. as due to the true preference. Errors due to the occurrence of metabolic trapping can be avoided by using suitable apparatus.