Tracheal Pulsation in the Flea

The tracheal phenomenon with which this paper deals was discovered accidentally during work on the effects of fumigants on the hedgehog flea, Archeopsylla erinaceae erinaceae.

The effects of HCN were particularly observed on the various body rhythms ; among these rhythms appeared the slow inflation and rapid deflation of the main branches of the tracheal system (called in this paper the “tracheal rhythm”). In the present paper the author does not attempt to give a dogmatic explanation of the phenomenon, as no wholly satisfactory one has arisen. As often happens, it is easier to show what does not cause tracheal pulsation than what does. It seems better, therefore, to give a description of what actually occurs, with three theories as to its cause, and their objections.

I should like here to express my sincere thanks for most useful advice, both theoretical and technical, to Dr V. B. Wigglesworth, Mr G. V. B. Herford and Dr G. Fraenkel.

The hedgehog flea was used for the present work, being one of the larger and less active fleas. Other fleas, such as the rat flea Xenopsylla cheopis Roths, or the dog flea have been used with the same results. The hedgehog, however, is a convenient animal to keep ; if two or three hedgehogs are kept in a pen of dryish earth the flea larvae can develop satisfactorily, and complete their life cycle. The adults cannot be obtained in a starved state in this way, but this was not necessary for the present work. The adult flea is long-lived, and can be obtained off the host in winter, when the hedgehog is torpid, though its parasites are still lively. The flea is caught on the hedgehog with a sucking tube. It is then held by the leg with forceps and transferred to the observation cage, which is constructed as follows; a pair of glass rods, about $14$ in. in diameter are placed along a microscope slide $14$ in. apart, and cemented in place with paraffin wax. Between them lie two glass tubes of the same diameter, also cemented with paraffin wax, with a $12$ in. space between their ends. This space is the gas chamber, whose upper rim is covered with paraffin wax. Within the chamber is built a glass platform almost flush with the top of the cage.

This is transparent, and is cemented with Canada balsam. The flea is held on this platform by the leg, and covered with a cover-slip, whose gentle pressure keeps the flea in position. (It is impossible to use narcotics to subdue the flea, as they destroy the tracheal rhythm, which does not return for several hours.) A hot needle is then run round the border of the coverslip, melting the wax beneath and sealing it in position. A second coverslip covers the open half of the cage, and the join between the two is hermetically sealed with vaseline, which must be carefully kept away from the body of the flea.

The cage is observed through a compound microscope, with a $23$ in. objective and a–15 eyepiece, and condenser. A 100 W. lamp is used, with a water heat screen, which keeps the temperature of the gas at about $12$ ° C. above room temperature. The experiments were run at room temperature which did not vary much from 20° C.

No appreciable difference can be seen between the effects of slowly moving and of still air, or of air at 85 % R.H. and that at 55–60% R.H.

The fleas succumb to the observation conditions as fast in one case as in the other. The usual length of life under observation conditions was about 24 hr.; after about 18 hr. the body rhythms (heart, gut) began to be erratic. Air was drawn through the apparatus with a vacuum pump. The different gases, when used pure, were drawn direct from cylinders, or mixed, and fed by water pressure from an aspirator. The flow was in all cases checked on a bubbler after passing through the cage. The change-over from gas to air and back was made by turning a three-way tap and switching on the pump.

The charts showing correlation of spiracular movement with spiracular rhythm were made on a revolving drum with two electric tapping keys worked by the hand and foot.

Respiratory system in A. erinaceae. The reader is referred to Wigglesworth’s paper (1935) for an illustrated account of the tracheal system, of which only a brief description is given here. The account applies to all species of fleas examined. Where abbreviations are used, Th. = thoracic spiracle, Abd. = abdominal spiracle. Thus, Abd. viii = the 8th abdominal spiracle.

The important spiracles for the present discussion are Th. ii, which opens on a small papilla behind and below the mesopleuron; Abd. i (in Wigglesworth’s paper = Th. iii) near the upper margin of the metepimeron; and Abd. viii, opening high up on the 8th tergite. Th. 1 and Abd. ii to vii are small and apparently unimportant. The structure of the abdominal spiracles is entirely different from that of the thoracic spiracles. In the abdomen, the trachea, before opening on the body wall, narrows abruptly and passes through a deep notch in a small crescentic chitinous rod lying across the trachea. When the ends of this rod are drawn together by an occlusor muscle the spiracle is closed.

The tracheae themselves, in all fleas examined, are not round in section, as in most insects, but more or less oval, except in small branches; they can therefore collapse more easily with reduced pressure. The whole system is covered, as in all insects, with a conspicuous epithelial sheath, which has about the same thickness over minute branches as over main trunks.

There occurs in the flea the phenomenon, possibly respiratory, of rapid, rhythmic contraction of the abdomen, similar to that shown by Hymenoptera and Orthoptera, though in the flea the period of the rhythm is about 2 sec. This contraction acts on the body fluids and gives corresponding sudden contractions of the tracheae. It usually occurs at times of exhaustion, or when the prothoracic ganglion has been severed.

Tracheal inflation and deflation. The tracheal rhythm is one of slow collapse of the main tracheal trunks, followed by sudden inflation. This inflation occurs regularly in any given flea, but the period between inflations may vary between 5 and 80 sec. in different fleas. It is very rarely that a flea fails altogether to show this phenomenon. Out of about 115 fleas examined in the course of the work perhaps half a dozen failed at any time to show it. Its appearance varies, however, with the physiological condition of the flea. In Fig. 1 is shown a young female flea with the tracheae in the inflated (black) and deflated (white) positions. In the male, and in the young female before the ovaries are ripe, the abdomen is under no great internal pressure. The tracheae are nearly round in cross-section, and lie contorted in the body. During deflation the volume of the tracheae diminishes owing to the elasticity of their walls. The curves and loops shorten and straighten; finally, when deflation is nearly complete, the walls in the largest trunks begin in a few places to fall together. On inflation the tracheal walls return to their nearly round crosssection, and the tracheae are again thrown into loops and curves. The slow decrease in length on deflation and sudden increase on inflation are particularly well shown in the leg, where the trachea is fixed by the articulations, and is only able to extend its length between them. The trachea is seen to be suddenly bent on inflation, and slowly to flatten again to a straight position.

In the gravid female flea, when the abdomen is distended with eggs, and presumably under internal pressure, the tracheae are already stretched, and cannot increase further in length. Deflation can only effect a transverse collapse of the tracheae, and a variation in length hardly takes place. The tracheae flatten until they are almost empty of air; upon inflation they return to their original oval crosssection. This type is shown in Fig. 2 which also shows the lengthening and shortening described above. The length of the sections between a–b and c–d should be compared in the two figures. During deflation the epithelium wrinkles and thickens itself over the tracheae. When they are again inflated this sheath is stretched. A third type is seen in old female fleas, where the tracheae appear to suffer from a type of sclerosis; the main trunks, irregularly darkened in appearance, remain motionless, and quite unaffected by inflation, which is taken up by the subsidiary tracheae, which are normally less responsive. When the main trunks cannot expand and collapse, the side branches are thrown into exaggerated curves.

At the moment of inflation the fluid in the tracheoles retreats towards their fine endings and during deflation slowly rises again, as in Fig. 3, which also shows the associated lengthening and shortening of thé tracheae. It has been shown by Wigglesworth (1935) that an atmosphere of oxygen causes a larger amount of fluid to pass into the tracheoles, by oxidizing the metabolites in the body fluid and so reducing its osmotic pressure. In an atmosphere of pure oxygen the meniscus of the fluid in the tracheoles rises to easily visible limits, and its movements can be studied under the high power. The tracheoles only respond to inflation and deflation if they are very short, and near actively inflated tracheae.

The tracheal pulsation is shown over the whole body, from the prothorax to the end of the abdomen, but is most obvious round the middle abdominal tracheae. In the thorax it is usually obscured by the thick body wall.

If the deflation begins to be visible soon after inflation has taken place, it has time to become deep before the next inflation ; if it only becomes visible at the end of the period it remains shallow. The length of the rhythmic period varies with the physiological condition of the flea. Starvation always has the effect of retarding tracheal rhythm. The quickest rhythms are found among gravid, fully fed females in the summer, when rhythms of 5 sec. are common ; but when the host has become torpid the fleas give a very much slower rhythm. The tracheal rhythm is inhibited by struggling, or any form of shock, such as wounding. This inhibition may last half an hour after pursuing the flea and placing it in the observation cage. During the motionless period the tracheae are deflated ; it is only after their gradual inflation that the rhythm can be expected to begin again. The period of inhibition can be avoided by running through the cage a slow current of oxygen, which almost invariably produces a rhythm within 2 min., and is useful in annulling the effects of amputation, etc.

This relation was shown to be extremely close, though the form varied from case to case. The flea could never be watched as a whole, under sufficient magnification, to see all the spiracles functioning together; it was possible to watch Th. ii, Abd. i and the tracheal rhythm together; or, immediately afterwards, Abd. viii and the tracheal rhythm, and to compare the results (Figs. 4, 5).

The bnly constant relation is that of Abd. i to the tracheal rhythm ; this spiracle always opens at the moment of inflation. Th. ii is variable: it may remain closed, or open rhythmically, for a short time, usually together with Abd. i but sometimes half-way through the beat. Abd. viii, the remaining large spiracle, is also rather variable. Its period also accords with the tracheal rhythm, and it usually opens on the inflation, but it may, as in Fig. 5 be distinctly late (Th. ii was in that case permanently closed). Charts are also given of the movements of Abd. vi and vii, which seem to play only a small part here. These charts, showing spiracular movement, were all made in an oxygen stream. In no case is the open period of a spiracle shown to be longer than half the period of one tracheal pulsation, and usually it is much less. All the spiracles are closed during the last third of a pulsation.

Deflation of the tracheae does not appear to begin until Th. ii and Abd. i are shut. It may appear very faintly just before Abd. viii closes, if its opening is very narrow, but is usually delayed until all three are closed. If, while all spiracles are closed, the flea be stimulated to struggle, so that Th. ii opens for a second, the tracheae are momentarily inflated. Almost immediately they return to the degree of deflation which they had previously reached. The deflation of the tracheae is always delayed by struggling until after the’ anterior spiracles have closed. During the period of immobility between placing the flea under observation and the commencement of the tracheal rhythm the movements of the spiracles are rapid and irregular, and are often independent. In changing to oxygen, a flea will always show a tracheal rhythm at once, with a corresponding retardation of spiracular movement usually to about 50 sec. between openings.

The effect of CO₂ in any concentration from 50 % in air (which produces complete coma) to 6·5 % in air, is to stop the tracheal rhythm immediately. If a female flea showing a tracheal rhythm of 11 sec. in air is subjected to a stream of 12 % CO₂ for 5 min. (four experiments) it needs 13 min. for the renewed air current to produce the tracheal rhythm again. This begins deeply and irregularly (e.g. 20–10–15–20 sec.) and in about 20 min. returns to the normal 11 sec. From weaker concentrations (five experiments) recovery takes Jess time, or may be almost immediate. The period of the tracheal rhythm, when it begins again, is always slower than normal : e.g. before exposure to 6·5 % CO₂ in air a flea gave a tracheal rhythm of 20–20–20–19 sec. The stream of CO₂ was then run through the cage for 5 min., stopping the tracheal rhythm at once. The air current was then renewed. 40 sec. later the tracheal rhythm began again at 25–25–23–23 sec. One flea, treated with 6·5 % CO₂, gave pulsations while in the gas, as follows :

• In air current : 16–16–16–17–17–16–17 sec.

• During 5 min. 0f 6·5% CO₂: Initial pause for 5 min., then spasmodic deep contractions: 40–24–70–40–30 sec.

• Air readmitted: 23–30–297–28–28–25–28–25–35 sec.

In 1·5 % CO₂ (five experiments) the tracheal rhythm is hardly affected.

Fleas were also observed in oxygen-nitrogen mixtures where the oxygen proportion was always greater than that in air. This has a stimulating effect on the tracheal rhythm, and will produce it when no rhythm has previously existed in air. The effects, in differing degrees, are the same for 36 % oxygen (six experiments), 75 % (one experiment) and 100 % (which was always afterwards used in routine work). The rhythm immediately becomes deeper and more obvious, and in pure oxygen distinctly slower.

In an attempt to estimate the control, if any, exercised on the tracheal rhythm by the central nervous system, parts of the chain of thoracic ganglia were destroyed by cutting. The flea was first conditioned for about an hour in a stream of pure oxygen, otherwise no tracheal rhythm occurred. The required ganglion was then cut through, and the flea was returned to the observation cage, and a stream of oxygen was passed through. The cutting was done from the ventral side with a sharp needle, and the wound was sealed off with vaseline.

Upon cutting through the prothoracic ganglion of a male flea with a tracheal rhythm of about 40 sec., the tracheae ceased to show any rhythm for about 15 min., after which there was a fairly marked rhythm of about 60 sec. Four fleas gave similar results.

In three further fleas the mesothoracic ganglion was destroyed. This had the effect of stopping the tracheal rhythm promptly and permanently ; cutting through the metathoracic ganglion had the same effect, though the retarding effects on the heart and gut were less marked than when the mesothoracic ganglion was cut.

1. A fairly large window (about half of one side of an abdominal stemite) was cut in segment 7 of an oxygen-conditioned flea. For about 4 min. the previous tracheal rhythm was suspended. It then began again strongly. There was no movement of the body fluid meniscus in the wound. On the other hand, when a flea was taken that showed marked abdominal contractions, a wound made in the prothorax showed very marked rise and fall of the body fluid meniscus. A tracheal rhythm which was operating at the same time with a different period had no effect at all on the level of the meniscus.

2. The posterior pair of legs were cut off through the proximal quarter of the femur, so that the large tracheal trunks were open to the atmosphere. The tracheal rhythm ceased at once, permanently.

3. A flea was subjected to reduced pressure, having been previously conditioned in oxygen for 1 hr., during which time it showed a strong 10 sec. rhythm. The chamber was then slowly evacuated to 23 cm. of mercury, taking 1 min. over the process. This vacuum was held for about 45 sec. and then released. The tracheal rhythm continued during evacuation, but with each beat the inflation was weaker, until at maximum evacuation the tracheae were permanently deflated. The tracheal rhythm appeared to have ceased at about 15 cm. of mercury. On releasing the vacuum, all tracheae immediately sprang back to the position of maximum inflation, and the tracheal rhythm was resumed.

Three explanations have been proposed for this tracheal activity. The first, and most obvious, is compression by contraction of the abdomen, or by pulsation of the heart or gut, whereby pressure would be transferred to the body fluids, and find a visible response in the tracheae as the only compressible parts of the body. Comparison of the respective rhythms in Fig. 6 rules out the influence of the heart or gut, but compression is undoubtedly produced at times by the rhythmic body contractions; these have, however, a quick rhythm, which may act synchronously with the tracheal rhythm but independently from it. It is also quite impossible ever to detect any sign of abdominal contraction or body fluid flow connected with the tracheal rhythm. The experiment on p. 334 showed that the meniscus in the wound, though responsive to abdominal contractions, was absolutely unaffected by the tracheal rhythm.

The second explanation attributes the tracheal rhythm to intrinsic contractions of the epithelial sheath surrounding all parts of the tracheal system. It is assumed that the tracheal epithelial system is innervated from the central nervous system, particularly from the mesothoracic ganglion (p. 334). Under this theory the initial pause in the tracheal rhythm after capture, wounding, etc., can be attributed to “shock”. Deflation, i.e. flattening, is produced by epithelial contraction, with accompanying thickening of the epithelium (p. 330), inflation by the release of this contraction which allows the natural spring of the walls to bring them back to their original positions and at the same time to suck in air by the principal spiracles which have opened at the moment required. However, it is difficult to imagine any sort of contraction that would flatten the oval tracheae as well as shortening them. Any such transverse or longitudinal contraction should have the mechanical effect of making the tracheae round in cross-section. The theory can, however, account for the rise of water in the tracheoles during deflation and its retreat on inflation by assuming that the epithelial sheath contracts equally over fluid-and air-containing sections alike. The fluid, not being compressible, must therefore rise in the tracheoles, even while the gas pressure is being increased.

The third explanation is “a collapse resulting from the disappearance of gas from the tracheae”, thus generating a negative pressure within the system. If reference is made to the relation between spiracular opening and tracheal rhythm (p. 332) it is found to support this theory quite well. The remarkable features are: (1) the dependence of inflation on Abd. 1 whose opening is apparently sufficient to inflate the whole system in an instant, when all other spiracles remain closed, and (2) the quite secondary importance of Abd. viii, which looks as large as Abd. i though its opening is seldom as wide. In the case where deflation begins shortly before Abd. viii closes, the brief and narrow air-leak which it provides in an otherwise closed system does not admit enough oxygen to keep up the pressure in the tracheae, which are being drained of oxygen all over the body. That a wide air-leak can inhibit the tracheal rhythm is apparently shown by amputating the hind legs (p. 335).

The effects of CO₂ and of oxygen also fit into this theory (p. 333). The depth and distinctness of the tracheal rhythm seem to depend on the length of the closed period of the spiracular system. In pure oxygen this is long relative to the open period, and gives good opportunity for deep inflation before oxygen want is felt and the spiracles open. Further, in oxygen the whole of the tracheal contents can be utilized by the tissues. Conversely, in CO₂, the spiracles remain open for a greater length of time and negative pressure has less time in which to develop.

The present theory also accounts for the pause in the tracheal rhythm after capture or struggling, as is seen by the very rapid spiracular rhythm on p. 333 which gives insufficient time for negative pressure to develop.

It seems likely that quite a considerable negative pressure could be set up before the resistance of the tracheal walls to collapse was overcome. It is not known how great a pressure is needed to flatten the walls, but in the experiment on evacuation the tracheal rhythm appeared to cease altogether at a vacuum of about 15 cm. of mercury; and although all spiracles were open, the natural spring of the tracheal walls could not hold them apart at this pressure.

It will be remembered that the tracheal rhythm was inhibited by cutting through the mesothoracic ganglion. The cutting through of this ganglion always stimulates Th. ii to continuous, rapid palpitation, and this in itself would be enough to inhibit the tracheal rhythm. Snodgrass (1936) quotes the case of a caterpillar in which the occlusor muscle of the spiracle is innervated from the median nerve trunk that arises from the ventral ganglion of the preceding segment. It has not yet been established whether the flea has a median nerve system; but if so, the mesothoracic ganglion might innervate Abd. i, and thus have a controlling effect on the tracheal rhythm. The temporary inhibition caused by cutting through the prothoracic ganglion can best be attributed to shock, as Th. ii may remain permanently closed after this operation. The reactions of Abd. viii to sectioning of the ganglia were not observed.

There remain to be considered the effects of evacuation. Here it is necessary to explain the permanent collapse of the tracheae with the spiracles fully open (the pressure being reduced both within the tracheae and around the body wall). The body wall, however, may be looked upon as a more or less rigid casing, i.e. it will keep the body fluids within it near the original atmospheric pressure which prevailed at the start of the experiment. Since the tracheae, being open to the air, are under very much reduced pressure, they will naturally collapse under the pressure of the body fluids. When the vacuum is released the resilience of the tracheae brings them back to their original inflated position.

The chief advantage of the last of the three explanations is that it assumes the existence of no special respiratory mechanism in the flea beyond the co-ordinated closing of all the spiracles, and the presence of collapsible walls in the tracheae, which are resilient enough to act as bellows and so suck in air through the narrow openings of the spiracles when such openings are provided. The mechanism may also be considered in the light of Krogh’s (1920) studies on the “diffusion method” of gas exchange between the atmosphere and the tissues via the tracheae. Krogh evolved a formula whereby the adequacy of the oxygen supply diffusing to the tissues depends upon the length and cross-section area of the tracheae concerned. Where the length was too great, or the cross-section too narrow the oxygen supply was jeopardized. In the present case diffusion is aided by the forcible indrawing of the gas by the tracheae released from forcible deflation.

A comparable phenomenon was seen by Babák (1912) in mosquito larvae kept submerged. The tracheae collapsed, somewhat as in the flea, some sections collapsing more than others, giving the appearance of a string of beads. On being refilled with air, the head tracheae were first inflated, but if the larvae were kept too long submerged, the tracheae did not refill at all. Babák attributed this deflation to muscular action, but the appearances that he describes can equally well be explained by differences in the resiliency of different parts of the tracheal walls. Deflation, in the mosquito larvae, took about 10 min. ; in the flea it can take place in under 10 sec.

It is not yet known how generally there occurs among insects a long period in which all the spiracles are effectively closed. Wherever it is found (and it is probably frequent) the phenomena observed in the flea must perforce accompany it. In insects with circular tracheae the decrease in volume will probably be longitudinal, as in young individuals among fleas; only oval tracheae are susceptible to transverse collapse. A comparative study of the tracheal rhythm waits, therefore, upon the adaptation of the technique to insects less accommodating than the flea.