Insect physiology

Caterpillars have lungs

Article metrics

For the past 350 years1, it has been the perceived wisdom that air is distributed to insect tissues through a series of branched tubes called tracheae and tracheoles, the main branches of the tracheae opening to the outside world at a series of segmentally arranged openings called spiracles. But a paper in the latest issue of the Journal of Insect Physiology2 gives us cause for a rethink. After a painstaking examination of over 2,000 electron micrographs, Michael Locke has challenged the universality of this ‘rule’ and concluded that many caterpillars possess lungs, designed to provide blood cells (haemocytes) with oxygen.

Figure 1


Alice confronts one caterpillar that clearly has lungs.

The tracheal cuticle (with its underlying epithelium) is continuous with that of the insect's body surface, and the flow of air in and out of the spiracles is often controlled by valves3,4. The tracheae themselves are continuous with the cuticular tubes of the intracellular tracheoles, which form branches that ramify throughout the tissues. The respiratory needs of the tissues are met by diffusion within the lumen of this tubular system, and by the exchange of oxygen and carbon dioxide across the thin walls of the tracheoles5,6. This mechanism applies to most tissues.

However, the pair of spiracles closest to the rear of many caterpillars — including the Brazilian skipper butterfly, Calpodes ethlius — are much larger than the others. Moreover, one of the tracheae from each of these spiracles gives rise to a branched tuft of very fine and thin-walled tracheal branches and tracheoles (Fig. 2, overleaf). The tracheae that give rise to these tufts are termed aerating tracheae to distinguish them from the other, conducting, tracheae. The tip of the tuft is anchored in the heart muscle, but the tracheoles do not ramify amongst the cells as in other tissues. Furthermore, much of the tuft is free in the haemolymph, and the tuft moves as the heart contracts and relaxes2.

Figure 2: The tracheal system in the last three segments of a live caterpillar.


Respiration in insects occurs by gaseous diffusion along the lumen of tracheal tubes that branch repeatedly. Blood cells lack their own permanent tracheae, but alight intermittently on a pair of specialized, thin-walled tufts that act like lungs.

Amazingly, although the existence and movement of these tufts has been known for more than 100 years7, only now has the significance of this been recognized2. Locke has found that the haemolymph becomes oxygenated as it passes through the tufts, which waft it towards the openings into the heart. Even more significantly, he has shown that the tufts provide way-stations for haemocytes to replenish their oxygen supplies.

Insect haemolymph contains over seven million haemocytes per millilitre in the C. ethlius caterpillar2, and there are several types of haemocyte. These engulf or encapsulate foreign material (including invading organisms)8, and may also be important in storing and distributing nutrients and/or maintaining sugar and amino-acid levels8,9. In most insects, the haemolymph is not involved in oxygen transport, and haemocytes were thought to be able to operate in the absence of oxygen. But this has now been called into question2.

Many of the haemocytes are loosely attached to tissue surfaces8 and, in C. ethlius, oxygen deficiency leads to a 60 per cent rise in the number in circulation to nearly 12 million per millilitre of haemolymph. Locke studied the structure of one type of haemocyte — the granulocyte. In a normal, well-oxygenated environment, granulocytes have irregular outlines, with processes (filopodia) on their surface. Their Golgi complexes are well developed and their secretory vesicles contain tubular structures. Under conditions of oxygen starvation, however, they become rounded and lose their filopodia, their Golgi complexes become smaller, and few (if any) tubular structures are seen in their vesicles2. Although there are always granulocytes trapped temporarily in the branches of the tracheal tufts, these become more abundant under oxygen starvation (Fig. 3a). Moreover, the granulocytes in the tufts have the characteristics of those in a well-oxygenated environment (Fig. 3b). This is a strong indication that the granulocytes are, indeed, coming to the tracheal tufts for oxygen2.

Figure 3: The branched tufts of tracheae and tracheoles that provide blood cells (haemocytes) with oxygen in caterpillars of the butterfly Calpodes ethlius.

a, The number of haemocytes (red) in a tuft increases when a caterpillar is subjected to oxygen starvation. b, Oxygen-starved granulocytes (a type of haemocyte; blue) entering a tuft resume the characteristics of those in a well-oxygenated environment (red). c, In the tokus — a ‘lung’-like compartment — the haemocytes become closely apposed to the thin-walled tracheae and tracheoles.

But the story does not finish here. The tracheal branch that gives rise to the tufts also passes backwards into a semi-isolated compartment at the rear end of the animal. Here, again, it becomes thin-walled, giving rise to tracheoles that end freely in the haemolymph and often run back along the sides of the tracheae to form knots. This compartment is called the tokus, and it acts as a ‘lung’ for the haemocytes. Haemolymph that has passed over the tufts periodically enters the tokus through openings alongside the tracheae. The granulocytes (and the other haemocytes) become associated with the free tracheal branches and tracheoles, but in a closer and longer-lasting fashion than in the tufts. The surface of the haemocyte that touches a tracheole becomes flattened to form a large contact area (Fig. 3c), increasing oxygenation of the haemocyte. Because the tokus is close to the heart, the aerated haemolymph can enter it immediately for recirculation2.

Locke's work is particularly exciting because it is not just a one-off phenomenon — indeed, the tufts and the tokus lung tracheae are found in caterpillars from all 13 families of lepidopterans that have been studied so far2. And Locke has clearly shown that, “although as a rule insect tracheae go to tissues ⃛. haemocytes go to tracheae”2.


  1. 1

    Swammerdam, J. The Book of Nature; or, The History of Insects (John Hill, London, 1637-1680).

  2. 2

    Locke, M. J. Insect Physiol. 44, 1–20 (1998).

  3. 3

    McArthur, J. M. J. Exp. Zool. 53, 117–128 (1929).

  4. 4

    Miller, P. L. J. Exp. Biol. 37, 237–263 (1960).

  5. 5

    Krogh, A. Pflüg. Archiv. Ges. Physiol. Menschen Tiere 179, 95–120 (1920).

  6. 6

    Wigglesworth, V. B. in Advances in Insect Physiology (eds Berridge, M. J., Treherne, J. E. & Wigglesworth, V. B.) 85-148 (Academic, London, 1983).

  7. 7

    King, H. S. Psyche III, 322-324 (1880).

  8. 8

    Gillott, C. Entomology 2nd edn (Plenum, New York, 1995).

  9. 9

    Crossley, A. C. Adv. Insect Physiol. 11, 117–221 (1975).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.