by Academician Vladimir SVIDERSKY, Sechenov Institute of Evolutionary Physiology and Biochemistry, RAS (St. Petersburg)

The flight of insects is one of the most intricate and perfect motor acts in nature. Up till now its evolution presents a lot of enigmas. When has a flight system emerged, how it worked in ancient fliers, and what vital problems they solved due to it? What kind of insects were first fliers on our planet? We believe that today we have answers to such questions.

WINGS... AND LEGS

Insects were first animals on our planet to fly. When did it happen? We cannot say for sure, but paleontolog-ical findings indicate that in the middle Carboniferous (about 320 mln years ago), these arthropods had a well-developed flight system and hence, could fly rather well. And if so, they started flying long before this period. In any case, only 100 mln years later there appeared flying pangolins–pterosaurs (they, by the way, died out in the Cretaceous) and 70 mln years later after that there appeared birds. Thus, a tremendously long period of time insects were the only fliers with all advantages of this unique feature. They used them well enough eventually to become the most numerous and prosperous animal class.

The flight of insects can be regarded from different viewpoints. I prefer a physiological approach. I shall have to start from the main problem, which is still unsolved–the origin of wings in insects. If in other flying animals, wings are modified fore limbs, in insects (according to a majority of scientists), wings have nothing to do with legs. According to the hypothesis most prevalent today, insect wings emerged from lateral folds of thorax tergites and at first served for gliding, and then

Geochronological scale showing the time of flight emergence in flying (and formerly flying) animals and duration of their existence.

acquired a capacity to move, providing a flapping flight. Unfortunately, this fine hypothesis is not persuasive from a physiological viewpoint.

I would not object if it were perfection–through evo-lution-of the gliding capacity. But speaking about the movements of future wings, we have to think primarily about respective muscles. How can a complex set of functions emerge from "nothing" in the upper part of the thorax? How do they "learn" that at this stage of development they have "to reunite" with strictly determined points in the wings? How will very intricate joints of the latter emerge in parallel with this?

To do justice to the so-called "leg" hypothesis, we must mention that this hypothesis is attractive for physiologists. According to it, wings in insects could "grow" from mobile lobes on the legs of ancient crustaceans. This hypothesis not only cancels the problem of mobility development of would-be wings, but it also proves that these fliers do not belong to a "declassed group", in which the origination of wings was different from other animals, which started flying (from legs). And, finally, this hypothesis is in line with a "crustacean" theory of the emergence of insects, propagated in recent years by specialists in molecular genetics. Physiologists advocate probable "kinship" of their wings and legs; they have been long studying the so-called bifunction-al muscles, used for walking and also during flights involved in the work of the flight system. The findings of scientists from our laboratory indicate that in dragon-flies, capturing spoils during flights, wings and legs work as a single functional system. Nevertheless, some specialists think that this hypothesis is not yet sufficiently convincing and there is still much to be done. Hence, the problem of origination of wings in insects remains unsolved.

And now, let us put these problems aside–for some time–and imagine that insects eventually did fly. What are they–first fliers on our planet, how they lived, and why did they die out? Unfortunately, there is some mis-understanding in this sphere (I shall call it myths), presented in many books, including scientific ones.

FOUR MYTHS

Myth number one. Dragonflies–progenitors of modern species–were the insects that inhabited the Earth during the Carboniferous. Their outward similarity brings us to this conclusion. However, the recent phylo-genetic studies indicate that ancient dragonflies should be referred to other orders (Protodonata, Palaeodicty-optera, Geroptera, etc.), though they have common roots with modern Odonata, and belong to other lines of development. Besides, the Protodonata were typical predators, Palaeodictyoptera were peaceful plant-eaters, while food preferences of Geroptera are not known. As for the ancient Odonata order, it emerged much later and is referred to the Permian (less than 300 mln years ago).

Myth number two. The ancient insects of the Carboniferous were giants (the span of wings in Meganeura monyi "champion" made up 71 cm!). True, there were some large insects, but middle-sized and even smaller species (smaller than modern dragonflies) predominated. The same is true of flying pangolins, which started flying during the Trias-sic. They included giants with more than 10 m wing span and tiny ones, no bigger than a modern sparrow.

Myth number three. Giant insects appeared as oxygen concentration in the atmosphere of the Carboniferous was very high, reaching 35 percent (21 percent today). The logic is as follows: respiration in insects is realized by trachea (not by lungs). They breathe in oxygen and release carbon dioxide at the expense of diffusion due to different partial pressure values at the inner and outer terminals of the tracheal tube. The higher oxygen content outside the body, the deeper it penetrates into the trachea, the longer they can be, and hence, the larger the insects are. Though the predominance of insects-not giants during the Carboniferous contradicts this myth, I shall dwell on respiration in more details, as it is essential for physiology of these arthropods.

The matter is (and it has been shown long before) that respiration in these insects is not at all diffuse. They breathe actively, with inspirations and exhalations, the respiration rate depending on exercise. This process is supported by the muscles of the trunk (mainly the abdomen) and spiracles. The former are innervated from the central, the latter from the vegetative nervous system. It is interesting that the respiration process during flights is facilitated in insects, as contractions and relaxations of wing muscles promote better ventilation of the tracheal system. Thus, the "oxygen hypothesis" cannot be applied to insects. To say nothing about other giant animals of the Carboniferous, for example, huge reptiles–sure, we cannot "suspect" them of tracheal diffuse respiration.

Myth number four. Giant insects died out as oxygen concentration in the atmosphere reduced and they had problems with respiration. However, this was the fate of almost all dragonflies only at the end of the Permian, while the concentration of atmospheric oxygen sharply reduced already by the end of the Carboniferous, and, the Geroptera insects, on the contrary, had disappeared even earlier, during the middle Carboniferous, when oxygen concentration in the atmosphere was the highest. So, the cause of dying out of dragonflies was different. What was it? We do not know yet, but let us point out other extinct giants, living during a different period: dinosaurs. Like ancient dragonflies, dinosaurs (I mean primarily predators here) were in a similar position: dinos dominated on the ground, dragonflies–in the air.

So, why have dinosaurs died out? Some paleontologists think that one of the causes could be rather prosaic: incapacity to save their progeny from new groups of animals, less vigorous, but more dexterous, for example, mammals. The matter is that huge dinosaurs laid rather small eggs (strength and porous structure of the shell limited their size) and did not hide them deep in the soil.

We have evidence that ancient dragonfly larvae lived not in water (as modern dragonflies do), but on the ground and became victims of predators appearing with evolution. This is what happened to dinosaurs. However, dragonflies found a way out from this seemingly hopeless situation: they protected their progeny by "hiding" the larvae into the water of calm springs and small lakes, where they were safe. We cannot say for sure whether this was so or otherwise, but the fact is that dragonflies with "water" larvae have survived up till now.

FLIGHT MASTERS

The apparent similarity in the structure and arrangement of wings, structure of the wing-bearing segment (the thorax) of extinct and modern dragonflies give us grounds to assert (with greater or lesser probability) that the work of the flight system could be largely similar in

Insect thorax skeleton structure (transverse section).

both. However, we think that we can learn some important details about its structure in ancient insects. How?

What we know about the flight system of modern dragonflies today. Our basic assumption is that all its functions are essential (otherwise they would not exist). And if a function does exist, there must be a structure supporting it. The question arises: What was essential for ancient fliers and what was absolutely superfluous at that time? In other words, let us try to understand everyday needs of ancient dragonflies. And then we will see what was different in the structure of their flight system as compared with that of modern, evolutionarily "advanced" dragonflies. We can tentatively call this approach paleontological-functional or, if you like it better, paleophysiological.

What are the potentialities of modern dragonflies, for example, Aeschm grandis–vigorous insects with remarkable maneuverability, belonging to the Heteroptera suborder, resembling their ancestors (particularly from the Protodonata order) more than other ones? After a detailed analysis of the position and functional potentialities of Aeschna grandis wing muscles, we revealed mechanisms due to which they perform accurate and rapid movements in flight (this study was carried out in collaboration with Svetlana Plotnikova, morphologist, Dr. Sc. (Biol.), and Valery Gorelkin, physiologist, Cand. Sc. (Biol.)). But first we shall dwell on the character of motor behavior of these creatures.

These extremely active predators catch mosquitoes, flies, horseflies in flight, and also larger insects with their legs. Due to their immense "appetite", they can spend hours in the air, trying to catch something. Their maneuvers in the air are truly marvelous: they can abruptly start or stop the flight, hover in the air, shoot upwards,   smoothly  glide,   fly  sideways,   with  the abdomen forward. It is remarkable that even turns they perform in two ways: drawing an arch (like other flying insects and birds) in rapid flight and, if they move slowly or hover over a certain place, they can turn their body by more than 90° making just two wing flaps! It is not a mere chance that many scientists consider dragonflies the greatest maneuver masters on the Earth. I think it is needless to assert that these exclusive abilities help them to capture not only the quickest insects, but also to escape from birds.

In order to better understand the mechanisms due to which they realize all these complex movements in the air, let us discuss the structure of their flight system. The skeleton of each segment of their bodies is formed by four main sclerites (shells): upper (tergite), lower (ster-nite), and lateral (pleurites). The thorax bearing the wings consists of three segments: anterior, middle, and posterior thorax. The pleurite is well developed with the pleural column located on it. It is a wing process, supporting wing movements. The wings are located in the 2nd and 3rd segments of the thorax between tergite and pleurites.

And now about wing muscles. From the functional viewpoint, they constitute two main groups: "motor" muscles, providing forward motion, and "steering" ones, making possible maneuvers in flight. In addition, the insect flight system includes the muscles involved in regulation of the elastic emission of the thorax compartment and in modification of the relative position of its mobile parts. The bifunctional muscles mentioned above can also be involved.

An insect wing is a two-arm lever with different upper arm lengths, and the pleural column serves as a supporting point for it. The wing can be made to move up and down by two principally different ways, and both (as we

Principle of work of direct-action muscles, providing up-and-down movements of the wing plate.

now know) are realized in the evolution of insects. The first way: alternative contractions of the muscles attached to the wing on both sides of the pleural column. In this case, the wing will move downwards from outside, from inside–upwards. The muscles cut upon the wing directly and hence, they are called direct-action muscles. The second way: the wing moves with the help of tergite, which is set to motion by the muscles not attached to the wing–indirect-action muscles. There exist two groups of these muscles. During contraction of tergosternal muscles, tergite goes down and "acts upon" the wing base near the pleural column, causing an upward movement of the wing plate. Contractions of longitudinal spinal muscles lead to bending of tergite upwards, resulting in downward movement of the plate.

Nine main muscles support the work of each of the four wings of the dragonfly. There is no strict division (as, say, in flies) into wing motor and steering muscles, and some of them serve for several purposes. For example, the downward flap of the wing is realized by 5 muscles, 2 of them are simple depressors and 3 can realize also steering functions. Three "simple" and one "plural action" muscles are involved in the upward movement of the wing. The longitudinal dorsal muscles are not developed in the dragonfly, and all depressors (muscles lowering the wing) are direct-action muscles. Lifting of the wing is "realized" by two indirect-action (tergosternal) muscles. Rather substantial resources are involved in realization of the forward motion in these fliers. The potent wing motor generates flying velocity of up to 96 km/h! It is interesting, that the tractive force in the dragonfly flight system is created during wing movements not only downwards, but also upwards.

However, they are known mostly for maneuvers, not velocity. And this is due to the plural-action muscles with the steering functions and "purely" steering muscles. During a straightforward flapping flight, they can turn wings round a longitudinal axis, changing their attack angle and hence, the lifting force and tractive force. But during a straightforward flight, these muscles function symmetrically on the right and on the left and hence, the trajectory of the flight is not "twisted", though its velocity can be changed, and rather essentially. Turning and other maneuvers are different: acting asymmetrically on the right and on the left, the steering muscles enable the insect perform rather complex maneuvers. All four wings of the dragonfly function independently as concerns phase, frequency, and amplitude of flaps. I have not yet mentioned that their wing muscles, even performing similar functions, differ by "size" and hence, by potency, due to which the flap force can be precisely "dosed".

Let us add that dragonflies move their wings not only up and down, change the attack angle, etc., but they can move them forward and backward in the horizontal plane irrespective of each other. The wings draw together or part from each other, continuing their "main" work. The aforesaid potentialities make these insects true masters. Could the ancient Protodonata predators or phytophagous Paleodictyoptera compete in anything with modern dragonflies?

WHY DRAGONFLIES NEVER FOLD THEIR WINGS?

According to the hypothesis, proposed in 1976 by Alexander Rasnitsyn, Dr. Sc. (Biol.), from A. Borisyak Paleontological Institute of the Russian Academy of

Principle of work of indirect-action muscles for the wing (at the expense of tergite movements). Dotted line shows the position of tergite, preceding contraction of tergosternal muscles (A) and longitudinal dorsal muscles (B). Contracting muscles are marked in red, relaxing ones are in pink.

Sciences, the ancestors of the herbivorous forms, phytophages, were the first to explore aerial environment. In order to escape their enemies–spiders–they often had to climb tall trees and then jump down, and to do all this properly, they had to learn first to glide and later on to fly. This idea seems quite probable, but we can also add: these insects could also jump from the trees for rapid change of their dislocation. It would be therefore logical to confine ourselves to an analysis of a probable structure of the flight system of first fliers–herbivorous forms. But we should not "forget" dragonfly-like predator insects, as they were still the first in a way. And I shall start–contrary to logics–from predators. What were their main problems in the distant Carboniferous?

One of the basic problems was feeding. The dragonflylike phytophages were already flying (for example, dicteoneurides), which presumably served as the main food for dragonfly-like predators. A bit later there appeared one-day flies and some of the Dictyoptera order in the air. In order to catch them and eat, our "heroes" did not need high flying speed. Thus, their wing motor muscles were most likely underdeveloped.

But, perhaps, they needed high speed to get away from enemies? They could get rid of those spiders just by flying from one place to another. Much more dangerous were reptiles. They deluded dragonflies by their initial immobility and then, when the insects turned out to be nearby, rapidly captured the gaping prey. But even here, ancient insects had to do just one thing–maximally rapidly fly up to the air and then "calmly", without hurry, escape the danger. However, they often had to fly under difficult conditions: in thickets, between trees. Thus, it was vitally important to be able to back up, turn around, fly sideways. Hence, their steering muscles were to be sufficiently well developed, and only the dragonfly flight system organization allowed to maneuver.

And now I should like to touch upon an important problem: why dragonflies, in contrast to many other flying insects, "failed to learn" (as it is often called) to fold wings along the trunk? According to textbooks, this capacity gave tremendous advantages in the struggle for life. By making the body more compact, insects could live in new, heretofore inaccessible places: in fissures, under stones, in the thickness of grass, etc. At the same

time they acquired extra potentialities to hide from enemies. And this is correct.

But I think that dragonflies did not need these potentialities: their "mentality" was quite different: they did not have to escape and hide, quite the contrary–other insects had to run/fly and hide from them more often! In order to escape voracious reptiles, dragonflies had to fly up to the air instantaneously. And their wings, always "ready" to fly up, were just what they needed. That is why it is more correct, I think, to say that these insects (later on our dragonflies) learned not to fold their wings for their safety, than to say that they failed to learn to fold their wings.

The question arises: Why had modern dragonflies failed to learn to fold their wings, though reptiles no longer threatened them? First, they became on average smaller and hence, more compact and more mobile than their predecessors. And secondly (and it is most important), they used their inability to fold wings as an advantage in new conditions too. For example, the dangers associated with the need to fly up rapidly persist up to the present time. The persecutor dragonflies spend day hours in the air, and their wings are always in a working state. The watcher dragonflies have a different life strategy (they do not fly for hours in search of prey, sitting in grass), use their spread wings for rapid flying up to capture a victim (for eating, not for safety). Thus, spread wings are essential for dragonflies even today. As for their movements on the ground, they are quite unfit for it, and their legs have other functions (for capturing the prey in flight, clutching to plants when "landing", for clearing the trunk).

And now a few words about probable structural and functional organization of the flight system in ancient phytophages. As they were the first insects to fly, it was to be at first better developed than in predator insects, "lagging behind" in conquering aerial environment. I.e., phytophages at first had certain advantages in the predator-victim pair. Their life at that time was rather easy: enough food, and in case of its absence, they could fly to another tree. They could escape from enemies (spiders) by just fluttering to a neighboring branch. In other words, they had no special stimuli for improvement of their flying capacities, and hence, the flight progressed slowly in phytophages.

A different thing is flight for dragonflies. If we say that for a wolf "walking is eating", we have to admit that these ancient insects could eat only due to their wings. No wonder that eventually their flight system became "technically" superior to its analogue in phytophages.

That is why when I read that the phytophages of the Carboniferous were excellent fliers, I think that it could not be so: at first they could not fly as they were just "mastering" this skill, and later on they could fly rather well in case of necessity–to avoid an "undesired" meeting with predators during flights from one tree to another. Generally they were passive hiding in the crowns of trees.

But then it turns out that in some cases the wings of ancient phytophage insects could be a nuisance. Perhaps, they folded them, not along the trunk, but vertically above the back, as modern Homoptera dragonflies? This hypothesis is indirectly supported by a "frightening" picture on their wings, which could scare predators. And it was to be shown to the enemy suddenly–presumably, by spreading the wings.

And the last intriguing circumstance. The ancient phytophages had two lobes with tendons on the anterior part of the thorax on the right and left, suggesting an idea that these insects could have six wings some time in evolution. However, paleoentomologists claim that these lobes were never mobile. Perhaps, they served as a sort of ram, which helped these animals "squeeze" through plant barriers, sparing the wings?

Thus, I have dwelt here on some aspects of life and behavior of ancient carnivorous and herbivorous dragonflies, but even these notes give us an insight into the structural and functional organization of their flight system. By using neurophysiological methods, it is possible to more precisely evaluate the functional contribution of each muscle in the work of the flying system of modern dragonflies and try to use these data for deciphering the potentialities and structure of such mechanism in ancient insects. In other words, to obtain a more detailed picture of the flight system development in ancient fliers. But this is a task of future research.