THE MALE housefly is one of the greatest living contortionists. Before it can mate, it must twist its tail end a full 360掳 about its axis. This twisting, or torsion, of the abdomen 鈥� accomplished by a complex system of muscles 鈥� causes the internal reproductive organs to loop around the gut, and at the same time twists and deforms the fly鈥檚 external body segments. More remarkable still is the fact that almost all males of the more than 100 000 species of flies perform these contortions, either through the full 360掳 or a less drastic 180掳 twist. How did a system that so distorts the body, both inside and out, ever arise?
Longitudinal torsion is an unusual phenomenon in animals. Spiralling, as in snail shells, antelope horns and the narwhal鈥檚 tusk, should not be confused with true torsion. In these cases the structure is not twisted about its axis, but is wrapped in ever decreasing circles around an imaginary core. In the snail, the tubular organs, such as the gut and nerve cord, are not twisted but retain their positions relative to each other.
Advertisement
If an animal鈥檚 organs were twisted, their function might be blocked or constricted, so torsion is rare in the animal kingdom. Indeed, twisting could only ever arise if it conferred some advantage on the animal. The classic example of torsion, described in almost every invertebrate zoology text, is the veliger larva of marine gastropods. During this larval stage, the body rotates through 180掳, which brings the mouth and anus close together (they are initially on opposite sides of the body). What advantage this torsion confers on the veliger is unclear, however.
By comparison, males of almost all flies, or Diptera, twist the rear segments of their abdomen at least 180掳 to mate, and more advanced flies rotate their abdomens a full 360掳. Even more remarkable, the full rotation has evolved independently in two separate lineages of flies. Although zoologists are not sure why torsion occurs in veliger larvae, there is little doubt that in dipterans it evolved to accommodate changes in mating behaviour. The degree of twisting is directly influenced by changes in mating habits. Most primitive flies mate on the wing in swarms; more advanced flies mate on some surface, such as a leaf, mud, rotten fruit or dung, where they often claim a territory or perform courtship displays.
A German morphologist first noted torsion of the male abdomen in 1897: he recorded the looping of the male ejaculatory duct around the gut in a blowfly. Although entomologists who studied flies were well aware of the external distortion of the tail segments of male flies, this looping of organs provided strong circumstantial evidence for an unusual internal twisting in the rear part of the abdomen. 杏吧原创s confirmed this when they found that in male blowflies the rear segments (including the genitalia) rotate before the fly emerges from its pupal skin. This 360掳 rotation of the genitalia is called circumversion. Students of more primitive flies, such as mosquitoes, also observed torsion, but in these cases the genitalia rotated only 180掳, a twisting known as inversion.
The abdomen of a fly is a tubular structure composed of 11 segments which telescope. In the male, the ninth to eleventh segments are fused to form a specialised genital 鈥榗apsule鈥� with appendages for clasping the female and transmitting sperm. Each segment in front of the genitalia has a hard but flexible plate above it, called a tergite, and one below it, called a sternite. Torsion is usually clockwise around the body鈥檚 long axis and normally only affects the genital capsule and two, or sometimes three, segments in front of it. Internally, in an untwisted state (as in most insects and all female flies), the gut lies in the upper part of the body and the reproductive tract lies beneath it, towards the insect鈥檚 belly. The two tracts run parallel along the abdomen. Rotation during torsion causes the reproductive tract to loop around the gut and also drags along the insect鈥檚 respiratory tubes, or trachea, and nerves.
While the results of rotation are easy to see when we look closely, working out how it happens is trickier. Insects have no internal skeleton. They rely on hydrostatic pressure and a combination of circular and longitudinal muscles to extend and retract the abdomen. When they contract the circular muscles, the body fluids push out the telescoped abdominal segments; when they contract their longitudinal muscles, the segments retract. Hans-Jurgen Dordel, of the University of Cologne, West Germany, showed in the early 1970s how torsion occurs in a chironomid midge, which rotates its rear through 180掳. He found that immediately after the midge emerges from the pupal skin, specialised oblique and circular muscles rotate the genitalia while at the same time shortening the abdomen by a third. The hard plates of the sixth, seventh and eighth segments become asymmetric, the direct result of being dragged during rotation. In other primitive Diptera, the male genitalia remain untwisted but the fly may rotate them temporarily up to 180掳 during mating, after which they resume their original position. Additional sets of muscles accomplish the second 180掳 twist found in advanced flies which undergo the full 360掳 circumversion.
Why flies should go through such contortions is a far more complex question. There can be no doubt that torsion of the male genitalia is directly related to their mating behaviour, and observations of mating flies support this. All flies couple with the genitalia in what is known as inverse interlock, so that the top of the male genitalia must join the bottom of the female genitalia. This mating position probably originated when insects mated on the wing, but it has been maintained in all flies, regardless of subsequent changes in their mating behaviour. Closely related groups of insects such as fleas (Siphonaptera) and scorpionflies (Mecoptera) couple in this way.
In 1922, the English entomologist Charles Lamb proposed a model for dipteran mating. Essentially, the manner in which male and female flies approach each other and start to mate is not always the same as the way they finish. Because of the awkwardness of the initial position, during which the flies must establish the inverse interlock, the pair often adopts a secondary or final position to complete the often lengthy period of coupling and transferring sperm.
In most flies belonging to the suborders Nematocera (such as crane flies, midges, blackflies and mosquitoes) and Orthorrhapha (such as horseflies, bee flies and robber flies), the male genitalia is inverted, either permanently after emerging from the pupa, or temporarily during mating. Sometimes members of the same family adopt quite different mating positions and with different degrees of torsion. For example, in the robber fly family, the flies belonging to some genera twist through 180掳 when necessary, others are permanently twisted through 180掳, and some twist only 90掳 while coupling. In this family no fixed pattern has developed. This, along with evidence from other families, suggests that inversion may have arisen independently several times.
In higher flies (the suborder Cyclorrhapha 鈥� including houseflies, fruit flies and blowflies) the ancestral inversion is supplemented by an additional 180掳 rotation, producing a full 360掳 twist. How does such a complete twist differ from the unrotated state? After all, twisting through 360掳 brings the genital capsule back to the same position it was in before rotation. However, consider those long sausage-shaped party balloons, which after twisting can be bent into various shapes. The unrotated abdomen is like one of these balloons, but held rigid by hydrostatic rather than pneumatic pressure. The first twist, through 180掳, creates a zone of weakness. A second 180掳 rotation creates a constriction, making the rear part of the male鈥檚 abdomen flexible. After circumversion, the genital capsule flexes downward, so that the male can approach a female from behind and mate with her in the 鈥榤ale dorsal鈥� position. This allows sperm to be transferred quickly and safely in a stable position without the awkwardness of the postures adopted by more primitive dipterans. This position is best suited to mating on a surface rather than in flight. Indeed, most of the flies in this group court and mate on a surface, usually a source of food, and they often perform some sort of mating display.
Fixing the male dorsal position must bring some advantage, because it has evolved independently in another group of flies, the long-legged flies (belonging to the family Dolichopodidae, a member of the Orthorrapha). In this family, the male genitalia start out inverted, then fold laterally through a 180掳 arc to end up in the same position as in the Cyclorrapha. Similarly, long-legged flies also mate on substrates and the males have developed a wide range of ornamentation for courtship display. This is an example of convergent evolution, demonstrating how similar selective forces in two separate groups can produce similar adaptations in both form and behaviour.
What has caused two separate groups of flies to develop similar mating systems of such complexity? As torsion involves the reproductive organs, sexual selection may provide an explanation. According to the theory of sexual selection, organisms which are more successful in obtaining mates leave more offspring and are thereby 鈥榝itter鈥�. In most insect mating, females choose among males, putting selective pressure on males to be more attractive to their prospective mates. Changes in mating behaviour, and consequent changes in shape and form, are often driven by such sexual selection.
Many entomologists believe that aerial swarming is the ancestral mating system of all flies, and that only later did some groups take to mating on substrates. In most flies that swarm the genitalia are twisted through 180掳. In a large swarm, females often have little opportunity to choose between males and so accept the first one they encounter. Because individual selection is more difficult in swarms, sexual selection would operate to produce a method of mating where males which are more attractive to females have greater success. This undoubtedly happens in some groups, such as the dance flies (the family Empididae), where swarming males of some species try to attract females with prey or silken balloons that they have woven.
Mating on a substrate is far better suited to sexual selection by female flies than swarming. So, female choice may be the selective pressure which has caused some flies to abandon aerial swarming for the ground, encouraging coupling in the male dorsal position, with the genitalia rotated to a remarkable 360掳. This shows just how powerful sexual selection can be in determining the morphology of animals.
Of course this is not to say groups which swarm (and only have inverted genitalia) are not successful. On the contrary, a visit to the northern forests in early summer should convince anyone that the lower biting flies (blackflies, mosquitoes and biting midges) are alive and healthy, no matter how 鈥榩rimitive鈥� their methods of mating might be.
Torsion to this extent is an extraordinary phenomenon. No one could ever have predicted the development of a 360掳 twist of the genitalia in any animal, much less assume that it was a helpful adaptation. Yet it has happened and to a group of flies which is one of the most diverse and abundant groups of organisms on this planet.
* * *
SWARMING FOR SEX IN THE SUMMER
WALKING along a river at dusk on a summer鈥檚 evening, you will have seen loose columns of midges flying rhythmically up and down over one spot. Continuing on, you sometimes find your head suddenly enveloped by these tiny flies, which follow you and refuse to disperse.
What are the midges doing? They are swarming, and normally they maintain a definite column over a particular spot, often a prominent object, such as a dark rock, known as the swarm marker. When you arrived, the swarm simply switched its marker to a more prominent and elevated object 鈥� your head. As for the function of the swarm, it becomes immediately obvious if you swing an insect net through the swarm and examine the catch: almost all the midges are male.
A wide range of winged insects swarm. Most obvious are the mayflies, the lower Diptera (many families of the Nematocera and Orthorrhapha), and some caddis flies (Order Trichoptera), but a scattering of species in other orders, some bugs (Order Heteroptera), some lacewings (Order Neuroptera), and some sawflies and ants (Order Hymenoptera) also swarm in flight. Swarming is strongly associated with wet habitats, and mayflies, caddis flies, and swarming families of the Diptera all have aquatic or semiaquatic larvae.
Swarming is triggered by cues such as the intensity of light (dawn or dusk are usual times) and relative humidity. In most cases males gather over a marker and fly upwards, and then drift down passively. Females are attracted to the swarm and fly into it. When a male sees a female flying above, he reaches up and grasps her from beneath, and coupling takes place on the wing. The pair then leaves the swarm to complete mating on nearby vegetation. After mating, the female searches for places to lay her eggs and the male rejoins the swarm.
Swarms increase the efficiency of mating in dispersed populations by drawing individuals together. (This form of swarming should not be confused with swarming in bees, the emergence flights of termites, or insect outbreaks such as locust plagues.)
Swarming has undergone remarkable modifications in one fly family, the Empididae. Milan Chvala, of Charles University in Prague, has studied these flies in detail. Some males, he found, carry large items of prey into the swarm to attract females. When the pair couples, the male passes the prey to the female who takes a protein meal during mating. Further modifications involve the male wrapping prey or other objects in a silken thread that it has spun, or ultimately weaving an empty silken balloon to attract females.
Even stranger is the 鈥榬eversed swarming鈥� of some empidids, where females swarm and wait for males to fly through. Some of these females have feathery hairs on their legs to make them look larger while swarming.
Aerial swarming is universal in the mayflies, one of the oldest groups of insects, dating to the Upper Carboniferous, some 300 million years ago. Adults live for only a day, during which they swarm, mate and die. As mayflies use their wings almost entirely for this reproductive flight, swarming may have been the original function of insect wings. Also of interest is the mode of coupling by mayflies. The male curves his abdomen so that the upper side of his genitalia joins the underside of the female genitalia. This is the 鈥榠nverse interlock鈥� of genitalia, which possibly originated as an adaptation for coupling in mid-air. Inverse interlock is found in all flies, suggesting that the ancestors of Diptera were swarmers, as are many primitive flies today.
Daniel Bickel is in the entomology department of the Australian Museum in Sydney.
