Back to: Projects

Distribution of Afrotropical Kite Swallowtails

Higher classification of the swallowtails

On this page you can read about the place of the Afrotropical kite swallowtails within the butterflies. We work down from the highest level: the classification of the butterflies as a whole. Then we deal successively with the subfamilies of the swallowtail family (Papilionidae); the tribes of the subfamily Papilioninae; the genera of the tribe Leptocircini (with a diversion on the name 'kite' swallowtails); and, finally, the division of the genus Graphium into subgenera.

Classification of the butterflies

Taxonomists divide the 18,000 or so known species of butterfly (the suborder Rhopalocera) into five families. The Papilionidae — swallowtails and parnassians — is one of the smallest families, with about 550 species worldwide. Though few in number, they tend to be large and spectacular and include the largest of all butterflies, the birdwings.
The name swallowtail comes from the occurrence of long tails on the hindwings of many of them. The name is somewhat misleading. The presence of tails is far from universal, as the Afrotropical Graphium themselves demonstrate: fewer than half the species possess them.

The other butterfly families are the skippers (Hesperiidae), whites and sulphurs (Pieridae), blues, coppers and hairstreaks (Lycaenidae), metal-marks (Riodinidae), and admirals, fritillaries, emperors, satyrs and milkweeds (Nymphalidae).

We have good evidence that the Papilionidae form a natural (monophyletic) group. The most convincing character supporting this idea is a structure found only in the swallowtails and parnassians. It is a forked, tongue-like organ called the 'osmeterium' that the caterpillar can thrust out if it is molested. It releases chemicals that can apparently influence potential enemies, such as ants. The adult butterflies also share a number of structural peculiarities not seen among other butterflies.

Classification of the swallowtails

Many taxonomists have attempted to discover the relationships or evolutionary history of the papilionids. In our opinion, the most reliable attempt was that produced in 1987 by Jim Miller, then of the American Museum of Natural History in New York.

Species numbers given below are taken from the Global Butterfly Species Register (GART).

classification of papilionid subfamiliesThe 550 or so species of the family Papilionidae are divided into three unequal subfamilies. Each of these seems to be a natural group, defined by features of the thorax, wings, male and female reproductive organs, and their feet.

The Baroniinae includes just one rather drab (and tailless) species, Baronia brevicornis, confined to Mexico.

The Parnassiinae is somewhat larger, with about 67 species with a largely holarctic distribution. Most are attractive species and the subfamily includes the well-collected Apollo butterflies (which lack tails) and a few tailed genera.

The largest subfamily is the Papilioninae, with about 480 species. It includes many of the largest and most spectacular butterflies. They are cosmopolitan in distribution, with the greatest diversity in the Old World tropics. Only one species occurs in the UK, a rare and endemic subspecies of the widespread (N Africa to Japan and N America) Papilio machaon. With typical understatement, the butterfly is known in Britain as 'The' swallowtail.

Classification of the Papilioninae

cladogram of papilionine tribesThe Papilioninae are further divided into four tribes.

The smallest tribe, the Teinopalpini, consists of just two species of spectacular green and gold Himalayan butterflies, the Emperor of India (Teinopalpus imperialis) and its even more gaudy sister T. aureus.

The Troidini consists of about 130 species in 12 genera. They are at their most diverse in South America and the Indo-Australian region (with just Pharmacophagus antenor from the Afrotropical region). The tribe contains many large and colourful butterflies, including the Birdwings.

With over 200 species, but just three genera, the Papilionini is the largest tribe. Most species are placed in the cosmopolitan genus Papilio (this has been divided into up to six genera by some taxonomists). The other two genera are both oriental: Meandrusa (three species) and Chilasa (13 species).

The fourth tribe — the kite swallowtails that we are interested in — has been given various names. For complex historical reasons we prefer Leptocircini (the linked page tells you why) rather than Lampropterini or Graphiini. There are seven genera recognised in the Leptocircini, with a total of about 140 species.

Classification of the Leptocircini

Cladogram of LeptocirciniFour genera are almost entirely confined to the Americas. The only exception is Protographium with 13 American species and just one, P. leosthenes, from Australia. The other American genera are Eurytides with 7 species, and Protesilaus and Mimoides, each with 11.

The other three genera are all Old World. Lamproptera includes two species of small butterflies with transparent panels in their forewings and extremely long hindwing tails. Iphiclides also contains two species, but they are both quite large and spectacular. One of them, I. podalirius, is common and widespread in southern Europe and the Middle East (and is called the 'scarce swallowtail' in Britain); I. podalirinus is from China and is somewhat darker.

The other genus is Graphium.

Some taxonomists have included either or both Teinopalpus and Meandrusa in the tribe, but the characters that have been cited in order to include them appear to be either unreliable, variable or primitive and thus insufficient to prove a genuine relationship.

Why 'kite' swallowtails?

Linnaeus placed all the 192 species of butterflies he knew in his original genus Papilio. This was in the 10th edition of his encyclopaedic Systema Naturae, which is considered the starting point of modern zoological nomenclature. He placed the swallowtails in two subdivisions, the Greek and Trojan Knights (Equites Achivi and Equites Trojani), naming them after classical heroes. He included the few parnassians that he knew in other divisions such as the Heliconii and Nymphali.
As the number of butterflies known to science grew rapidly at the end of the eighteenth and into the nineteenth century, Linnaeus's original Papilio was divided up into the families we know today. Each of these families included one or more genera. The swallowtails remained in their original genus of Papilio.

The subdivision of the Papilionidae did not begin until early in the nineteenth century. But, although some genera were split from it, the increasing number of described species meant that Papilio grew increasingly unwieldy. Various ways of subdividing the genus were tried. One that persisted, perhaps because of the prominence of its proponents, Lord Walter Rothschild and his employee Karl Jordan, and because of the high profile of some of the publications in which the name was used, was 'Kite swallowtails' or its German equivalent, 'Segelfalter'. They divided Papilio into three sections.

  • The 'Aristolochia—swallowtails', were so named after their larval food plant and are equivalent to the tribe Troidini.
  • The 'fluted swallowtails', equivalent to the Papilionini, were defined by structural characters of the larvae.
  • The 'kite swallowtails' were also defined by the larval structure, but given their name because 'the most characteristic forms . somewhat resemble a paper kite.' The kite swallowtails included the species that now form the Leptocircini.

The name Graphium was first introduced in 1777 by Giovanni Antonio Scopoli (1723—1788), for a rather disparate group of about 57 species characterised by having the wings banded, but neither tailed nor ocellate. The name was largely ignored until Francis Hemming selected Papilio sarpedon as its type species in 1933, effectively restricting Graphium to the composition we use today. It was not until 1944, when E.B. Ford published a classification of the subfamily Papilioninae, dividing it into nine genera including Graphium, that the name came into general use.
As we now understand it, the genus consists of about 90 species from Africa and tropical and subtropical Asia. Campbell Smith and Dick Vane-Wright are trying to understand the phylogeny of the whole genus, including whether the Afrotropical species form a natural division, which has been subject to disagreement among previous workers.

Subdivisions of Graphium

Graphium has itself been divided into five subgenera.

Graphium (Pazala) contains six or seven species of striking — and strikingly similar — striped and swordtailed species found from Kashmir through the Himalaya and into southern China and northern Thailand and Indo-China. It has often been thought of as the most primitive subgenus — as sister to the rest of the genus. Our results suggest this may not be the case.

The 12 or so members of Graphium (Pathysa) are also striped and swordtailed, but with a greater variety of pattern than G. (Pazala). They also occur over a wider range: from northern India to the Philippines.

Graphium (Paranticopsis) occupies a similar range. The 12 or so species show a variety of patterns largely on the theme of pale marks on a dark background. Most lack tails, though G. phidias does have short narrow ones.

The 39 Afrotropical species (see checklist) have usually been placed in the subgenus Graphium (Arisbe). Our results suggest that this subgenus may not be a natural group. About one-third of the species are swordtailed; many of the remainder may be mimetic (see introduction).

Graphium (Graphium), the 'nominate' subgenus, may be the most problematic. There are about 30 species found throughout the Indo-Australian region. They show a wide variety of patterns, including several with green marks on a dark background. Some species have well-marked tails; in others they are rudimentary; yet others lack them altogether. An earlier team of researchers (Saigusa et al., 1977, 1983) divided them into four large species groups. Those groups each appear to be well-founded, with good characters to define them. But there do not appear to be any good characters linking them together to the exclusion of the others (and thus defining the subgenus).

It may well be that our ideas about the division of the whole genus may have to be revised.

Part of the reason for our study is to try to illuminate part of this complex pattern.