The text below is grayed out because it is not intended to be read. It is a necessarily imperfect OCR of the original and is only used by a search engine.
Journal of The Lepidopterists' Society
Volume 14 1960 Number 3
HARMONIZATION OF CONCEPTS OF HIGHER CLASSIFICATION OF THE PAPILIONID^E
by Eugene Munroe and Paul R. Ehrlich
Previous work carried out independently by us on the basis of largely different sets of characters (Ehrlich, 1958; Munroe, 1953, and in press) led to interpretations of papilionid classification that were in harmony in major outline, though they differed in several points. Consultation and joint examination of certain characters has permitted resolution of all the points of difference and of some points of uncertainty in our previous arrangements. The present paper can be considered a supplement to and revision of our individual contributions on this subject.
Our earlier classifications agreed in considering the Baroniinae, Parnassiinae and Papilioninae as subfamilies of Papilionidae, in considering the phyletic separation of Baroniinae as considerably antedating that of the Parnassiinae and Papilioninae, in considering the Zerynthia and Parnassius groups as not fundamentally distinct, and in associating Lamproptera with the Graphiini and Cressida and Euryades with the Parides complex. We differed mainly in that (1) Ehrlich classified Teinopalpus in a separate tribe of the Papilioninae, whereas Munroe placed this genus in the Graphiini; and (2) Ehrlich placed Battus and provisionally Troides and Ornithoptera in the Papilionini, whereas Munroe placed them with Cressida, Euryades and. Parides. Also, (3) although we are agreed that genera have been far too finely divided in recent work on several groups of butterflies, Munroe divided the genera Parides and Graphium as understood by Ehrlich, who did not carry his work to the generic level. This division was based on genitalic and other characters. Munroe attempted a more detailed phylogenetic classification than did Ehrlich, whose interests were in a broader field. Some points of major uncertainty remained, however, in Munroe's classification. The most important of these were: (1) the phylogenetie relationship of the
169
170
Munroe & Ehrlich: Papilionidae
Vol.14: no.3
Parnassiinae and the Troidini; (2) the relationship of the red-tuberculate Aristolochia-teeding larval type to the green, Lauraceae-feeding type; and (3) the origin and internal phylogeny of the Papilionini.
Points of Previous Difference
The points of difference seem best resolved as follows:-
(1) Teinopalpus appears to have real affinities with the most primitive Graphiini, but it differs in the inflation of the frons and the associated hypertrophy of the palpus, in lacking sclerotized patagia, in having smaller tentorial crests, in the weak development of the cubito-vannal cross-vein, in the sexual dimorphism and specialization of the pattern and wing-shape, in the atypical wing-venation, and in the reduced antennal scaling. It appears to be a matter of individual preference whether the differences or the resemblances should be emphasized. Possibly the best solution is to recognize Teinopalpiti and Graphiiti as sub tribes of Graphiini. Teinopalpus lacks the spine of the prodiscrimen, as do other Graphiini except Dabasa. The statement to the contrary in Ehrlich (1958) resulted from an error in proofreading.
(2) Troides and Ornithoptera, as stated by Munroe and hinted by Ehrlich, are typical members of the Cressida - Farides complex. The name Troidini is to be preferred over Cressidini for the union of the two tribes recognized by Ford, as names based on Troides are older in the literature and Ford's Troidini form by far the larger constituent of the combined group. Ornithoptera and Troides are undoubtedly derived from a common ancestor more like Troides than Ornithoptera, for Troides has no specialization of pattern or wing-shape that is not found or suggested in Ornithoptera, whereas Ornithoptera has greatly specialized pattern and strong sexual dimorphism and has lost the Parides-like sex-scaling on the anal margin of the hind wing, still evident in Troides, and has substituted a new set of sex-scaling on the upper surface of the fore wing. The retention of slight sclerotization of the patagia and the longer free course of Ri indicate a separation of the Troides-Ornithoptera complex from a position near or below the base of the present genus Farides, but this conclusion requires confirmation from examination of a wider range of species. Battus is far more widely different. The narrow spinasternum and the absence of definite sensory pits on the under surface of the antenna indicate a separation of Battus from the remaining Troidini before the separation of the Cressida - Euryades and the Farides - Troides stocks. These four genera agree in having the two specializations, broad spinasternum and definite sensory pits. On the other hand, the detailed structural correspondence of the Aristolochia-feeding, fleshy-tubercled
1960
Journal of the Lepidopterists' Society
171
larva, of the arcuate, flanged pupa, and of the male genitalia indicates beyond possibility of coincidence the direct relationship of Battus with the other Troidini. Battus is, however, by far the most distinctive genus of the tribe, and it is probably best here also to recognize two subtribes, Battiti and Troiditi.
(3) The generic separations advocated by Munroe appear to be reasonably founded. In the Troidini, Pachlioptera, comprising P. hector, aristolochiae, polydorus and allies, is a compact group, certainly closely related to Parides and presumably derived from it, but differing in the radical character of near-abortion of the valves and pseuduncus and enlargement and sclerotization of the socii apparently to take over a clasping function. These differences are supported by minor differences in the structure of the female genitalia and in the larvae and the pupae. Cressida and Euryades are not closely related to Pachlioptera as assumed by Ford on the basis of an observation copied from Talbot and based in the first place on misinterpretation by the latter of an external examination of the male genitalia. In the Graphiini, division of the New World Kite Swallowtails (Eurytides), which, like the related Old World genus Lamproptera, retain the uncus and socii as a composite structure, from the main group of Old World species (Graphium), in which the uncus is aborted, appears to be satisfactory and to define homogeneous groups. The three groups of Old World Kite Swallowtails that retain the free Ri that is characteristic of most Eurytides also all have peculiar genitalia. The Australian species leosthenes retains the uncus weakly as a fingerlike structure; Munroe has proposed the genus Protographium for this species. Protographium is apparently a relict connecting link between the large genera Graphium and Eurytides. The payeni - evan group is very distinct in genitalia from Graphium; it resembles it in having lost the uncus, but has large horny socii, articulated firmly with the eighth tergite; it also differs from Graphium in the short: cell and distorted discocellulars, in the weak sclerotization of the patagia and in having a small spine on the prodiscrimen, lacking in Graphium. This group forms a very distinct genus for which the name Dabasa is available. The European species podalirius, together with its Tibetan representative podalirinus, constitute the third group with free Ri. In this group the male genitalic structure is fundamentally similar to that of Graphium, but the valve is very simple and unusually elongate. The larva is unusual in appearance and food plant and has segmental red spots. This combination of characters would support separation of these two species as the genus Iphiclides. This separation has the advantage of making Graphium homogeneous for anastomosis of Sc and Ri. However, the separation is a weak one, and some students may prefer to unite these
172 Munroe & Ehrlich: Papilionidse Vol.14: no.3
two genera. To summarize, these separations represent either rather large phyletic cleavages (Eurytides, Graphium, Dabasa) or distinct and well-characterized side-lines or relics (Pachlioptera, Iphiclides, Protographium). They are on a considerably coarser plane than the divisions currently fashionable in such groups as the Blues and Coppers. With the possible exception of the separation between Iphiclides and Graphium, they should be retained.
Points of Previous Uncertainty
(1) Relationship of Troidini to Parnassiinae and to other Papilioninae. The key to this problem appears to be the cubito-vannal vein. This was previously considered by Munroe to be a primitive character, but we now believe this interpretation to be erroneous. No similar vein is found in other families of butterflies, so far as our reading extends; it is certainly lacking in a variety of forms examined by us. No comparable vein is found in higher groups of moths, including Cossidae and Castniidae. The vein is also absent or rudimentary in the two other subfamilies of Papilionidae, Baroniinae and Parnassiinae. The nearest parallel we have found outside of Papilioninae is the sclerotized root of 1st V found in Limenitis and certain other Nymphalidae. It is possible that the vein present in Papilioninae is such a root stalked for a certain distance with Cu; alternatively it may be a structure formed de novo. It is noteworthy that in Teinopalpus, which has perhaps the most primitive genitalia of any papilionine, the vein is weakly developed. At any rate it seems highly probable that this vein is an unusual specialization, characteristic of the Papilioninae, and that it indicates with great probability that this subfamily is a natural and monophyletic group, as would be suggested by their considerable uniformity in wing-venation and palpal structure, and by the possibility of harmonizing the different larval and pupal types without undue stretching of the imagination. As the Graphiini have more primitive genitalia, legs and antennae than the Troidini, which resemble Parnassiinae mainly in the early stages and in the presence in certain species of a sphragis, we must conclude that the Troidini are derived from an ancestor which on the defining characters of scaled legs, scaled antennae and free, complete male genitalia we would be forced to refer to the Graphiiti. There is therefore no question of direct relationship of the Parnassiinae and Troidini. There must have been a graphiine intermediary, though probably one more closely similar to generalized Troidini and Parnassiinae than are existing species of Graphiini. We may recall at this point that the existing troidine genus Battus has apparently never acquired two of the most prominent speciali-
1960
Journal of the Lepidopterists' Society
173
zations of the remaining Troidini. This involves the corollary conclusion that the Parnassiinae-like features of Euryades and Cressida are secondary; this applies particularly to the narrowed, Parnassius-Mke valve. The sphragis in these genera and in certain primitive Parides may indeed be derived from the same source as that of Parnassius, though perhaps only as a common tendency, rather than as an overt character in an unbroken line.
(2) Relationship of red-tuberculate, Axisto\ochi&-feeding larva to green, sometimes brown, sometimes spinose, Lauracese - or Rutacese - feeding larva. The solution of this problem is bound up with that of the previous one. If we assume, as we apparently must, that the; Papilioninae are a homogeneous, monophyletic group, it is ascribing too much to coincidence to suppose that the red-tuberculate, Aristolochia-ieeding larva has been independently developed in the two groups Parnassiinae and Troidini from smooth, green, Lauraceae-feeding, skipper-like or pierid-like ancestors. On the other hand, it is not hard to imagine a reversion from Aristolochiaceae to the lauraceous food-plant that characterizes several primitive hesperiids and pierids. The acquisition, perhaps several times, of a cryptic, green or brown, coloration would also be easy. It may be noted further that larvae of many species in Graphiini and Papilionini retain red spots and/or tubercles in the final instar, while all known species are tuberculate and many red-spotted in earlier instars. This permits a decision between the alternative hypotheses presented in Munroe's paper. The red-tuberculate, Aristolochia-\deeding larva must be primitive, at least for the Parnassiinae and Papilioninae. It follows that the primitive Graphiini must have had red-tubercled larvae, probably somewhat Zerynthia-like in aspect, and feeding on Aristolochia. The hearsay reports of the larva of Lamproptera that were cited by Munroe (in press) appeared to fill this need, but unfortunately Mr. Kent Wilson (in litt.) informs us that these reports were erroneous and can be disregarded. Mr. Wilson's description of the early stages of Lamproptera will be awaited with interest. Larvae of the Eurytides lysithous group have tuberculate and spotted larvae, but there is a possibility of mimicry of Parides larvae. In our present state of knowledge it cannot safely be assumed that all red spots and all tubercles are homologous. Comparative morphological study of a wide range of larvae is badly needed. It may be noted parenthetically that the larva of the primitive though aberrant Teinopalpus is described as being smooth and green. It may be repeated for good measure that we have no knowledge or incomplete knowledge of several of the really critical life histories for phylogenetic interpretation, viz.: Teinopalpus, Lamproptera, and Dabasa from Asia and Baronia from Mexico.
MUNROE & EHRLICH
PAPILIONID^
Par ides
Pachlioptera Dabasa PAPILIONINI
Troides Ornithoptera
Battus
Iphiclides
Cressida
BATTITI TROIDITI
TROIDINI
PARNASSIINAE
BARONIINAE
PAPILIONIDAE
Graphiurn
Protographium
Eurytides Euryades
Lamproptera
Teinopalpus
GRAPHIITI TEINOPALPITI
GRAPHIINI
PAPILIONINAE
_____I
PIERIDAE
Apparent sequence of separation of papilionid lines. The significant feature of this diagram is the order of the bifurcations.
174
1960 Journal of the Lepidopterists Society 175
(3) Origins and internal phytogeny of the Papilionini. This problem continues to present many points of uncertainty. The most probable origin of the tribe still appears to be from the higher Graphiini. The presence of the spine of the discrimen is a point of similarity between Papilionini and Troidini that can be added to the well-known ones. This spine is lacking, however, in Battus, the troidine genus that has other PapiZ/o-like characters. Dabasa, unlike other Graphiini, has the spine, though weakly. This would support the possibility that this genus arose close to the point of separation of Papilionini from Graphiini. On the whole it seems likely that the spine has been acquired independently in the Papilionini and the Troidini. It is obvious that final answers on the precise ancestry of the Papilionini are not possible on the evidence available to us. Still less can the vexed questions of the internal phylogeny of the Papilionini be properly resolved. This must await the results of more comprehensive studies than time or material have permitted us to make.
References
Ehrlich, Paul R., 1958. The comparative morphology, phylogeny and higher
classification of the butterflies (Lepidoptera: Papilionoidea). Univ. Kansas
sci. hull. 39: 305-370. Munroe, Eugene, 1953. The phylogeny of the Papilionidae. Vroc. 7th Pacific sci.
congr., Auckland, 1949 4: 83-87. Munroe, Eugene, in press. The generic classification of the Papilionidae. Canadian
ent., supplement 17.
(EGM) Entomology Research Institute, Research Branch, Canada Agriculture,
Ottawa, Ont., CANADA and
(PRE) Dept. of Biological Sciences, Stanford University, Stanford, Calif., U. S. A.
A NEW ENTOMOLOGICAL SERIAL
The Society Library has received the first two numbers of Esakia published by the Hikosan Biological Laboratory, Faculty of Agriculture, Kyushu University, Fukuoka, Japan. They include nine short papers by present or former members of the laboratory shaff; among these are two by Hiroshi Kuroko describing new microlepidoptera (Gracilariidae and Cosmopterygidae). The issues are well illustrated by photos and line drawings. Members desiring further information are advised to write to the Chief of the Laboratory, Professor Keizo Yasumatsu.
P. F. Bellinger