INFORMATION TO USERS

Transcription

1 INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. University Microfilms International A Bell & Howell Information C om pany North Z e eb Road. Ann Arbor. Ml USA / /

2

3 Order N um b er A r ev isio n o f th e L e p to ly cin i (C o leo p tera :L y cid a e) w ith a d iscu ssio n o f p a ed o m o rp h o sis Miller, Richard Stuart, Ph.D. The Ohio State University, 1991 C opyright 1991 b y M iller, R ichard S tu art. A ll rights reserved. UMI 300 N. Zccb Rd. Ann Arbor, MI 48106

4

5 A REVISION OF THE LEPTOLYCINI (COLEOPTERA: LYCIDAE) WITH A DISCUSSION OF PAEDOMORPHOSIS DISSERTATION Presented in Partial Fulfillment of the Requirements f the Degree Doctor of Philosophy in the Graduate School of The Ohio State University by Richard S. Miller, B.A., M.S. * * * * * The Ohio State University 1991 Dissertation Committee: Approved By: Dr. David J. Horn Dr. Donald E. Johnston ta Dr. Charles A. Triplehorn Dr. Barry D. Valentine Advisor Department of Entomology

6 Copyright by Richard S. Miller 1991 I

7 To Dr. Roy A. Crowson for his study and his provocative queries ii

8 ACKNOWLEDGEMENTS How dull it is to pause, to make an end, To rust unburnished, not to shine in use! As though to breathe were life itself! -A. Tennyson, The immensity of biology can either elate or overwhelm the student. There are, essentially, two responses to this conundrum. One is the quixotic charge here and there at the fog of unexplained phenomena. The other is to succumb to a Faustian despair of ever encompassing the morass of little suspected knowledge. While fully appreciating the absurdity, I chose the former when most of my cohort were increasing their biological fitness (sensu Colinveaux, 1982:394). This current milestone in my studies has given me pause to reflect on those who have contributed to my formal and informal education. It is impossible to differentiate adequately between them and, thereby, classify them - for they have made varying contributions of infinitely subtle distinction. A "lumper" might thank Western Civilization as probably more than 997. of my thoughts are a result of inculcation from this source. A "splitter" would begin a list of individuals. Aristotle, a member of Western Civilization, has urged moderation and, so, I will mention a few stellar contributions. Of course Niki comes to mind first. Besides her readily apparent artistic abilities demonstrated in this dissertation, her support and encouragement have significantly contributed to its conclusion and my iii

9 happiness. Again, my family has been a source of support, even though they kindly find my interests "different." I thank them for everything. My friends and former office- mates of B & have been inspirational in our discussions. These peers include Michael A. Ivie, William F. Abeles, James B. Stribling, Paul S. Cwikla (really only 309-A), and recently Peter W. Kovarik. I thank Nick Calderone, an honorary member of our clique, who has done much to enrich my appreciation of philosophy and other arcane subjects. The various members of the faculties of CSUS, UCD, and OSU who have contributed to my education are appreciated. Especially important among these are Harold Wiedmann and Rollo Darby; Martin C. Birch; Charles A. Triplehorn, Barry D. Valentine, David J. Horn, Donald E. Johnston, Norman F. Johnson, and J. Bruce Griffing respectively. The staff of the OSU biology library has been immensely helpful. I would like to thank them all, and especially Susan Ward, for their patience, understanding, and assistance. Several of ray colleagues around the world have assisted in various aspects of my work including the loan of specimens. The latter are listed in the materials and methods section, although I would like to acknowledge all here. This paragraph is reserved for my reading committee should they decide to pass me after the laborious task of reading this manuscript. Additionally, I would like to thank Dr. David Horn for his question during my general exam on testability by use of hypothesized phylogenies. It was seminal to some of the ideas developed here. iv

10 Finally, I thank all those natural historians and scientists, whether trained or untrained, who for whatever motives have left their thoughts in print. If I have been remiss in failing to cite their work, I here apologize and can only plead the finite limitations of a human mind. History <ie. pattern) is not intrinsically useful unless its lessons reveal process. It cannot explain process. Hence, the validity of a phylogenetic hypothesis is not as important as the testable questions it compels. This dissertation gently asks how ontogeny relates to beetle evolution and the significance of paedomorphosis in cantharoid evolution. I hope this relationship of ontogeny and phyiogeny can be examined more cogently by providing this arena for further study. If I have failed, the enclosed cladograms must be viewed merely as art to be stored in some attic. Art is the culmination of human imagination, but it was not my purpose. v

11 VITAE 15 February Nativity, Los Angeles, California Bachelor of Arts in Biology with a concentration in Zoology, California State University, Sacramento University of California, Davis Economic Entomologist I- III, California Department of Food and Agriculture, Sacramento Master of Science in Entomology, The Ohio State University, Columbus Insect Identifier, United States Department of Agriculture present... Ph.D. in Entomology, The Ohio State University, Columbus PUBLICATIONS The Buprestidae of the Virgin Islands. Florida Entomologist 67(2 ): (co-authored with M. A. Ivie) The behavior of Calooteron reticulatum (F. ) larvae (Coleoptera: Lycidae). Ohio Journal of Science 88(3 ): FIELDS OF STUDY Major Field: Entomology vi

12 TABLE OF CONTENTS ACKNOWLEDGEMENTS... iii VITA... vi TABLE OF CONTENTS... ix LIST OF TABLES... xi LIST OF FIGURES... xi i INTRODUCTION... 1 MATERIALS AND METHODS... 3 Methods... 3 Systematic analysis... 3 Taxonomic conventions... 6 Paedomorphosis... 7 Material studied ONTOGENY AND PHYLOGENY Introduction Definition of the p r o b l e m An example A model Synergism of morphology and systematics Phylogenetic implications Possible taxa for further study PHYLOGENETIC RELATIONSHIPS WITHIN THE CANTHAROIDEA vi i

13 Relations within the Elateriformia Monophyly of the Cantharoidea Current hypotheses of Cantharoid relationships Family relations within the Cantharoidea Omalysidae Plastoceridae Cantharidae Omethidae Phengodidae Telegeusidae Dri 1 idae Lampyridae Lycidae THE TRIBE LEPTOLYCINI Key to Genera of Leptolycini Genus Abrolvcus New Genus Genus Anti 1 lolvcus New Genus Genus Ceratoprion Gorham Genus Flabel locaenia Pic Genus Leptol vcus Leng and Mutch ler Genus Pseudacroleptus Pic Genus Pseudoceratoprion New G e n u s Phylogeny of the Leptolycini Monophyly of the Leptolycini Lycid relationships of the Leptolycini Hypothesized Sister Group vi i i

14 Character discussion Relationships within the Leptolycini Phylogenetic Conclusions Biogeographical implications Concluding remarks LIST OF REFERENCES APPENDIX (Illustrations) ix

15 LIST OF TABLES TABLE 1. Character states found in Thvlodrias Possible results of a test as proposed by Lauder Character states of Leptolycini x

16 LIST OF FIGURES FIGURE 1. Theoretically available phylogenetic reconstructions for paedomporphosis CprogenesisD (after Westheide, 1987:850). 2. Models of possible ontogenetic pathways of a character. 3. Hypothesized phylogeny of the genera of the Leptolycini. 4. Abrolvcus bicolor New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Right antenna. e. Right hindwing, 5. Abrolvcus cubensis New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. 6. Abrolvcus darlingtoni New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Lateral aspect head and prothorax. 7. Abrolvcus iviei New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. 8. Abrolvcus omalvsiformis New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Lateral aspect head and prothorax. 9. Abrolvcus sandersoni New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Lateral aspect of head and prothorax. e. Internal male genitalia. XI

17 10. Anti1lolvcus auratosuratus New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Last abdominal sternite. e-f. Ultimate and penultimate abdominal tergites - e. internal view, f. external view. 11. Anti 11olvcus elongatus New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Lateral view of head and thorax. 12. Anti 1lolvcus flavomarginatus New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. a. Lateral aspect head and prothorax. 13. Anti 1lolvcus semiflavus CChevrolat). a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. 14. Ceratoprion bordoni New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Right antenna, e. Right hindwing. 15. Ceratoprion chaelae New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. 16. Ceratoprion mandibularum New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. right antenna. 17. Ceratoprion nigrum New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Last abdominal sternite. e-f. Ultimate and penultimate abdominal tergites - e. internal view, f. external view, g. Internal male genitalia. IS. Ceratoprion periosus New Species. a-c. Aedeagus - a,

18 dorsal aspect, b, lateral aspect, c, ventral aspect. 19. Ceratoprion serricorne Gorham. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Lateral aspect head and prothorax. e. Left hindwing. f. Left antenna. 20. Flabellocaenia bourgeoisi Pic. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Lateral aspect head and prothorax. e. Last abdominal sternite. e-f. Ultimate and penultimate abdominal tergites - e. internal view, f. external view. 21. Flabel1ocaenia elegantulus New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Lateral aspect head and prothorax. 22. Flabellocaenia leticia New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. 23. Flabellocaenia iolei New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Lateral aspect head and prothorax. e. Frontal view of head. 24. Flabellocaenia pecki New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Lateral aspect head and prothorax. 25. Flabellocaenia rimae New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. 26. Leptolvcus adiaphorus New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. xi ii

19 Last abdominal sternite. e-f. Ultimate and penultimate abdominal tergites - e. internal view, f. external view. 27. Leptolvcus brunneus New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Lateral aspect head and prothorax. e. Internal genitalia. 28. Leptolvcus dominicensis New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. 29. Leptolvcus effeminatus New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Lateral aspect head and prothorax. 30. Leptolvcus flavicol1 is Leng and Mutchler. a-c. Aedeagus a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Lateral aspect head and prothorax. 31. Leptolycus heterocornis Leng and Mutchler. Habitus of adult male. 32. Leptolvcus heterocornis Leng and Mutchler. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect, d. Lateral aspect head and prothorax. e. Right hindwing. 33. Leptolvcus heterocornis Leng and Mutchler. Habitus of adult female. 34. Lepto1vcus heterocornis Leng and Mutchler. Habitus of 1arva. 35. Pseudacroleptus caeruleus New Species. a-c. Aedeagus a, dorsal aspect, b, lateral aspect, c, ventral aspect. 36. Pseudacroleptus neblinus New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. xi v

20 Lateral aspect head and prothorax. e. Rigth wing. 37. Pseudacro1eptus neblinus New Species. a. Thorax - dorsal aspect on right and ventral aspect on left. 38. Pseudacro1eptus neblinus New Species. a. Head - ventral aspect. b. Right maxillary palp. 39. Pseudacro1eptus obscuricolor Pic. Habitus of adult male (Type). 40. Pseudoceratoprion bechvne New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Lateral aspect head and prothorax. 41. Pseudoceratoprion brasi1iensis New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Right antenna. 42. Pseudoceratoprion colombiensis New Species. a-c. Aedeagus - a, dorsal aspect, b, lateral aspect, c, ventral aspect. d. Lateral aspect head and prothorax. xv

21 INTRODUCTION Although they have been the object of little recent study, lycids present some interesting questions in both their biology and phylogeny. Presently, the literature of this group is confined largely to alpha taxonomy. Exceptions occur in the discussion of "trilobite" larvae culminating in the work of Mjorberg (1925), some discussion of mimicry (Darlington, 1938; Parsons, 1940; Linsley et a l., 1961; Emmel, 1965; and Burke, 1976), and a few papers on larval feeding (Mjorberg, 1925; Withycombe, 1926; McCabe and Johnson, 1979 and 1980; Miller, 1988). Because there has been little synthetic work on a global scale, phylogenetic relationships within the family are obscure. Kleine's catalogue (1933) is the last treatment of the entire family, but it is, perhaps, useful only as a list of described taxa in the New World. Although he had worked on lycids for 26 years, Kleine s studies were largely confined to the Old World, especially the Indo-Malaysian region. He published only two papers limited to alpha taxonomy on the New World fauna (Kleine, 1941 and 1942). As a result, he had to rely heavily on the earlier work of Bourgeois (1905) and vague statements such as "voisin de" of Pic for the relationships of Neotropiaal members of the family. A further complication in the systematics of the Lycidae is that much of the taxonomic literature is too brief and emphasizes only color

22 characters which are often of little use. This has necessitated the examination of types before names can confidently be placed on any Neotropical specimens. This chaotic state has done little to induce studies in the biology of the Lycidae and, thus, they remain largely unknown. What i known, however, is tantalizing. For example, some species are known t aggregate as adults or larvae, but no species is known to aggregate in both stages. Aggregations are often composed of numerous distantly related species and are thought to serve as models for mimetic complexes. Although lycids are thought to be an evolutionarily recent group (Crowson, 1972), several of the widespread and apparently basal groups have " larviforra" females which are incapable of flight. The foregoing had prompted me to study one tribe within the Lycidae, the Leptolycini, but eventually required the examination of ever larger problems that culminated in the following study of some aspects of cantharoid phylogeny and an examination of paedomorphosis.

23 MATERIALS AND METHODS The methodology discussed here serves two purposes. First, it serves as a basis for evaluation of my work. Second, it references my methodology so it need not be repeatedly discussed in fucure work except in its difference from this approach. Systematic Analvsis. The data gathered in this study were analy2ed using the methodology of phylogentic systematics as proposed by Hennig (1965, 1966) and developed by numerous other workers (see Wiley, 1981 and Ax, 1987 for a review and list of references and, also, recent issues of Systematic Zoology, Systematic Botany, and Cladistics among others). This method assumes that 1) evolution occurred, 2) speciation can be recognized by novel characters, 3) relationships can be recognized by the recency of these shared novelties (ie. synapomorphies), and 4) multichotomies are rare. Therefore, in addition to a hypothesis of character cladogenesis, the resulting tree is also a hypothesis of the phylogenetic relationships within the taxon. These hypotheses will be called phylogenetic hypotheses for brevity, but no ancestor - descendant relationship is implied. A further assumption for transposition of phylogentic hypotheses into the Linnean classification is that evolution is hierarchical. 3

24 Species Recognition. The principle method of recognizing species was the morphological species concept. This implies that some difference (eg. autapomorphy) was found consistently between populations in which no intermediates were found to bridge the gap (ie. character distributions are discrete, not continuous). Exemplars. When analyzing higher taxa, exemplars are the only feasible method of study. Their selection was largely a result of availability of specimens, but attempts were made to find any aberrant cantharoids for study. Besides the material borrowed from the various museums and individuals mentioned in the acknowledgements, my study of character states within the Coleoptera was restricted to my personal collection (RSMC) and the Ohio State University collection (OSUC). No attempt was made to list these specimens studied by repository, but the majority were from the RSMC. Character Weighting. No character state was arbitrarily weighted. However, implicit weighting must have occurred inadvertently by choice of characters examined, by availability of material, by human cognition, and by my cultural bias. Additionally, characters were intentionally weighted a priori if they were the result of loss or reduction, if they were changes of degree, or if they were thought to be affected by paedomorphosis. Loss was weighted because homology is notoriously difficult to document when it cannot be seen. Lineages supported by reduction or loss character states were considered suspect and additional support was sought. If increased homoplasy resulted from their inclusion in the analysis as a synapomorphy, then the character states were recoded as

25 unique novelties and the resulting data were reanalyzed. The rationale for such action was not to produce a more parsimonious cladogram, but to draw attention to other homoplasy that might provide more fruitful investigation. Acceptance of hypotheses was subsequently based on parsimony. The homology of recoded character states remains uncertain (see below) and is so noted. Presence - absence character states were sought preferentially for analysis. The use of those states resulting in a change of degree, such as shape, was minimized to those with only two states. If a third state is found, it is often not possible to polarize transformation series except by a priori assumptions of parsimony or aesthetics. A case in point is the presence of a square, a transversely rectangular, and a longitudinally rectangular pronotum. Even if the functional out group resolves the basal state, the remaining states have three possible transformation series which can be resolved for use as a synapomorphy only arbitrarily in the absence of studies of allometry during ontogeny. Such states can, however, be used as autapomorphies. Characters whose expression were demonstrated to be affected by paedomorphosis were not used in the analysis of paedomorphic species. This rationale will be further discussed during the treatment of ontogeny and phylogeny (p. 12). Homology. The homology of characters was hypothesized using the criteria of Remane (1956). These are the positional criterion (Remane, 1956:30), the similarity criterion (Remane, 1956:42), and the criterion of intermediates (Remane, 1956:45). These and the criterion of correlation of transformation series (Hennig, 1966:96) are accepted only

26 as operational rules for pragmatic reasons during hypothesis formation. The actual test of a hypothesis of homology can be ontogeny which is not just "complementary to the criterion of out - group comparison" as stated by Wiley (1981:154). The latter is a phylogenetic criterion sensu Hennig, but ontogeny can be an actual independent test subject to falsification sensu Nelson (1978). Polarization. The outgroup method (Watrous and Wheeler, 1981 ) was used during analysis with consideration of the qualifications discussed by Madison et a l. (1984). Use of Nelson's ontogenetic rule (1978) is provisionally rejected because of the difficulty of assessing its falsifiabi1ity (ie, differentiating between a more general state due to basal position or one due to paedomorphosis ). Additionally, the rule assumes a uniform transformation which may or may not be real. When a sufficiently large data base becomes available, ontogenetic character precedence may prove useful in analysis (see pp. 28 and 36) if caenogenesis with polarizable character states can be demonstrated, but it remains unexamined. Taxonomic Conventions. Diagnoses. A combination of both syraplesiomorphies and apomorphies was often used to construct diagnoses. Although the former has information content, it is not sufficient for diagnosis of a species. Newly discovered taxa are more likely to have such characters than apomorphies and, hence, may not be separable or even recognized. Taxa, after all, are composed mostly of symplesiomorphous character states.

27 Keys. No attempt was made to construct phylogenetic keys. Although such keys are arguably more stable, it must be remembered that the results of a phylogenetic analysis are only hypotheses. It seems unfair to user the reader to examine apomorphic character states that are perceived only with difficulty if more easily observed plesiomorphic characters are available. Again, additional data may refute the current hypotheses and necessitate an additional key. However, the cladogram (Figure 3) can be so used. Classification. The taxonomic hierarchy was constructed using the traditional Linnaean hierarchical method. As there is no known algorithm for deciding the limits of superspecific taxa, their delimitation is at the discretion of the specialist. The merits of combining taxa to show relationships instead of separating them to emphasise uniqueness (Ball 1975:149) were considered, but stability was also a major consideration. In all cases hypothesized monophyly was the final arbiter. Subgenera are not used. They will only be retained when their names are already available and the lineage is still in need of analysis. Species groups suffice in other cases and do not clutter the literature unnecessarily. PAEDOMORPHOSIS. Paedomorphosis was examined to determine its impact on morphology and, hence, on phylogenetic analysis. Its presence was recognized by wingless, "larva-like" females. A species of Dermestidae amenable to culturing, Thvlodrias contractus Motschulsky, was selected for

28 comparative study, because its phylogenetic relationships are well supported. The methodology followed here is a documentation of character states found in both sexes in a paedomorphic species and subsequent comparison with character states found in other members of the same family. Study of the lineage was primarily restricted to adult material from the RSMC, but was supplemented by OSUC and MCZC material Genera examined consisted of Dermestes L., Attagenus Latreille, Novel si s Casey, Trogoderma Dejean, Globicornis Latreille, Megatoma Herbst, Dearthrus LeConte, Anthrenocerus Arrow, Ctesias Stephans, Thaumoglossa Redtenbacher, Hemirhopalurn Sharp, Orphinus Motschulsky, Crvptorhopalum Gu6rin-Menevi1le, Anthrenus Schaeffer, Apsectus LeConte, Trinodes Dejean Evorinea Beal, and Thvlodrias Motschulsky. This constitutes 18 of 43 genera in 7 of the 9 subfamilies recognized by Mroczokowski <1968) excluding Orphilinae which has been moved to the Nosodendridae by Ivie (1985:64). In the latter family, which is hypothesized to be the sister to the Dermestidae, Erickson and Nosodendron Latreille were examined. both Orphi1is Additional dermesti genera were not available, and the literature was consulted for their character states. Characters whose states are unknown in the rest of the Dermestidae are not discussed. The differences in character state trends of the paedomorphic dermestid were then compared to those of other putative paedomorphic species. In a few cases specimens were available, but it was also

29 necessary to consult the literature. Cantharoid * specimens available included Lamovris noctiluca L. male and females, Drilus and Selasia males (Drilidae), Omalvsus males (Omalysidae) and males of numerous genera of Phengodidae and the female of Phengodes. Also, Diphvllostoma fimbrata Fall (Scarabaeoidea ) in the personal collection of Michael A. Ivie (MAIC) was examined. Thvlodrias Mots, was not illustrated here as ample illustrations can be found throughout the literature. These include Hinton (1945: figs ), Franciscolo (1975: figs. 1-7), and Suss and Foggato, : figs. 1-22). As there is an insufficient data base to incorporate ontogenetic data in a phylogenetic analysis using the methodology of Nelson (1978 and 1986), alternative approaches were sought. Material Studied The following museums and collections provided material for study. Examination of the material studied for each species of leptolycine will demonstrate the paucity of that material. AMNH ANSP American Museum of Natural History, Central Park West at 79th Street, New York, New York Dr. Lee H. Herman, Jr. Academy of Natural Sciences of Philadelphia. 19th and The Parkway, Philadelphia, Pennsylvania Mr. Donald Azuma and Dr, Daniel Otte. Even though my studies in progress corroborate Lawrence's (1987) action synonymizing the Cantharoidea with the Elateroidea, I will continue to use the former name for convenience during the discussion of those lineages formerly placed in the superfamily within this paper.

30 10 BMNH British Museum (Natural History). London, SW7 5BD, England, Great Britain. Dr. C. M. F. von Hayek. CASC CISC CNCI FSCA California Academy of Sciences, Goldengate Park, San Francisco, California Drs. David H. Kavanaugh and Norman Penny California Insect Survey, University of California, Berkeley, California Dr. John A. Chemsak Canadian National Collection, Agriculture Canada, Biosystematics Research Institute, Ottawa, Ontario K1A 0C6. Mrs. Jean McNamara and Dr. Ales Smetana Florida State Collection of Arthropods, Florida Department of Agriculture and Conservation Service. P.O. Box 1269, Gainesville Florida, Dr. Robert E. Woodruff. HAHC Personal collection, Drs. Anne and Henry Howden. Carlton University, Ottawa, Ontario K1A 0C6 ICCM Carnegie Museum of Natural History. Pittsburgh, Pennsylvania Dr. John E. Rawlins INHS Illinois Natural History Survey. 607 East Peabody Drive, Champaign, IL Dr. Donald W. Webb. ISNB IZAC Institut Royal des Sciences Naturelles de Belgique, 29 rue Vautier, Brussels, Belgium. Instituto de Zoologia, Academia de Ciencias de Cuba, Habana, Cuba. JARC Personal collection, Dr. J. A. Ramos. Box 1046, Mayaguez, Puerto Rico MAIC Personal collection, Dr. Michael A. Ivie. Department of Entomology, Montana State University, Bozeman Montana MCZC Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts Dr. Scott R. Shaw MHMB Naturhistorisches Museum. CH-4001 Basel, Augustinergasse 2, Switzerland. Dr. Michael Brancucci MIZA Museo Instituto Zoologia Agricola, Universidad Central de

31 11 Venezuela, Maracay, Venezuela. Dr. J. Luis Joly T. MNHP Musee National d histoire Naturelle, Paris. 45, Rue Buffon, Dr. Jean J. Menier NMNH OSUC RSMC TAMU UCDC National Museum of Natural History, Smithsonian Institution, Washington, D. C. Drs. Robert Gordon and Paul Spangler. The Ohio State University Collection of Insects and Spiders, Department of Entomology, Columbus, Ohio. Dr. Charles A. Triplehorn Collection of author. Texas A. & M. University Entomological Museum, Department of Ent ^ology, College Station, Texas Dr. Lorace R. Burke. University of California, Davis Dr. Robert Schuster. UPRM University of Puerto Rico, Mayaquez, Puerto Rico Mr. Rafael Ingles ZMHB Zoologisches Museum, Museum fur Naturkunde, der Humboldt - Uni versitat zu Berlin, DDR 104, Berlin. Dr. Manfred Uhlig I would also like to thank the following individuals for material that was used in this study and that can now be found in the RSMC: Mr. William F. Abeles, Mr. Patrick Bleuzen, Father Carlo Brivio, Dr. Fortun6 Chalumeau, Dr. Shawn Clark, Dr. Paul S. Cwikla, Dr. Phillip DeWailly, Dr. Michael A. Ivie, Dr. Norman F. Johnson, Mr. Peter W. Kovarik, Dr. John F. Lawrence, Drs. Charles and Lois O'Brien, Dr. Luis E. Pena G,, Dr. Andrew Penniman, Dr. John A. Shuey, Dr. James B. Stribling, Dr. Barry D. Valentine, and Dr. Charles Withrow. Material for the study of Thvlodrias contractus Motschulsky was from my private culture supplemented by specimens from a culture maintained by Dr. Barry D. Valentine.

32 ONTOGENY AND PHYLOGENY "We still do not know the mechanics of evolution in spite of the over-confident claims in some quarters, nor are we likely to make much progress in this by classical methods of paleontology or biology -White, 1966:8. "The most incomprehensible thing about nature is that it is comprehensible" Albert Einstein Or, that we imagine it comprehensible. While attempting to resolve phylogenetic relationships within the Leptolycini (Miller, 1991), a paedomorphic lineage within the Lycidae (Coleoptera ), I encountered unusually high rates of homoplasy and inexplicable biogeographic patterns that could not be resolved by usual techniques. Hypothesizing that the anomalous development of the leptolycines might be causal, I studied another paedomorphic species, Thvlodrias contractus Motschulsky (Dermestidae), which is more amenable to laboratory culturing. Additionally, the familial and superfami1ial relationships of the latter species is more strongly supported than the relationships within the Cantharoidea. The results revealed unexpected assumptions about paedomorphosis, why these implicit assumptions might lead to error, and possible modifications of phylogenetic analysis that could alleviate this problem. During phylogenetic analysis, systematists attempt to minimize 12

33 13 assumptions to those that are congruent with evolution as it is currently understood. Since von Baer's (1Q28 ) pioneering studies of embryology, innumerable attempts have been made to incorporate ontogenetic data while examining evolutionary relationships (Gould, 1977). However, since the repudiation of Haeckel s views of ontogeny and phylogeny, the study of ontogenetic and evolutionary processes has remained largely separate (Raff and Kauffman, 1983:355) and the relationship of these processes little studied. The relationship of ontogeny and phylogeny within the Coleoptera is no exception. Ontogeny has not been used extensively in beetle systematics, primarily because so little is known of the immature stages. Newton (1990:208 ) has estimated that only 15 7, of the Staphylinoidea, which comprises about one fifth of all described beetles, are known in the immature stages at the specific level. Although the extent of our knowledge of other lineages varies, Newton's estimate approximates the level of study throughout the Coleoptera. Additionally, the embryological literature is at best contradictory (Crowson, 1981:361) and postembryonic development within the Coleoptera has been examined only in a limited number of species. Even the relatively few descriptions of larvae are often inadequate for recognition. This needs to be rectified, but indubitably will require many years of study. This ignorance in turn may impede understanding of diverse developmental phenomena within the Coleoptera such as paedomorphosis (neoteny of authors, see below). Paedomorphosis occurs in a number of lineages within the order. The best known examples are in the fireflies and glowworms of the former

34 14 Cantharoidea, now Elateroidea (Lawrence, 1987)*. However, the phenomenon has been documented or is here hypothesized in the Archeostemata CMicromalthidae, Scott (1938)3, and several superfamilies within the Polyphaga including Staphylinoidea CStaphy1inidae, Aleocharinae, Crowson (1981:389)3; Histeroidea CHisteridae, Crowson (1974, 1981:389)3; Dascilloidea CKarumiidae3, Elateroidea CElateridae, Cebrionidae3; Cantharoidea sensu strictu CLampyridae, Phengodidae, Drilidae, Lycidae, and Omalysidae, numerous authors3; Cleroidea CMelyridae, eg. Apteroma1achius namibensis Wittmer, 19603; Bostrichoidea CDermestidae, eg. Thvlodrias contractus Motschulsky!; Tenebrionoidea CRhipiphoridae, eg. Rhioidius Thunb., studied by Besuchet (1956); and Meloidae, eg. Hornia Riley studied by Linsley (1942)3; and Curculionoidea CScolytidae, Crowson (1981:390)3. In addition, there are many other taxa that display paedomorphic tendencies that remain unstudied. Many of these paedomorphic lineages, if not most, appear to have had independent origins. Because paedomorphic, "larviform" females are widespread within the Cantharoidea and are proposed to be basal in several lineages within the superfamily (Crowson, 1972:62), the phenomenon of paedomorphosis must be examined before any meaningful discussion of these lineages is possible. This is prerequisite because the effect of paedomorphosis on character states is unknown. Hypothesizing paedomorphic ancestors within the Cantharoidea, however, creates anomalies that require ad hoc explanation (Crowson, 1972). In addition to possessing hypothesized paedomorphic ancestors, the Cantharoidea are considered to express basal character states

35 15 secondarily within the Coleoptera (Crowson, 1967, 1972, 1981; Geisthardt, 1977; Lawrence, 1982, 1987; Lawrence and Newton, 1982). If this secondary expression of basal states is related to a paedomorphic ancestor, it may have application to other lineages within the Coleoptera. Such convergence to apparent basal character states may produce misleading results during phylogenetic analysis if the phenomenon is not addressed. Again, the alternative possibility that these taxa, assumed to be secondarily "primitive," are in fact basal, should not be rejected without reexamination. Even though knowledge of paedomorphosis is essential to understanding several, if not many, lineages within the Coleoptera, it has remained unknown except from an intuitive morphological perspective. Besides the original descriptions, the literature consists of redescriptions of paedomorphic forms such as the female Omalvsus fontisbellaquei Geoffroy (Geisthardt, 1977 ). Two major exceptions are the excellent and extensive morphological studies of Geisthardt (1974 and 1979 ) of adult Lamprohi za splendidula (L. ) and Lampvris noctiluca L. However, Geisthardt did not discuss ontogenetic or evolutionary processes in relation to morphology in either paper. Cicero (1988) attempted to study the phenomenon using a restricted data set of only females, but this analysis suffered from a number of problems (Miller, in prep. ) and those discussed below. The physiological aspects of paedomorphosis have been studied only in Lampvris noctiluca L. Naisse (1966, 1968, and 1969) found that sexual dimorphism was the result of differential titers of juvenile hormone during development. She also demonstrated that the testes,

36 16 which produce an androgenous hormone, are responsible for male character states and will masculinize a female when experimentally placed in the abdomen of a female larva. Davydova's (1967) ablation of the corpora allata and the corpora cardiaca in males of the same species failed to produce wingless females and her experimental implantation of 2-6 of these glands in female larvae did not induce wing formation in females. She did not report any effect on the gonads. A correlation of paedomorphosis with the X-0 sex determination within lineages has been noted (Crowson, 1972), but Smith and Virki (1907) indicate the loss of the Y - chromosome is widespread throughout the Coleoptera and occurs in many lineages where paedomorphosis is unknown. This karyotype constitutes 14 of the examined species in the order (loc. c i t.. p. 90) and their origins often appear independent. However, the X-0 system may be necessary, but not sufficient for paedomorphosis, and requires further examination. Interestingly, I have found no records of karyotype studies of actual paedomorphic species. Although little is known about comparative morphology and physiology of paedomorphosis in beetles, we know even less of it as a function of developmental or evolutionary processes. As a result, systematic examination of paedomorphic species has created a number of enigmata. For example, Crowson (1972:49) hypothesized that the character states found in the lycids Platerodrilus Pic, Lvropaeus Waterhouse, and Puliticola Mjorberg might be basal in the Lycidae. The latter is known to have flightless, paedomorphic females and females are unknown in the other genera, but are probably also paedomorphic. Why,

37 then, do most presumably derived lycid genera have fully flighted males 17 and females? He (loc. cit.. p. 64) implied that the male might carry the necessary genes that were reincorporated into the females of derived lineages similar to the tarsomeres of Cryptophagidae. Clearly, he was troubled with his conclusions and flagged this problem for future cantharoid workers (Crowson, 1972). Geisthardt (1977) questioned Crowson's concerns, basing his conclusions on the widespread occurrence of paedomorphosis in some lampyrid genera. Unfortunately, he did not support his argument with a phylogenetic hypothesis demonstrating the non-derived nature of the lampyrids discussed. More recently, Cicero (19QQ) examined ontogeny and evolution within the Cantharoidea. In so doing, he proposed a discrete classification of developmental stages for female adult beetles, assumed the basal origin of paedomorphosis (his neoteny ) within a restricted Cantharoidea (Lampyridae, Phengodidae, Drilidae, but not Cantharidae, etc. ) that supports the hypotheses of Crowson (1972) and Geisthardt (1977). While such an examination of the character states of adult females alone can be fruitful, its implications require further study. For instance, 1 ) a classification of developmental stages can offer insight into process, but should not be accepted hastily because a discrete classification is arbitrary and, possibly, misleading if the phenomenon is in fact continuous. Current knowledge of ontogenetic processes has not clarified the nature of development over evolutionary time and, hence, it is unknown whether it is discrete or continuous. 2) Ignoring male morphology in favor of female morphology is no more

38 18 efficacious than studying only males. More importantly, 3) the homology of developmental character states must be demonstrated and not assumed. If homology is assumed and, again, the series is assumed to be a continuum from one extreme to the other, there is an enhanced possibility of succumbing to the Haeckelian argumentation of "ontogeny recapitulating phylogeny - an idea that was rejected four decades ago (see Gould, 1977 for an extensive review; Miller, in prep. ). Such an approach can prove disastrous when character states within lineages are unexamined and assumed to be homologous throughout the entire family without a phylogentic analysis (See Aiberch, 1985 for a discussion of the dangers of Haeckelian recapitulation in systematics ). I here suggest that implicit Haeckelian argumentation remains the major hurdle to understanding phylogenetic relationships within the Cantharoidea and that it has yet to be addressed. Because of the central nature of this problem and because of recurrent problems in polarization of character states within the Lycidae, a group of particular interest to me, study was initiated to examine this process and its phylogenetic implications. Definition of Problem. Gould (1977:484) defined paedomorphosis as "the retention of the ancestral juvenile characters by later ontogenetic stages of descendants." Aiberch et a l. (1979) presented models to define and explain this and other forms of heterochrony. Because of the precision of their definitions and because it is not possible to differentiate paedomorphosis further (ie. neoteny or progenesis) without supporting developmental data, only their more inclusive term will be

39 19 used in this discussion. Another term currently used in discussions of paedomorphosis is "larviform". It is, however, imprecise. Very few adult beetles are essentially indistinguishable from the larvae (eg. all known female Phengodidae, some Micromalthus. and Puliticola of the Lycidae). Most are similar to their larvae in only a few characters or in their general facies. Thus, "larviform" is an inappropriate adjective for all but a very few beetles and the other imagoes would more properly be called paedomorphic adults. A number of character states are commonly associated with paedomorphosis within the Coleoptera. They may be present in any combination, but their expression is normally manifest in extreme sexual dimorphism. Probably the most commonly used character for recognition of the phenomenon is the loss or reduction of wings and elytra. Additionally, the "retention" of more abdominal segments is often noted. Occasionally, the externally undifferentiated meso - and metanotum are discussed. These are comparative characters that require recognition of the opposite sex or sister taxa to evaluate. However, any one such character is usually not a necessary or sufficient condition to assume paedomorphosis. Given this imprecise means of identification, how may paedomorphosis be differentiated from other evolutionary phenomena? A putative example of paedomorphosis from the dermestid lineage, Thvlodrias contractus Motschulsky, was examined to clarify possible trends in character states due to the phenomenon. Adult Thvlodrius contractus was first described (Motschulsky,

40 1839:75) as a malacoderm near Malachius Fabricius (now in the Melyridae, Cleroidea). Reitter (1894) moved it to the Drilidae which is currently placed in the Cantharoidea. Discovery of the larvae revealed that it was in fact a dermestid (Hinton, 1945:260). Dermestids are placed currently in the Dermestoidea of Crowson (19Q1) and Lawrence (1982) and, most recently, Ivie (1985) has placed the Dermestidae in the Bostrichoidea. Both hypotheses are in agreement that the sister of the Dermestidae is the Nosodendridae and that other close relatives include the Bostrichoidea / Bostrichidae. Neither of these hypotheses place these taxa in the Elateriformia where the Drilidae (Cantharoidea) is currently placed. Even though the Dermestidae is probably known better than most families because of its economic status, its higher classification is in need of review. Currently, Thvlodrias is placed in its own subfamily (Crowson, 1967:74; Mroczkowski, 1968:158; Suss and Fogato, 1978), although it also has been given monotypic tribal status (Beal 1959:99) as well as superfami1ial status (Franciscolo, 1975). In all these cases the differences are emphasized, but no relationship is evident. Hinton (1945:258) suggested that Thvlodrias is most closely related to Trinodes based on examination of larvae. Although I have not seen larvae of the latter genus, the literature seems to support his conclusion. In the absence of a phylogenetic analysis of the Dermestidae, the sister group relations of Thvlodrias remain obscure. Therefore, general character states within the family and its hypothesized sister group were examined and compared with those in Thvlodrias males and

41 21 females (Table 1 ). As can be seen, several characters are significantly affected by paedomorphosis. Sclerotization. The degree of sclerotization is less than other dermestids and nosodendrids examined. This lightly sclerotized cuticle is characteristic of other paedomorphic beetles, although it varies in degree. Historically, this character state was used as a means of grouping taxa. The constituents of the Malacodermata, literally meaning "soft skin" (Gk. ), included most of the lightly sclerotized beetles at some point in the history of beetle classification and all of those currently placed in Cantharoidea. Compactness. The 'coadaptation' of various sclerites (Crowson, 1981:60) is significantly less than found in other dermestids examined or in the Nosodendridae. It is difficult to ascertain if other paedomorphic species are less compact, as their presumed relatives are also not as compact as many beetles. Sexual Dimorphism. The sexual dimorphism of Thvlodr ias is much greater than found in other Dermestidae. In fact, unlike the odd beetle, it is often necessary to dissect specimens to determine sex in the other members of the family. Again, the sex of nosodendrids are not as apparent externally, but sexual dimorphism in other studied paedomorphic species is pronounced to extreme. Antennae. The antennae in both sexes of Thvlodrias are more elongate and the club is less compact than in other dermestids and nosodendrids examined. There are fewer antennomeres (10 in males and 9 in females) and the male club is extremely elongate, but some dermestids

42 22 Table 1. Character states found in Thvlodrias contractus Motschulsky males and females compared to the general states found in other Dermestidae and Nosodendridae. Character T. contractus Other Dermestidae females males (and Nosodendridae) Sclerotization 1 ight 1ight heavy Coadaptation loose 1oose compact Retraction of head no no yes Shape of head Globose Elongate Elongate Elongation of 2-6 some more short Antennomeres Number eye facets reduced reduced? not reduced Cryptop1euron unchanged unchanged unchanged Cavities for front no no yes leg retraction Retractibi1ity of no no yes front legs Prosternal carina absent absent present Mesonotal incomplete complete complete differentiation Scutellum caudally fused free free Elytron absent present, present, elongate oval to obovate Metasternal carina absent absent present (absent) Metanotal 1ittle complete complete differentiation Intercoxa1 distance very wide normal normal of metacoxae Metendosterni te strongly weakly normal modified modified Sclerotization of light heavy heavy Metepimeron Hindwing absent present, present +/- reduced Metafemoral cavity absent absent present Ratio Trochanter- normal elongate normal femur length Abdominal Sternites 7 exposed 7 exposed 5 exposed Sexual Dimorphism pronounced pronounced siight

43 23 (eg, Thorictodes Reitter) have fewer antennomeres also. Examination of the antennae of both sexes of Lampvri s noctiluca L. did not reveal significant size differences, but Phengodes females have larval-type antennae. Again, the position of antennal placement on the frons is unaffected in both sexes of T. contractus. This placement in the Coleoptera appears to be restricted to a track between a ventral position over the mandibular bases and a median position found near the vertex in taxa such as Chrysomelidae, Lampyridae and Lycidae. The nosodendrid and dermestid condition, including Thvlodrias. is ventral near the mandibular bases. Other examined paedomorphic species with an adult antennal form are likewise unaffected. Compound Eyes. The number of eye facets is reduced in the female and appears variable, but a larger sample size is indicated. The male eye facet number may also be slightly reduced, but the taxonomic literature of dermestids does not discuss ommatidial number. The few other dermestids available for study had more facets than Thvlodrias. Lampvris noctiluca females have significantly fewer ommatidia than males and Phengodes females have no compound eyes. Prosternal carina. All dermestids examined have a carina that extends the length of the intercoxal process, except it is absent in Thvlodrias males and females. The carina is also present in Nosodendron. but not other paedomorphic species examined, nor their probable sister groups. Cryptopleuron. Only a few genera were examined, but the cryptopleuron appears to be unaffected by the phenomenon in both males