Megaloptera : Corydalidae
Total Page:16
File Type:pdf, Size:1020Kb
Systematic Entomology (1981) 6, 253--290 Systematics of the dobsonfly subfamily Corydalinae (Megaloptera :Corydalidae) MICHAEL J. GLORIOSO* Department of Entomology, Ohio State University, Columbus, Ohio, U.S.A. ABSTRACT. The genera of Corydalinae are redefined, and representative characters are figured for each genus. New character sources, such as mouthparts and internal female genitalia, are investigated, as well as traditional male genitalia and wings. Allohermes is synonymized with Protohermes, Doeringia with Platy- neuromus. Intergeneric relationships are hypothesized on the basis of a cladistic analysis. Acanthacorydalis and the New World genera form a monophyletic group, as do Protohermes and Neurhermes, and Neuromus and Neoneuromus. Chloroniella belongs in the Acanthacorydalis - New World lineage, but exact placement is uncertain. A phyletic sequence classification is proposed on the basis of the cladistic analysis. Introduction American Covydulus cornutus are excellent angling bait, while dried larvae of Pvotoizevmes The Megaloptera, long considered among the grandis, referred to as 'magotaro mushi', were most primitive of holometabolous insects considered a remedy for infant emotional (Weele, 191 O), contains two families, Sialidae irritation in Japan (Kuwayama, 1962). and Corydalidae. The Corydalidae, easily Relationships within the Corydalinae are separated from the Sialidae by presence of poorly understood, and generic limits are ocelli, non-bilobed fourth tarsomeres, and poorly defined. In this study I redefine the large size, contains two subfamilies, Coryda- genera and postulate intergeneric relation- linae and Chauliodinae. The Corydalinae are ships based on shared derived character states, restricted to North and South America, South or synapotypies. In searching for sufficient Africa and Asia, while the Chauliodinae also characters to more confidently establish occur in Australia and New Zealand. relationships I also hope to generate enough Adult Corydalinae are among the largest characters and knowledge of character states and most bizarre appearing of living insects, to eventually resolve intrageneric relationships. males of Ac~ur~thucorydu1l.uand ('orydalu~ being noted for disproportionately large Tuxonomzc lz~story mandibles. The formidable appearance of Davis (I 903) divided the Sialidae into the Acanthacorydalis has led at least one collector Sialidinae and Corydalinae, and Weele (1909) to regard this insect as extremely dangerous divided the Corydalinae into two tribes, (McLachlan, 1899). Neuromini and Chauliodini. Some authors While not as spectacular, the larvae may be (Esben-Petersen, 1924; Barnard, 1931 ; of specialized importance. Those of the North Kimmins, 1948; Kuwayama, 1962) recog- * Editors' note: We record, with regret, the un- nized Neuromini, while Tillyard (19 18) and timely death of Ilr Michael J. Clorioso on 17 Octoher many subsequent authors called this group the 1980, and thank Professor C. A. Triplehorn and Dr 0. S. Flint for their help with proof reading. Corydalinae. Lestage (1927) noted the name should be based on reducing Correspondence: Professor Charles A. Tripleh~n, C'orydulus, Department of Entomology, Ohio State University, Neuromini to a junior synonym. 1735 Neil Avenue, Columbus, Ohio 43210, U.S.A. Some early authors (Tillyard, 1918; 0307-6970/81/0700-0253 $02.00 O 1981 Blackwell Scientific Publications 751 Lestage, 1927; Barnard, 193 1) considered the 2200 specimens of Corydalinae and Chaulio- Sialidae and Corydalidae distinctive enough to dinae borrowed from museum and private merit familial status, with Weele's tribes ele- collections. Individuals and institutions pro- vated to subfamilies, while others followed viding material for study were: P. C. Barnard, Weele's categorical rankings (Banks, 1940, British Museurn (Natural History), London, 1943; Kimmins, 1954). More recently (e.g. England (BMNH); D. H. Kavanaugh, Chandler, 1956; Flint, 1973; Evans, 1978) the California Academy of Sciences, San Sialidae and Corydalidae have been considered Francisco, California (CASC); G. Ekis, separate families. Carnegie Museum of Natural History, Weele (1910) was the first to revise the Pittsburgh, Pennsylvania (CMNH); J. E. H. Corydalinae, recognizing eight genera and Martin, Canadian National Collection of thirty-four species and menhoning possible Insects, Biosystematics Research Institute, relationships. Navas (1 9 15) proposed Ncur- Ottawa, Ontario (CNCC); L. L. Pechuman, hermes to replace Gray's Hermrs, homony- Cornell University, Ithaca, New York mous with a gastropod genus, and Esben- (CUNY); J. Kethley, Field Museum of Natural Petersen (1 924), Navhs (1 925) and Lestage History, Chicago, Illinois (FMNII); L. A. (1927), proposed additional genera. Lestage Stange, Florida State Collection of Arthro- (1927), in his catalogue of Indo-Chinese Cory- pods, Gainesville, Florida (FSCA); N. D. dalinae, included a list of world genera and Penny, Instituto Nacional de Pesquisas da postulated affinities based on wing venation. Amazonia, Manaus, Brazil (INPA); L. Since 1927 work on Corydalinae has been Dieckman, Institut fiir Pflanzenschutzfor- mainly species description, though Banks schung Kleinmachnow, Eberswalde, DDR (1940) included a key to Chinese genera, (IFPK); C. L. Hogue, Los Angeles County Kimmins (1948) reviewed the Protohc~rmesin Museum of Natural History, Los Angeles, the British Museum, and Penny (1977) listed California (LACM); M. M. Pearce, Museum the New World species. of Comparative Zoology, Ilarvard University, Cambridge, Massachusetts (MCZC); M. J. Glorioso, Columbus, Ohio (MJGC); J. With the exception of Corydalus c.orrzutu.s, Legrand, MusCum National dlHistoire the natural history of Corydalinae is poorly Naturelle, Paris, France (MNHN); C. A. known, though the larvae are probably all Triplehorn, Ohio State University, Entomo- lotic inhabitants. Larval lifespan varies from 1 logy, Columbus, Ohio (OSUC); N. W. Britt, to 5 years (Evans, 1978), this being correlated Columbus, Ohio (NWBR); P. A. Adams, with latitude in C.cornutus (Brown & Fullerton, California (PAAC); G. W. Byers, Fitzpatrick, 1978). Some species have been Snow Entomological Museum, University of observed occupying specific microhabitats Kansas, Lawrence, Kansas (SEMK); S. J. within a stream (Geijskes, personal communi- Merritt, Texas, A & M University, College cation), while label data indicate a correlation Station, Texas (TAMU); W. F. Barr, Univer- between species and stream type or altitude. sity of Idaho, Entomology, Moscow, Idaho Adults are crepuscular or nocturnal and (UIDC); H. P. Brown, University of readily collected at lights. Though some Oklahoma, Zoology, Norman, Oklahoma evidence exists indicating feeding on liquids (UOKL); 0. S. Flint, Jr, P. J. Spangler, United (Parfin, 1952), adult life is generally short, States National Museum of Natural History. ranging from 3 to 13 days for C.cornutus Entomology, Smithsonian Institution, (Davis, 1903; Parfin, 1952). Adult emergence Washington, D.C. (USNM); K. K. Gunther, of some species is seasonal in certain regions, Zoologisches Museum, Humboldt-Universitat, while little seasonality is observed in other Berlin, Germany (ZMHU). localities. Seasonality may be influenced by temperature or rainfall, but also occurs in Methods regions with no distinct seasons. Criteria for gerzera delimitution Material Mayr (1969) considers a genus to be a This study is based on examination of about monophyletic category containing a single Systemutlcs oj the dobsonfly subfam~lyCorydal~nae 255 species or group of species separated from However, the splitting of genera only on the other genera by a decided gap. I likewise basis of number of species should be conser- believe genera should be monophyletic, but vative to avoid proliferation of genera. in the sense of Hennig (1965, 1966), and, In this study several autapotypies involv- ideally, genera should be recognizable by dis- ing male genitalia define distinct lineages, continuities in character states. As the degree with no intergradation between lineages. As of difference between genera or other supra- these monophyletic units are also recognizable specific taxa is subjective, criteria for recog- on the basis of other characters, I propose nition of supraspecific categories should be that each be given generic status. explicitly stated to allow subsequent workers to judge the validity of such taxonomic deci- Phylogrnefzc and class~jccatorymethods sions (Ekis, 1977b). Before supraspecific classifications are My phylogenetic hypotheses and classifi- developed, worldwide knowledge of the cations are based on the results of a cladistic investigated taxon should be available (Erwin, analysis, the principles of which were outlined 1975). This requires investigation of inter- by Hennig (1 965, 1966). Phylogenetic classifi- specific, but not intraspecific variation cations, including phyletic sequencing, a (Herman, 1970), and an attempt should be means of deriving a classification from an made to examine all species of the larger asymmetrical branch of a cladogram without taxon before inferring generic limits. While superfluous supraspecific categories, have knowledge of all interspecific relationships been discussed by Wiley (1979b). Proponents is not essential, major monophyletic lineages of phyletic sequencing include Nelson (1972, based on synapotypies must be defined, as 1974), Cracraft (1974), Schuh (1976) and well as the relationships of each lineage to the Wheeler (1979a, b). others. After these relationships