PROCEEDINGS OF THE California Academy of Sciences Vol. 42 SAN FRANCISCO PUBLISHED BY THE ACADEMY 1979-1982 COMMITTEE ON PUBLICATIONS Tomio Iwamoto, Chairman and Editor Frank Almeda Daphne F. Dunn William N. Eschmeyer Frank H. Talbot (US ISSN 0068-547X) The California Academy of Sciences Golden Gate Park San Francisco, California 94118 PRINTED IN THE UNITED STATES OF AMERICA BY ALLEN PRESS, INC., LAWRENCE, KANSAS CONTENTS OF VOLUME 42 Pages No. 1. WILDER, D. DEE. Systematics of the Nearctic Ptilodexia Brauer and Bergen- stamm (Diptera: Tachinidae). Published March 2, 1979 1-55 No. 2. McCosKER, JOHN E. The snake eels (Pisces, Ophichthidae) of the Hawaiian Islands, with the description of two new species. Published March 2, 1979 57-67 No. 3. CHEMSAK, JOHN A., AND E. G. LINSLEY. Review of the Rhinotragini of Mexico (Coleoptera: Cerambycidae). Published March 2, 1979 69-85 No. 4. KAVANAUGH, DAVID H. Studies on the Nebriini (Coleoptera: Carabidae), III. New Nearctic Nebria species and subspecies, nomenclatural notes, and lec- totype designations. Published December 22, 1979 85-133 No. 5. IWAMOTO, TOMIO. Eastern Pacific macrourine grenadiers with seven branchi- ostegal rays (Pisces: Macrouridae). Published December 22, 1979 135-179 No. 6. FRITZSCHE, RONALD A. Revision of the eastern Pacific Syngnathidae (Pisces: Syngnathiformes), including both Recent and fossil forms. Published July 2, 1980 181-227 No. 7. BRADBURY, MARGARET G. A revision of the fish genus Ogcocephalus with descriptions of new species from the western Atlantic Ocean (Ogcocephalidae; Lophiiformes). Published July 2, 1980 229-285 No. 8. Poss, STUART G., AND WILLIAM N. ESCHMEYER. Xenaploactis , a new genus for Prosopodasys asperrimus Giinther (Pisces: Aploactinidae), with descrip- tions of two new species. Published July 2, 1980 287-293 No. 9. ROBERTS, TYSON R. Sundasalangidae, a new family of minute freshwater salmoniform fishes from Southeast Asia. Published March 5, 1981 295-302 No. 10. ALMEDA, FRANK. New and reconsidered species of Miconia (Melastomata- ceae) from Costa Rica and Panama. Published March 5, 1981 303-314 No. 11. BRIGGS, THOMAS S., AND DARRELL UBICK. Studies on cave harvestmen of the central Sierra Nevada with descriptions of new species of Banksula. Published June 24, 1981 1 315-322 No. 12. LINDBERG, DAVID R., AND JAMES H. MCLEAN. Tropical eastern Pacific lim- pets of the family Acmaeidae (Mollusca, Archaeogastropoda): generic criteria and descriptions of six new species from the mainland and the Galapagos Islands. Published June 24, 1981 323-339 No. 13. SMALL, GREGG, J. A review of the bathyal fish genus Antimora (Moridae: Gadiformes). Published June 24, 1981 341-348 No. 14. FRICKE, RONALD. The kaianus- group of the genus Callionymus (Pisces: Callionymidae), with descriptions of six new species. Published October 26, 1981 349-377 No. 15. ROTH, BARRY. Distribution, reproductive anatomy, and variation of Monade- nia troglodytes Hanna and Smith (Gastropoda: Pulmonata) with the proposal of a new subgenus. Published October 26, 1981 379-407 [iii] No. 16. JONES, ROBERT E. Food habits of smaller marine mammals from northern Cali- fornia. Published October 26, 1981 409-433 No. 17. KAVANAUGH, DAVID H. Studies on the Nebriini (Coleoptera: Carabidae), IV. Four new Nebria taxa from western North America. Published October 26, 1981 435-442 No. 18. GRYGIER, MARK J. Dendrogaster (Crustacea: Ascothoracida) from California: sea-star parasites collected by the ALBATROSS. Published May 14, 1982 443-454 No. 19. ERWIN, TERRY L. Small terrestrial ground-beetles of Central America (Carab- idae: Bembidiina and Anillina). Published May 14, 1982 455-496 No. 20. ROBERTS, TYSON R. The Bornean gastromyzontine fish genera Gastromyzon and Glaniopsis (Cypriniformes, Homalopteridae), with descriptions of new species. Published May 14, 1982 497-524 No. 21. ZULLO, VICTOR A., AND RAJ B. GURUSWAMI-NAIDU. Late Miocene balanid Cirripedia from the Basal Wilson Ranch beds ("Merced" Formation), Sonoma County, northern California. Published May 14, 1982 525-535 Index to Volume 42 _ 537-550 tiv] PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES Vol. 42, No. 1, pp. 1-55; 113 figs. March 2, 1979 SYSTEMATICS OF THE NEARCTIC PT1LODEXIA BRAUER AND BERGENSTAMM (DIPTERA: TACHINIDAE)* By D. Dee Wilder Department of Entomology , California Academy of Sciences, Golden Gate Park, San Francisco, California 941 18 ABSTRACT: A revised classification of the Nearctic prosenine genus Ptilodexia Brauer and Bergcnstamm (Dip- tera: Tachinidae) is presented. A total of 8,000 specimens and type material for nearly all species were studied. All seventeen previously described valid species of Ptilodexia are diagnosed and illustrated. Five new species, P. sabroskyi, P. pacifica, P. californica, P. westi, and P. maculata are described and illustrated. A key to the Nearctic species is presented. The following new synonymies are made: P. conjuncta (Wulp) (= Rhynchodexia simulans Wulp); P. contristans (Wulp) (= K. punctipennis Wulp); P. carolinensis Brauer and Bergenstamm (= P. neotibiaKs West, P. minor West); P. halone (Walker) (= P. hucketti West); P. harpasa (Walker) ( = P. leucoptera West, Dinera robusta Curran); P. rufipennis i Mai-quart ) (= Drxia cerata Walker, D. albifrons Walker, Rhynchodexia confusa West, K. translucipennis West, Rhynchodexia dubia Curran); P. major (Bigot) (= Dexiosoma fumipennis Bigot, Rhynchodexia fraterna Wulp, R. omissa Wulp); P. incerta West (= P. proximo West; Rhynchodexia elevata West). The biology of these parasitic flies is reviewed and possible host-parasite relationships are discussed. The taxonomic significance of numerous morphological characters in the genus and the subfamily is discussed. The phytogeny of Ptilodexia and its nearest relatives is discussed; six species groups are separated, and an evolu- tionary tree presented for these groups. The contemporary and historical zoogeography of the genus is discussed as it pertains to host and parasite distribution. A distribution map is presented for each species treated. INTRODUCTION larvae of certain scarabaeid beetles. They are of Flies of the genus Ptilodexia are large calyp- economic interest because they are known par- trate Diptera belonging to the Tachinidae, a fam- asites of such Pests as PhyUophaga spp., Po- ily of exclusively parasitic flies. Ptilodexia Pillia J<*P°"<™ Newman, and Macrodactylus adults are commonly collected on flowers during subspinosus (Fabncius). The genus is distribut- the summer months. The larvae parasitize the ed throughout the New World. Although specimens of Ptilodexia are com- mon in collections, few are correctly identified. * Contribution from the Frost Entomological Museum, Sabrosky and Arnaud (1965) made no attempt Pennsylvania State University. This research was supported to give synonymies Or distributions in their cat- by the Agricultural Experiment Station Project No. 2070, and alog listing of 27 species names. The special constitues Contribution No. 561 1 from that station. An earlier problems leading to such confusion in Ptilodexia version of this paper was submitted to the Graduate School, ^ j} djfficu,t jn associating t.he S6X6S, (2) a Pennsylvania State University, as a thesis in partial fulfill- . ment of the requirements for the degree of Doctor of Philos- h'gh degree of mtraspecific variation, (3) an UH- ophy. usually low degree of interspecific variation, and [1] PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 42, No. 1 (4) the lack of consistent traditional morpholog- ical characters. The purpose of this study is to revise the clas- sification of Ptilodexia on the basis of all avail- able material, to analyze the life history of its species, and to determine evolutionary and zoo- geographical trends which may apply also to other members of the family Tachinidae. MATERIALS This study was based on over 8,000 specimens borrowed from various museums. In addition, type-specimens for most of the known species and numerous representatives of other genera in the Prosenini were studied. Institutions loaning material used in this study were as follows: American Museum of Natural History (AMNH), Arizona State University (ASUT), British Museum (Natural History) (BMNH), California Academy of Sciences (CASC), University of California Berkeley (CISC), Canadian National Collection (CNCI), Cornell University (CUIC), University of Ne- braska (DEUN), Field Museum of Natural His- tory (FMNH), Florida State Collection (FSCA), Iowa State University (ISUI), Los Angeles County Museum of Natural History (LACM), Leningrad Museum of Natural History (LMNH), Museum of Comparative Zoology, Harvard Uni- versity (MCZC), Michigan State University (MSUC), Ohio State University (OSUC), Okla- homa State University (OSEC), Oregon State University (OSUO), Paul H. Arnaud, Jr., Col- lection (PHAC), Yale University (PMNH), Pur- due University (PURC), South Dakota State University (SDSU), University of Kansas (SEMC), Staten Island Institute of Science (SIIS), University of Oklahoma (SMSH), Texas A & M University (TAMU), University of Ari- zona (UAIC), University of Alberta (UASM), University of California Davis (UCDC), Uni- versity of California Riverside (UCRC), Univer- sity of Idaho (UICM), University of Montreal (UMIC), University of Michigan (UMMZ), Utah State University (USUC), National Museum of Natural History (USNM), Vienna Museum of Natural History (VMNH), Washington State University (WSUC). METHODS The male genitalia of Ptilodexia species are partially obscured on dry, pinned specimens. To examine them, the posterior half of the abdomen was removed, placed in a solution of 10 percent KOH, and heated until the structures were suf- ficiently softened to be dissected easily. They were later rinsed twice with water and twice with acetic acid, placed in glycerine, and ex- amined. They were stored in a microvial pinned beneath the insect. Illustrations of the genitalia were made using an ocular grid. The postabdomen was anchored to a small piece of soft wax on the bottom of the dish of glycerine to prevent it from drifting about. Drawings of the heads were made by project- ing photographic transparencies of them onto drawing paper. Manipulation of the projector provided images of uniform size. The image was then traced with a hard pencil; the details were filled in after thorough examination of the spec- imen with a dissecting microscope. All measurements were made using an ocular grid, calibrated by a stage micrometer. Ratios were calculated from these measurements. Because of the extreme intraspecific variabil- ity of Ptilodexia, a description which included all variation would be unwieldy and would be similar for each species. To make the descrip- tion more useful and manageable in size, only one specimen, the holotype, is described. A complete synonymy and list of citations are given for each species included in this study. New species are thoroughly described and di- agnosed. For previously described species, only a diagnosis is presented. Known information on each species is summarized and notes regarding types and nomenclature are presented. Intraspecific variation is discussed thoroughly for all new species. With previously described species, variation is discussed only where it is necessary for species identification. Complete data from each specimen, including sex, locali- ty, collector, depository, and other information, have been recorded by Wilder (1976) and there- fore are not presented here. Wilder (1976) also gives complete redescriptions for all previously described species of Ptilodexia. The scope of this revision has been limited for practical reasons. Inclusion of the southern Mexican species would have doubled the num- ber of species treated, and the material available for these species is wholly inadequate. All Nearctic species are treated herein. WILDER: NEARCTIC PTILODEXIA ACKNOWLEDGMENTS I extend my sincerest appreciation to Dr. C. W. Sabrosky, who initiated this study and with- out whose help this revision could not have been completed. Dr. Sabrosky also generously loaned types and was always available to answer ques- tions about Ptilodexia and related tachinids. I thank Dr. K. C. Kim for his valuable assis- tance during this project. His enthusiasm and drive served as a constant inspiration. Persons and institutions loaning type material, for which I am grateful, are: Dr. R. W. Crosskey (BMNH), who was extremely generous in loan- ing 56 type-specimens for this project; Dr. R. Litchtenberg (VMNH); Dr. V. Richter (LMNH); Dr. L. L. Pechuman (CUIC) who kindly loaned me the types of 14 species for an extended pe- riod of time; also J. C. Scott (MCZC), Dr. P. Wygodzinsky (AMNH), and Dr. F. C. Thomp- son (USNM) who helped in uncovering some important syntypes. I am indebted to Dr. P. H. Arnaud, Jr., and the Department of Entomology, California Academy of Sciences, for providing facilities and valuable assistance during my time on the West Coast. Thanks are also due to Dr. D. C. Rentz and D. H. Kavanaugh for their frequent assistance with many problems. I finally thank my husband, George Zelznak, for his unending encouragement and optimism throughout this study. BIOLOGY The larvae of Ptilodexia flies parasitize and kill their scarabaeid larva hosts. The adult flies, however, feed on nectar and they spend consid- erable time probing at flowers, particularly com- posites. While feeding they become covered with pollen and probably act as pollinators. Ptilodexia conjuncta and P. agilis adults have been observed pollinating the flowers of dwarf mistletoe (Arceuthobium cyanocarpum). Adults of Ptilodexia have been collected with pollinia attached to the tarsi. Adults are collected in many environments. Members of some species are collected at the seashore, while others have been taken at alti- tudes as high as 3,000 m in the Sierra Nevada and the Rocky Mountains. These flies have been taken at UV light, Malaise, and other flight traps. But the most productive method of col- lecting seems to be sweeping flowers, especially composites such as Baccharis and Solidago. The occurrence of specimens of Ptilodexia is, as with most parasites, seasonal. At times, hundreds of individuals of one, two, or even three species may be collected simultaneously at one kind of flower. In other years the flies will be rare — perhaps representing the normal build- up and decline of a parasitic population. In some areas, such as Long Island, New York, and Riv- erside, California, specimens of certain species have been collected on the same dates every year for ten or more years. However, both these areas have been extensively surveyed regularly by specialists and may represent the actual sit- uation, which is not seen in other areas simply because of poor sampling. There is no information on the mating habits of Ptilodexia spp. Despite the numerous speci- mens collected, few have been pinned in copula. Males usually emerge before females, but the place and time of mating is unknown. The female carries hundreds of tiny larvae in her abdomen during larviposition. It is not known if she simply broadcasts them or if she places them directly on the soil. Neither is it known if the female is able to locate areas of host density, nor if the larva has the sensory capacity to find a host. If the larvae are depos- ited in ajar, they will wander along the sides for two to three days before dying. First-instar lar- vae of Prosena siberita survive a week or more in the soil (Clausen 1927). The larvae are presumably quite easy to rear if the host larvae can be kept alive under labo- ratory conditions. During the development of the parasite larva, a defensive response of the host causes a respiratory funnel to appear at the point of attachment to the host. This is a scler- otized funnel-shaped structure which encloses the caudal end of the larva and can be seen through the integument of the host. The larva feeds on the internal fluids and fat body of the host and finally leaves the host to pupate in the soil. By this time, the larva has ingested nearly all the contents of the host. Based on Davis's (1919) data, Ptilodexia lar- vae overwinter within the host. The pupal stage is quite short, cold temperatures are not re- quired to complete development. The host scarabaeid larvae, or white grubs, belong to the subfamily Melolonthinae, which PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 42, No. 1 includes such common and economically impor- tant genera as Phyllophaga, Popillia, and Mac- rodactylus. The dynastinine scarabaeids of the genus Aphonus are also parasitized by these flies. Champlain and Knull (1944), and Peterson (1948) implicate Ptilodexia canescens as a par- asite of the cerambycids Saperda calcarata and Rhagium lineatum. Most species of Phyllophaga have a two- or three-year life cycle. Popillia japonica and some of the other hosts have one-year life cycles. The life cycle of Ptilodexia probably does not exceed one year. Early-instar white grubs are attacked in the fall and fed upon until the following spring or summer when the parasite pupates. Adult flies emerge shortly thereafter. According to Davis (1919), when infested host larvae are brought indoors in the fall, the Ptilo- dexia larvae will continue their development within the host, pupate, and emerge as adults in the winter, without interruption. It is thus pos- sible that in areas with long seasons, two or more broods could develop. My samples from Texas, southern California, and some south- eastern states appear to support this. Two broods per year are possible only if there is an ample supply of grubs of the proper stage feed- ing in the soil. In these areas of bivoltinism, vari- ant populations in the species concerned are ap- parently more common than in areas where only one brood is possible. One can extrapolate to the tropical regions where even more genera- tions per year are possible, and the number of species and the variation among species is phe- nomenal. It is not known if any species of Ptilodexia is host specific, but some species are known to have more than one host. For example, P. car- olinensis can complete its life cycle either in Phyllophaga rugosa (fide Davis 1919, as Ptilo- dexia abdominalis) or in Popillia japonica. On the other hand, Ptilodexia maculata and P. prexaspes have restricted ranges and show little intraspecific variation — perhaps indicative of host specificity. Ptilodexia harpasa and one of its reported hosts, Macrodactylus subspinosus, are sympatric. Many questions remain unanswered regarding the relationship of host preference and specia- tion in Ptilodexia. In some species, local aber- rant populations are found that only vaguely re- semble the typical population. The possibility exists that these aberrant populations have shift- ed to a host significantly different from the com- mon one. It is conceivable that such a shift could be an early step in speciation. Many factors involving the relationship of host and parasite presumably influence the ap- pearance of the adult fly. These factors include number of fly larvae per host, instar of parasi- tized host, rate of host development, and rate of parasitoid development. Specimens of P. caro- linensis developed in Phyllophaga grubs differ greatly from those developed in larvae of Popil- lia. These factors contribute to the extreme intra- specific variability in Ptilodexia, perhaps ulti- mately leading to speciation in the group. Care- fully controlled breeding experiments are needed to help understand the effects on the parasites of the host and host environment. TAXONOMIC CHARACTERS Most of the specific characters previously used by Ptilodexia taxonomists are subtle, dif- ficult to see, and unstable, sometimes differing not only between individuals but also on each side of the same specimen. Of the characters traditionally used to separate species, many have been either stable within the genus (e.g., "arista plumose") or different within a species (e.g., color, wing venation). Most earlier work- ers lacked sufficient study material to recognize normal intraspecific variation. Certain charac- ters these early workers used were good, but they are more useful used in combination with certain other characters. Sexual dimorphism has caused problems in the taxonomy of Ptilodexia. Abdominal color and color pattern, and leg color frequently differ between sexes; hence the two sexes of some species have been described under different names. I have freely used raw measurements in my descriptions, recognizing nonetheless that they are of limited use in this group. Proportional measurements are generally more useful, and I include them in my descriptions and diagnoses. The standards for these proportional measure- ments are head height and length of the first an- tennal segment — both measurements which are proportional to general size. This study is limited to dried adult flies, so structural characters are the only ones which form the basis of my classification. What follows WILDER: NEARCTIC PTILODEXIA is a brief discussion of the taxonomic impor- tance of the various physical characters and how their states are determined. The headings and general organization are similar to those used by Crosskey (1973a). It is hoped that other workers in the Tachinidae will adopt the same format, eventually bringing some order to the study of variation in the family. Body Color and Vestiture The general integumental color of Ptilodexia adults is a dull brown, although adults of certain more-advanced species may be black or testa- ceous, and teneral specimens are generally pal- er. The color of the scutellum compared to that of the rest of the notum sometimes is specifically useful (e.g., P. planifrons-P. contristans); how- ever, in adults of some species (P. rufipennis), it also varies intraspecifically. The color of the abdomen varies from reddish or testaceous with a dark longitudinal stripe, to concolorous black to testaceous. Although abdominal color may be of occasional taxonomic value, it almost always varies between males and females of the same species. In the female it is frequently concolor- ous, with the longitudinal stripe indistinct or ab- sent. General body color sometimes varies cli- nally, and in some species, smaller, darker populations exist in the northern parts of the range. Vestiture characters can aid in distinguishing members of different species. These characters seem to vary independently of integumental col- or. Facial tomentum varies from extremely heavy — totally obscuring the underlying integ- umental color-to fine and sparse. Occasionally there is a pattern or spot of color in this vestiture which can be distinct for a species (e.g., P. con- tristans, P. canescens). Facial tomentum may be dull (P. westi) or strongly shining (P. incerta). The color of the facial tomentum varies intra- specifically. The tomentum on the pleuron is of little taxo- nomic value, and that on the notum is only slightly more useful. The notal tomentum on adults of some species (e.g., P. westi, P. arida) is heavy, abundant, and almost flocculent, near- ly obscuring the integumental color, whereas on those of others (P. mathesoni) it is so fine that the notum appears polished. In members of oth- er species (P. conjunct a), the tomentum is flat- tened and shiny, giving the notum, or parts of it, a coppery hue. Usually notal tomentum is arranged in longitudinal stripes, but this striping varies among individuals and is of little diagnos- tic value. Notal tomentum also varies between the sexes, usually being heavier in the female. Tomentum on the mediotergite can be fine or heavy; in specimens of P. agilis and P. mathe- soni, however, it is absent. Abdominal tomentum varies more between species than between sexes. It is generally gray- ish and tessellate, although in some individuals it may be gold or brownish; and it may be shin- ing or dull. Only in adults of one species (P. mathesoni) has the tessellate pattern been re- placed by a more uniform distribution of tomen- tum, and even then, only in the males. In mem- bers of a few species (e.g., P. pacifica, P. ponderosa), the grayish tomentum is the only vestiture on the abdomen, but in most there are patterns of brown or gold tomentum which can aid in distinguishing species. In P . rufipennis adults the bases of the median marginal setae on the third and fourth abdominal tergites are sur- rounded with gold tomentum; adults of P. agilis have a marginal band of it on the third and fourth tergites; those of P. arida have brown tomentum overlying the longitudinal stripe. These characters of general color and vesti- ture show a greater degree of variation within Ptilodexia than within all other Nearctic Pro- senini, with the possible exception of the most closely related genus, Mochlosoma. When examining specimens for colors and patterns of tomentum, it is imperative that they be viewed from several different angles. Often a pattern can be distinguished only if it is seen obliquely. Chaetotaxy and Hairing Chaetotaxy is of minor taxonomic use in Ptilodexia, as it varies widely within species but little between them. Frequently, numbers and even the presence of setae vary from one side of the specimen to the other. This phenomenon is represented in the descriptions by separating the two states with a slash (1/0). Hairs differ from setae in being much smaller and finer. The taxonomically useful setae on the head are the oral vibrissae (Fig. 1). There is usually one pair (two in specimens of P. contristans). The size and spatial relationships between the vibrissae and the peristomal setae can aid in identifying adults in some species. In P. rufi- pennis adults the peristomals immediately below PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 42, No. 1 frontal villa frontal seta arista vlbrlssa eplstome perlstomal seta haustellum ninth terglte surstylus poslorbltal seta adeagus FIGURES 1-3. Structure of generalized Ptilodexia. Fig. 1. Head, anterolateral view. Fig. 2. Head, lateral view. Fig. 3. Genitalia, lateral view. the vibrissae are short, becoming longer with distance from the vibrissae; in P. conjuncta adults, on the other hand, the peristomals are subequal and nearly as long as the vibrissae. Numbers of peristomal setae vary between and within species, but considerable overlap be- tween species is common. The number and size of frontal setae vary, but these are even less reliable characters than are the number and size of the peristomal setae. The ocellar, postocellar, internal vertical, and external vertical setae show some intraspecific differences in size and number. I have described these differences, but they are too variable to be used diagnostically. The postorbital setae vary in length between species, but not as much as between the sexes. The hairs which are some- times inserted between them have minor signif- icance. In adults of some species the postorbit- als are long and closely spaced, while in those of others they are interspersed with fine setae half the length of the postorbitals, and in still others they are interspersed with tiny hairs. Another group of taxonomically useful hairs on the head are those immediately ventral to the postorbitals. Members of species such as P. planifrons and P. mathesoni are characterized as having two to four irregular rows of dark hairs between the postorbital setae and the yellow or white occipital hair. Members of other species have only one row of these dark hairs (P. cali- fornica), and those of others have none or just a few scattered hairs (P. maculata). This char- acter also shows much intraspecific variation and must be used cautiously. Perhaps the best diagnostic character in Ptilo- dexia is the hairing of the parafacials (herein de- fined as the sides of the head bounded by the apex of the second antennal segment, the oral vibrissae, the frontal suture, and the anterior eye margin). The presence, size, distribution, and color of these hairs are extremely variable but species specific. There is slight variation in the characteristics of these hairs between males and females belonging to the same species; that is, the parafacial hairs of the female are slightly sparser, finer, and are not inserted as far ven- trally on the parafacial as they are on the male. These hairs may be absent (P. rufipennis)( long, dark, and abundant (P. planifrons); sparse and pale (P. halone)', minute and occurring only on the upper anterior parafacial (P. incerta}; strong and concentrated at the lower edge of the eye (P. canescens)\ strongly inclined anteriad (P. contristans)', or inclined ventrad (P. harpasa). Many other combinations exist. It appears that this character can be of diagnostic value even in the Tropics, where there are many undescribed species. I have illustrated the character state for parafacial hairs in every Nearctic species herein described or diagnosed. The nature of the para- facial hairs is also of importance in Mochloso- ma, where they are always present, but not in the other Nearctic Prosenini. Parafacial-hair characteristics appear fre- quently in the key to species. When the hairs are small and pale, specimens must be examined carefully from several angles; often it is the bas- es of the hairs rather than the hairs themselves which are visible. The parafrontal hairs are of much less taxo- nomic value. They are usually present, dark, and are either sparse or abundant. WILDER: NEARCTIC PTILODEXIA Thoracic chaetotaxy is of little diagnostic im- portance in Ptilodexia. The numbers of such se- tae as sternopleurals, notopleurals, and posta- lars are generally constant within the genus. Others such as posthumerals, presuturals, ac- rostichals, dorsocentrals, and scutellars vary somewhat between species, but they also show considerable intraspecific variation. The number of humeral setae and discal scutellars are more constant, but must be used in combination with other characters to aid identification. The length and density of hairs covering the Ptilodexia thorax vary between species. These, however, are difficult characters to divide into easily defined states. Propleural hairs do not oc- cur in Ptilodexia adults, but are present in mem- bers of several closely related genera. They have diagnostic value at the generic level. Another group of hairs on the thoraces of these flies is the infrasquamal setulae, small hairs inserted beneath the point of attachment of the squamae or calypters. In adults of some related genera in the Prosenini, these are always absent. In some Ptilodexia adults their absence may be a reliable specific character state (e.g., P. canescens, P. maculata}, but in others, their absence carries less taxonomic importance. Six- ty percent of the specimens of P. rufipennis ex- amined had infrasquamal setulae, but they were present in only twenty percent of P. incerta specimens. This character is of equal value in both sexes. When using this character, one must realize that the 'absence' of infrasquamal setulae indicates absence on both sides of the body. Hairs and setae on the legs have little diag- nostic value; often the setal length reflects total body size more than any specific difference. The exception to this is the length of the antero- and posteroventral setae on the posterior leg of the male, which show species-level variation. These are difficult to measure, however, and have not been used in this revision. Other setae on the femora show some taxonomic potential, espe- cially the presence or absence of anterior setae on the posterior femora. Numbers of abdominal setae vary intraspecif- ically, but they usually vary around a certain number which can be defined for some species. The presence or absence of median marginal se- tae on the first syntergite can be a useful char- acter. The number of median discal and median mar- ginal setae on the third and fourth tergites is useful in separating members of some closely related species (e.g., P. californica and P. pa- cifica), while in others it shows considerable in- traspecific variation. The presence or absence of lateral discal setae on these tergites will sep- arate members of distantly related species. Length and density characters of abdominal hairs have about the same taxonomic value as those characters in thoracic hair. That is, they differ and seem to be constant among members of a species, but are difficult to separate into character states. Although hairing on the genitalia varies only slightly between species, the presence of strong setae on the ninth tergite (epandrium) is an ex- cellent diagnostic character in adults of P. con- tristans and P. westi. Head Head characters in Ptilodexia are of more use taxonomically than characters of any other part of the fly. Included are those of chaetotaxy, which have been discussed in the previous sec- tion. Drawings of the head, with the terms used in this paper, are presented in Figures 1-2. The head, in members of this genus, is wide and boxlike. The parafacials and genae are wide and covered with fine, dull-lustered tomentum. The genae are usually reddish, contrasting with the whitish parafacials and genal dilations. The velvety-appearing frontal vitta extends from the vertex to the frontal suture. The third antennal segment is rarely longer than twice the length of the second and bears an arista covered with long fine hairs. Between the antennae is a raised ridge or carina which does not protrude beyond the antennae. The epistome may or may not pro- trude. Mouthparts are similar to those of other calyptrate flies with the mentum length from 0.3 to 0.8 times the head height. The width of the parafacial is of considerable diagnostic use in Ptilodexia. Ptilodexia rufipen- nis adults have narrow parafacials, while in P. conjunct a adults they are quite wide. This char- acter is easier to evaluate in males than in fe- males, and it varies more among Ptilodexia adults than among Mochlosoma adults. The shape of the frontal vitta can be of use in this genus. Below the ocellar triangle and be- tween the eyes, the frontal vitta is usually oblit- erated (Fig. 4b), the parafrontals becoming con- PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 42, No. 1 FIGURES 4-6. Comparison of specimens of Ptilodexia. Fig. 4. Comparison of adults of two generalized species of Ptilodexia showing variation in width of epistome, position of oral vibrissae, and width of frontal vitta; A. head, anterior view, with frontal vitta not obliterated and distance of oral vibrissae from oral margin greater than distance between oral vibrissae; B. head, anterior view, with frontal vitta obliterated and distance of oral vibrissae from oral margin less than dis- tance between oral vibrissae. Fig. 5. Comparison of antennae of adults of two generalized species of Ptilodexia; A. antenna, showing length of plumosity on arista shorter than length of second antennal segment; B. antenna showing length of plu- mosity on arista longer than length of second antennal seg- ment. Fig. 6. Ptilodexia ponderosa (Curran), holotype, head of female, lateral view. tiguous. In members of some species (e.g., P. canescens, P. halone), however, the parafron- tals do not touch, and the frontal vitta is contin- uous from the antennal base to the ocellar tri- angle (Fig. 4a). This character varies to a similar degree in specimens of Mochlosoma. The size and shape of the carina are useful at the supraspecific level. Among Ptilodexia species, the carina is sometimes distinctly shaped (e.g., P. planifrons). Carina characters vary more among Mochlosoma than Ptilodexia adults, in general being wider and better devel- oped in individuals of the former genus. In mem- bers of several closely related genera such as Diner a and Hesperodinera, the carina is strong- ly developed, protruding from between the an- tennae, appearing almost bulbous, and visible from the lateral aspect. In Rhamphinina adults it is short, narrow, and strongly keeled. The ca- rina shows no sexual dimorphism in size and shape, and except in a few cases it is species specific. The oral vibrissae and adjoining areas hold characters of taxonomic value in Ptilodexia. These are the width of the depression between the bases of the oral vibrissae and the distance of the vibrissae from the epistome (sclerotized oral margin). Members of some species of Ptilo- dexia have this area wide and short; an example is P. rufipennis, in adults of which the vibrissae are far apart and close to the oral margin (Fig. 4b). The opposite state is shown in P. prexaspes adults, which are characterized by vibrissae that are far from the oral margin, but not far apart (Fig. 4a). In Rhamphinina specimens, the area is four times as high as wide. This character is diagnostic in a few species, but in others it varies intraspecifically. It shows no sexual dimor- phism. In evaluating this character, physical measurement is necessary; estimate is inade- quate. The invisible line connecting the vibrissae should pass through the center of their bases, and the sclerotized margin of the epistome should serve as the ventral boundary. Width is measured only at the vibrissae; height is mea- sured mesially. The epistome generally protrudes in Ptilo- dexia and Mochlosoma specimens, and al- though both genera show variation, in Ptilodexia adults it is sometimes species specific. In adults of species like P. prexaspes, P. canescens, and P. halone, the oral margin projects slightly if at all. As a result, the lower anterior portion of the head is vertical in profile (Fig. 71), and in some cases the anterior margin of the head protrudes further anteriad at the antennae than at the vi- brissae. Ptilodexia conjunct a adults show the opposite state, the epistome projecting strongly, as does the lower anterior portion of the face (Fig. 26). In other prosenines the character shows less intraspecific variation than it does among species of Ptilodexia. The length of the haustellum is an extremely valuable taxonomic character in Ptilodexia (and Mochlosoma}, and it is also the chief difference distinguishing Ptilodexia from Mochlosoma specimens. In individuals of the former genus, the length of the haustellum varies from 0.3 to WILDER: NEARCTIC PTILODEXIA 0.9 times the head height, and the shape is broad and linear or slightly tapered; it is rigid in all individuals. Mochlosoma specimens have the haustellum much longer than the head height, and narrow and flexible. In other Prosenini this character serves to separate genera. In Ptilo- dexia I have compared the length of the haus- tellum with the head height and used the result- ing ratio as a diagnostic character which varies consistently between species, little within species, and not at all between the sexes. When using this character in the key, actual measure- ments must be made; estimating the ratio is dif- ficult because a slender haustellum appears lon- ger than a broad one of the same length. There is intraspecific variation in the length of the haustellum among members of a few species. Among P. rufipennis (as well as P. arida, P. carolinensis, and P. pacified) specimens, the haustellum length varies locally. The length can be short in members of one population and no- ticeably longer in those of another. The char- acter is still useful, though, since the variation remains within easily expressed values. The length and shape of the palpi vary slightly between members of different species of Ptilo- dexia. The length is expressed, in this paper, as a fraction of the haustellum length. Some Pro- senini, such as Prosena and Senostoma speci- mens, have short stubby palpi; and in Atelog- lossa adults they are completely absent. This character varies among Mochlosoma specimens much as it does among those of Ptilodexla. In members of P. arida and P. prexaspes, the length of the palpi may nearly equal the length of the haustellum, while in those of P. obscura, it is rarely more than 0.3 times the haustellum length. The antennae possess some useful taxonomic characters: length, shape, and arista plumosity. The length of the third segment is herein ex- pressed in terms of its relationship to the rela- tively constant second segment. Measurement of the second segment is taken from a slightly anterodorsal aspect and is the longest dorsoven- tral length of the segment. Among Prosena and Senostoma adults, length of the third segment is approximately twice the length of the second; in those of most other Prosenini, it is considerably shorter. Among Mochlosoma species the length varies, the most usual state being the third segment equal to 1.4 to 1.5 the second. The same is true in Ptilodexia species, where this character can be used to separate adults of some species. Ptilodexia sabroskyi adults have a short third segment, subequal to or shorter than the second, whereas those of P. rufipennis have the third segment up to twice the length of the second. Specimens of P. obscura sometimes have the third antennal article broadened apically instead of slightly pointed as it is in members of most species. The length of the plumosity on the arista is an excellent diagnostic character in Ptilodexia. I have expressed it in relation to the length of the second antennal segment. The arista, including the plumosity, is measured at its greatest width (Fig. 5). In specimens of P. rufipennis and P. harpasa, two species with long third antennal segments, the length of the plumosity is greater than twice the length of the second antennal seg- ment, while in those of P. planifrons and P. prexaspes, the length of the plumosity is less than or equal to the length of the second seg- ment. This character is especially useful in sep- arating adults of the closely related P. califor- nica and P. sabroskyi. Thorax Most of the thoracic characters used in this revision have been discussed in the sections on vestiture and chaetotaxy, the remaining ones are those of the mediotergite, legs, and wings. Adults of Ptilodexia have a typical calyptrate thorax with the mesothorax highly developed, and the prothorax and metathorax reduced. The scutellum is small; ventral to it is the bulging postscutellum, which distinguishes members of the family Tachinidae. The pleuron is typical of other calyptrate flies. The propleuron is bare, although the rest of the pleuron is beset with fine hairs and numerous groups of setae. The legs are long with extremely long tarsi. The wings are also long, the venation typical of ca- lyptrate flies. The mediotergite is the oval arched area ven- tral to the postscutellum. In members of some species the mediotergite is dorsally polished, although other parts of it may be tomentose. Care must be taken when observing this char- acter, since on adults of some species the me- diotergite has a thin layer of tomentum and still appears shiny. Leg color is a taxonomically useful character. In species where the color is similar in members 10 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 42, No. 1 of both sexes, pale-colored legs are diagnostic. The color of the tarsi can also separate members of different species (e.g., P. halone and P. prex- aspes}. Ptilodexia maculata specimens are dis- tinguished by distinct femoral patches which, although present on members of other species, are strikingly evident on those of P. maculata. Other species exhibit dimorphism in leg color, the males with dark legs, the females with pale legs (e.g., P. agilis, P. arida). Two other species, P. rufipennis and P. pacifica, have fe- males with pale legs and males with legs of var- ied color. Wing venation is useful in distinguishing members of some genera in this tribe (Nimio- glossa). Within Ptilodexia (and Mochlosoma), however, it is of dubious value. Wing color is constant within species and can be used diag- nostically. Adults of P. contristans have the wings distinctly darkened basally, while those of P. mathesoni have the entire wing darkened. The colors of the squamae, epaulet, and basi- costa also show slight differences between mem- bers of certain species, but they can be varied among those of others. Abdomen and Genitalia In Ptilodexia adults as in most Nearctic Pro- senini, the abdominal tergites meet ventrally, entirely obscuring the sternites. The first tergite is actually composed of two fused segments; the next three tergites — third, fourth, and fifth — are conspicuous. The sixth tergite is fairly broad, the edges not meeting ventrally (but embracing the fifth sternite); it and those remaining are withdrawn into the fifth tergite. The next two tergites are fused and become the seventh syn- tergite, which is fairly narrow in Ptilodexia members (not much wider than the epandrium), with its surface oriented posterodorsally, as is the epandrium (the ninth tergite). This pattern is similar in Mochlosoma members, but in other Prosenini it is different. In Prosena adults, for example, the ninth tergite appears to be fused with the seventh and eighth, and in members of Hesperodinera, the fused seventh and eighth tergites are exposed and greatly enlarged, the surface facing posterad, the epandrium forced beneath the abdomen. Abdominal color is varied intraspecifically but is still useful as a key characteristic distinguish- ing members of some species. In the key pre- sented herein, when the abdomen is described as reddish laterally, at least the second and third tergites (of the male — the character is not as consistently applicable to the female) have the integument reddish or rufotestaceous laterally. On specimens in which the abdomen is concol- orous dark brown or gray, there may be a slight rufescent cast along the margins of the tergites. This state should not be confused with the pre- vious one, in which the reddish color extends from the anterior to posterior margins of the ter- gites. In specimens of Ptilodexia, the external male genitalia (Fig. 3) have taxonomic value. The characters which vary slightly between mem- bers of different species are the shape of the ninth tergite, the shape of the surstyli, and the shape of the cerci. These characters are useful in distinguishing members of a few species, but sometimes vary more intra- than interspecifical- ly. Only in species with extraordinarily modified members (e.g., P. westi, P. rufipennis) can the external genitalia be called diagnostic, and even then, they must be dissected for characters to be examined properly. Often the genitalia of adults of Mochlosoma and Ptilodexia are iden- tical. The above-mentioned characters vary greatly between specimens belonging to differ- ent genera and are of excellent supraspecific group characters. Internal genitalia are generally not useful in separating members of species of Ptilodexia. The aedeagus is nearly identical in members of this genus and those of Mochlosoma . The ejac- ulatory apodeme, however, is useful in distin- guishing specimens of some species or species groups. Its shape can be distinct, as in P. con- tristans, P. planifrons, and P. rufipennis mem- bers; between many of the species, though, it does not vary. Female genitalia show no striking diagnostic differences, with the notable excep- tion of the surface sculpturing of the spermathe- cae which, with high-magnification studies, may reveal specific differences. The reproductive systems of both sexes of Ptilodexia have been described by Townsend (1938). The larvae of Ptilodexia have never been de- scribed, even though there is a figure of a mature larva and the puparium in Davis (1919). The na- ture of the cephalopharyngeal skeletons of first instar larvae (from the abdomen of gravid fe- males) has been used to separate species in some genera of tachinids (Archytas), but the character is of no use in Ptilodexia. Greene (1922) de- WILDER: NEARCTIC PTILODEXIA 11 scribed the puparium of an unknown species of Ptilodexia (erroneously determined as P. tibi- alis). Of the useful diagnostic characters, none works to separate members of all species from those of all others. Most of these characters are of high value in distinguishing members of the derived species, but when members of certain primitive species (P. carolinensis, P. major) are examined, they lose much of their value and more characters must be considered in making identifications. PHYLOGENY Present attempts to reconstruct the phylogeny of a genus or tribe in the Tachinidae are based on incomplete data and should be considered extremely tentative at best. Characters used at generic and tribal levels are so unstable that con- vergence, loss, and acquisition occur repeated- ly. Most of the species, and probably many of the genera, are unknown or poorly defined on a worldwide basis. Host relationships are largely unknown. For the phylogeny of Ptilodexia, Neotropical species and representatives of closely related genera were carefully examined to infer apo- morphic and plesiomorphic states. Character matrices were then constructed and phylogenet- ic trees inferred. This method works well when trying to construct probable relationships in higher categories, but for relationships among species it is not adequate. This is becuase the characters distinguishing species are generally more unstable than those distinguishing families or tribes. Many specific characters can be lost or regained easily. Relationships within the Prosenini can be in- ferred only after examining members of the tribe on a world basis. I have not had the opportunity to do this. I have seen a few representatives of the North American genera, none of the exclu- sively Neotropical genera, one of an Australian genus, and one of Prosena, a worldwide genus. Most of the species in these genera can not yet be identified with existing keys. The characters used to infer relationships be- tween genera are facial carina, space between vibrissae, propleural hairs, and haustellum length. The form of the facial carina provides a good generic character, much as it does in the tribe Rutiliini (Crosskey 1973a). In members of Prosenini, the carina was lost once, although slight expression is common in members of some species of Ptilodexia and Mochlosoma. The area between the vibrissae is another sta- ble generic character. In Ptilodexia and Moch- losoma adults, and in those of some of the other genera, this area is depressed and may be flat or slightly concave, while in members of the more primitive genera, it is slightly to strongly con- vex. In the primitive genera, the oral vibrissae are situated at or slightly above the oral margin, while in Ptilodexia, Mochlosoma, and Rham- phinina adults, they are inserted distinctly dor- sal to the epistome. In Arctophyto and Milada adults, their placement is intermediate between that in members of the primitive and the derived genera. Propleural hairs are lacking in members of Sentstoma and Prosena, two of the more prim- itive genera. They have been lost in Hespero- dinera adults and are never seen in those of the advanced genera. The haustellum is longer than the head height in specimens of Prosena, Prosenoides, Nimio- glossa, and Mochlosoma, but is much shorter in those of the other genera. Also, there is great variation in haustellum length among Ptilodexia species (from 0.3 to 0.9 times head height). The position of Ptilodexia within the Prosen- ini is advanced. Assuming an Oriental center of origin for the group, there is a wealth of forms (members of which have a broad, blunt, facial carina) in the Oriental and Palearctic regions. Some of these, such as Prosena and Dinera, are widespread. It was probably a form similar to Dinera, Milada, and Arctophyto which, along with its hosts, crossed the Bering Land Bridge during favorable conditions in the late Tertiary. Subsequent radiation before and during the Pleistocene must have been great, for Ptilodexia is the most diverse genus of prosenines in the western hemisphere. Dinera and Arctophyto-Milada remain on both sides of the Pacific with relatively few species. Other small genera which proably orig- inated from this complex are Ateloglossa, Dol- ichocodia, Prosenoides, and Hesperodinera, members of which have retained the inflated ca- rina, and Myoceropsis, Rhamphinina, and Nim- ioglossa, whose members have lost it. Ptilodexia and Mochlosoma, both large gen- era, probably had similar origins, Mochlosoma from a more restricted ancestral line. Although 12 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 42, No. 1 the strong carina has been lost, it is expressed to a moderate degree in members of some species in both genera. It is my opinion that the only character which separates Ptilodexia from Mochlosoma adults, the haustellum shape and length, is a phylogenetically sound one, that is, its origin in Mochlosoma is monophyletic. Al- though radiation of both genera has been great, divergence between Mochlosoma and Ptilo- dexia members is only slight. Ateloglossa and Hesperodinera had their origins from a Dinera-like ancestor, members of the former having lost the palpi and the latter the propleural hairs, but both having retained the squarish head and inflated facial carina. Pro- senoides adults, on the other hand, bear a closer resemblance to Prosena specimens, and the two may be closely related. Both Nimioglossa and Rhamphinina had their origins early in the Ptilo- dexia-Mochlosoma line, their members having diverged from the ancestral forms in having the area between the oral vibrissae depressed and the vibrissae placed considerably above the oral margin, as do representatives of Ptilodexia and Mochlosoma. Within Ptilodexia, characters indicating rela- tionship are difficult to determine. The ancestral and derived states of a few of these characters have been inferred. Small size and dark color seem to be primitive states within Ptilodexia, while the derived states are large size and pale color. The primitive state of the length of the third antennal segment, the length of the plumosity on the arista, and length of the haustellum is an intermediate one, with the derived states being short and long. The presence of both parafacial hairs and infrasqua- mal setulae appear to be primitive within this genus. Primitively the oral margin is narrow and projecting, while the more advanced forms show it to be wide and not projecting. The primitive, typical shape of the ejaculatory apodeme can be seen in most species (Fig. 44). Members of some of the advanced species have this structure mod- ified in various ways (Fig. 29, 35), although those of others do not. The genus Ptilodexia in North America is comprised of six loosely knit groups. The first, and probably most primitive of these, is the agilis group, which consists of P. agilis, P. ob- scura, and P. mathesoni. The next group is the carolinensis group, with member species P. carolinensis, P. halone, P. prexaspes, and P. canescens. The harpasa group is composed of P. rufipennis, P. arida, P. harpasa, and P. pon- der osa. In the major group are P. major, P. incerta, P. maculata, and P. flavotessellata. The conjuncta group contains P. conjuncta, P. planifrons, P. contristans, and P. westi; and the californica group, P. californica, P. sabroskyi, and P. pacifica. The character states which segregate mem- bers of these groups are vague and difficult to define, but since the groups appear to have both a zoogeographical and morphological basis, they will be discussed. Their relationships to each other are somewhat less clear. Members of the agilis group are small, dark flies with little red color on the abdomen and a short haustellum. The legs of the females of P. agilis and P. mathesoni are pale, while those of P. obscura are dark. The mediotergite is pol- ished or shiny in members of all three species. This is probably the most primitive group of Ptilodexia; P. agilis members perhaps being similar to those of the prototype of the genus. Ptilodexia agilis is a widespread western form, occurring into central Texas; P. mathesoni, closely related, is a northern form found in New York, Michigan, and eastern Canada. Ptilodexia obscura has a distribution from the Appalachi- ans to the Rocky Mountains and is nearly com- plimentary to that of P. agilis. Species included in the carolinensis group have members with a nonprojecting epistome and a nearly vertical anterior head profile. All adults have relatively short plumosity on the arista and the abdomen reddish laterally. Two of these species, P. halone and P. prexaspes, have limited east coast distributions. Ptilodexia ca- nescens occurs across the northern United States and Canada, while P. carolinensis is widely distributed from the east coast to the Rocky Mountains. Ptilodexia carolinensis is probably the oldest of the four, P. canescens, P. halone, and P. prexaspes being derived from it during the Pleistocene. The harpasa group is the most ambiguous of all, containing most of the Antillean and many Mexican species. Members of these species all have long antennae and long plumosity on the arista; some lack parafacial hairs. Ptilodexia ru- fipennis occurs from the east coast to the Rocky Mountains and across Canada; P. harposa is more restricted, P. arida is restricted to the Southwest and Mexico, while P. ponder osa is WILDER: NEARCTIC PTILODEXIA 13 CALIFORNICA CONJUNCTA AGILIS OBSCURA CAROLINENSIS MAJOR HARPASA FIGURE 7. Inferred phylogeny of Ptilodexia species groups. probably a West Indian species, with one record from Florida. Assuming that P. harpasa is the closest to the ancestor of the group, P. rufipen- nis became the most widespread and P. arida and P. ponderosa radiated in the southern lati- tudes. Ptilodexia major, the most primitive member of the next group shows slight similarities to specimens of P. harpasa. It ranges widely throughout the Midwest, Southwest, and Mex- ico. Ptilodexia incerta has an eastern distribu- tion almost exactly complementary to that of P. major, while P. maculata and P.flavotessellata are restricted in the Southwest and Midwest. These species all have members with short, pale, parafacial hairs. The next group, conjuncta, is probably de- rived directly from the ancestral agilis group and consists of only western species. Ptilodexia conjuncta, its most primitive member, ranges throughout the Rocky Mountains from Canada into Mexico and west to California. Ptilodexia planifrons and P. contristans extend from the southwestern United States into Mexico; and P. westi, a close relative of P. contristans, is re- stricted to the extreme southern Midwest and the Southwest. The last, or calif ornica group, is related to the conjuncta group and probably had a similar or- igin. Its three species are confined to California and the West Coast, and members of these taxa show similarities only to members of the con- juncta group. A graphic representation of the relationships of these species groups is given in Figure 7. Pos- sible events at the numbered branching points are as follows: 1. Major east-west split. Eastern group mem- bers with parafacial hairs extending low on the face; western group members with parafacial hairs high on the face. 2. Widening of face; lighter general color. Some females with yellow legs; lengthen- ing of plumosity on arista. Some advanced members radiating widely; radiation into Mexico and Antilles. 3. Lengthening of haustellum, increase in size, widening of face. Extensive radiation into Mexico. 4. Slight increase in size. Flattening of oral margin and anterior facial margin; short- ening of haustellum and plumosity on arista. 14 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 42, No. 1 5. Shortening of haustellum. Radiation and isolation in southwestern United States and Mexican Pleistocene refugia. 6. Pleistocene isolation in California. De- crease in abundance of parafacial hairs and length of haustellum. ZOOGEOGRAPHY Although the dispersal powers of Diptera are relatively great, the distribution patterns seen in Ptilodexia seem to be dependent upon those of their hosts, the Scarabaeidae. Distribution of some Scarabaeidae are well known, and their possible histories have been discussed in several papers (Howden 1963; 1966). All statements made in this section are ten- tative. The patterns discussed are those of species of Ptilodexia, but interpretations of those patterns are those which have been of- fered for some of the species of Scarabaeidae. No host specificity has been found, and it is only speculation that similar patterns of Ptilodexia and their scarabaeid hosts are due to similar his- tories. Howden (1966) stated that North American species of Phyllophoga show a decline in num- bers from Georgia to Canada and from Texas or Arizona to Nebraska. This is true of Ptilodexia. He also stated that if certain areas of Texas were included with Arizona and New Mexico, there would be little overlap between the eastern and western faunas (approximate dividing line, 100th meridian). This holds true for the most part in Ptilodexia. However, a number of eastern species occur all the way into British Columbia in the northern parts of their ranges. As with Phyllophaga, many of the southern Arizona rec- ords represent the northern limits of Mexican species. When plotting the centers of distribution of species of Ptilodexia, it was noted that six species groups could be defined geographically. These were the same six groups which had been structurally and zoogeographically defined above. Although these groups show that the phylogeny presented herein has a zoogeograph- ical basis, they are not the best groupings for discussing zoogeography. I have categorized the species of Ptilodexia into six zoogeographical groups, based on their complete distributions rather than centers of dis- tribution. The relationships of these categories may give insight into the historical zoogeogra- phy of the group. The first of these is an extreme northern pat- tern shown by P. canescens and P. mathesoni. The distribution is almost exclusively in areas which were previously glaciated. The range of P. mathesoni (Fig. 17) is restricted to Michigan, New York, and eastern Canada. Its ancestral and most closely related species, P. agilis, oc- cupies a large area from the Rocky Mountains west, extending eastward into Texas (Fig. 12). Ptilodexia canescens, from the Caroline nsis group, inhabits the northern United States and Canada from Newfoundland to British Colum- bia. In the East it extends southward only to the previous front of the Wisconsin glaciation, while in the West it extends southward into eastern Idaho, western Wyoming, and northern Utah (Fig. 67). The distribution of a species in pre- viously glaciated areas without representation south of the glacial front is fairly uncommon (Ross 1965). This deglaciated area may have of- fered considerable opportunity for expansion to certain Scarabaeidae and their Ptilodexia para- sitoids. The next group has a widespread distribution, throughout the eastern United States into the plains states and, in some cases, even further west. Of these species, only P. harpasa (Fig. 82) lacks representation in the lower Midwest. The other species, P. carolinensis (Fig. 62), P. incerta (Fig. 103), P. obscura (Fig. 22), and P. rufipennis (Fig. 88) occur widely throughout the Midwest and the East. The remaining eastern distribution pattern is that of P. halone and P. prexaspes. Both of these species belong to the carolinensis group; they are closely related and complementary in distribution. Ptilodexia prexaspes occurs in Florida and along the Atlantic coast to Virginia (Fig. 77), while P. halone is found in Mississip- pi, Tennessee, and along the coast from Mary- land to New York (Fig. 72). It is possible that these relatively uncommon species are host-spe- cific parasites of some of the large, flightless scarabs found in the Southeast and discussed by Howden (1963). Two species, P. agilis and P. conjuncta, have large western ranges, the former from British Columbia to Texas and west to the Pacific coast (Fig. 12), the latter from Mexico to British Co- WILDER: NEARCTIC PTILODEXIA 15 lumbia, west to the coast (excluding California), and eastward through Canada to Ontario (Fig. 27). Ptilodexia calif arnica (Fig. 47), P. pacific a (Fig. 52), and P. sabroskyi (Fig. 57) have ranges which are restricted to the west coast of the United States. It is possible that some unique local populations are parasitic on the large flight- less genera of scarabs (such as Pleacama) which survived in situ during the Pleistocene. The remaining distribution group is the most common in Ptilodexia, occurring in at least sev- en species. This is a southwestern distribution, with species which may have had Mexican re- fugia. Four of these are Mexican species whose ranges extend northward into the mountainous regions of Arizona and New Mexico, rarely into Utah and Idaho. These four are P. contristans (Fig. 37), P. planifrons (Fig. 32), P. maculata (Fig. 108), and P. arida (Fig. 93), all recently differentiated. It is possible that the ranges of many other Nearctic Mexican species also ex- tend into these areas, but specimens have not yet been taken by collectors. Two of this southwestern group, P. westi and P. flavotessellata, apparently do not range into Mexico. The former occurs broadly along the international boundary from central Arizona to eastern Texas and into Oklahoma and southern Kansas (Fig. 42), while P. flavotessellata occurs in northern New Mexico, Colorado, and Ne- braska (Fig. 1 13). The last species in the southwestern group is P. major. Its distribution is a combination of the ranges of the previous two groups, extending from Mexico (where it is widespread) into the mountains of Arizona, New Mexico, and Colo- rado and through Texas into the Plains in Ne- braska and Kansas (Fig. 98). Unlike those of Phyllophaga (Howden 1966), eastern species of Ptilodexia frequently occur from Georgia to southern Ontario; others range broadly across the northern part of the United States and Canada. I see this deviation as a re- sult of the vagility of these flies and the probable capability of developing in different hosts, fac- tors which may account for other deviations from typical scarabaeid distributions. Inferring the historical zoogeography of Ptilo- dexia is extremely speculative. The genus, as we know it, probably evolved on this continent, its ancestor reaching the area via the Bering Land Bridge during the Tertiary. By the onset of the Pleistocene, most of the species were probably already established. Pleistocene cli- matic fluctuations must have affected the distri- bution patterns we see in the genus today. The eastern species in our fauna may have occupied southeastern Pleistocene refugia, most of them expanding westward in the north after the ice sheets retreated. Ptilodexia agilis seems to have been much more widespread at one time, one of the species derived from it being found only in the Northeast. While P. agilis may have had a wide refugium, P. conjuncta and P. californica perhaps survived the Pleistocene in Mexico and California, respectively, separated by the extensive desert barriers of the time. The three species endemic to California were prob- ably separated from the other species at a rela- tively early time, closely resembling each other considerably more than any other species. Their refugia were in central and southern California, and subsequent recolonization proceeded no further north than the southern limits of the ice sheet. The southwestern groups could have sur- vived the glacial periods in situ or in Mexican refugia. It appears that P. major was once a widespread species, extending well into Mexico before the Pleistocene and giving rise to many species there. The Southwest, including Texas and Arizona, has the largest number of endemics. In the warmer parts of the country such as these, more generations per year are possible, and evolution can proceed at a faster pace than in the north. This ripay account, in part, for the large number of endemics; it also helps explain the numerous divergent populations seen in California and Texas as well as the tremendous diversity of the genus in Mexico. Genus PTILODEXIA Brauer and Bergenstamm Ptilodexia BRAUER AND BERGENSTAMM, 1889:119 (Type- species, Ptilodexia carolinensis Brauer and Bergenstamm, 1889, by original designation.) Myoceropsis TOWNSEND, 1915:23 (Type-species, Rhynchio- dexia flavotessellata Walton by original designation.) Rhamphinina, authors, not Bigot. Rhynchiodexia, authors, not Bigot. Rhynchodexia, emend. Wulp, 1891. The genus Ptilodexia Brauer and Bergen- stamm is confined to the New World. It is best represented in the Neotropical region, as is the entire tribe Prosenini. PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 42, No. 1 DIAGNOSIS. — Members of Ptilodexia can be distinguished from those of all closely related species of Nearctic Prosenini by the following combination of character states: propleuron bare; facial carina sometimes well developed (but never broad, blunt, and separating the an- tennae); haustellum shorter than the head height; apical cell open or closed at wing margin; and infrasquamal setulae usually present. DESCRIPTION. — Color black to reddish, usu- ally with thorax dark and abdomen pale with dark longitudinal stripe. Head with face broad in profile, anterior margin usually vertical; epi- stomal margin somewhat projecting; frontal vitta strongly narrowed between eyes, often obliter- ated; facial tomentum heavy to sparse, dull to shining, color varied, but usually grayish; para- facial hairs varied, absent or present; carina long, not much deeper than width of third an- tennal segment; postocular setae long; one or two pairs of oral vibrissae; epistome generally projecting to some degree; haustellum rigid, var- ied, from 0.3 to 0.8 times head height; palpi long. Antennae with length of third segment varied from one to two times length of second; arista with length of plumosity more than length of second antennal segment. Thorax with propleu- ron bare; mesonotum strongly or weakly tomen- tose, usually indistinctly striped; three or four pairs of presutural and postsutural acrostichals; a tuft of small hairs on postalar wall. Wing length 2.5 times width; apical cell open or closed at wing margin; infrasquamal setulae present or absent. Legs dark, tibiae lighter in most cases, posterior tarsi very long, 1 .5 times length of tib- ia; claws and pulvilli long. Abdomen broad, conical, tomentum usually in large irregular patches; numerous median discal and marginal setae on abdominal tergites; four abdominal ter- gites visible, lateral margins meeting ventrally. Ge nit alia slightly withdrawn, terminal, axis ver- tical; cerci and surstyli variously modified. Fe- male differs from male in following ways: frontal vitta wide with sides subparallel; eyes widely separated; profrons a little wider; frontal and peristomal setae not as abundant; vertex with few hairs or setae; postocular setae shorter and sparser; height of eyes distinctly less; frontal orbital setae present (Fig. 6). Thorax with fewer setae and hairs and more heavily tomentose; thorax and legs frequently lighter in color with fewer and shorter major setae and hairs; tarsal claws and pulvilli much shorter. Abdomen broader, shorter, and much more heavily to- mentose; usually with fewer median discals and often lacking other setae; integumental color uniform brown or gray in many females, even when it is marked in males of the same species. Brauer and Bergenstamm erected the genus Ptilodexia in 1889 for the North American species P. carolinensis. There has been much confusion since that time regarding the limits of the genus. This is because several characters normally constant within tachinid genera vary among Ptilodexia species. These characters in- clude the presence or absence of parafacial hairs and infrasquamal setulae. Prior to 1889, Macquart and Walker described species belonging to Ptilodexia in the genus Dexia Meigen, while Bigot (1885) created the genus Rhamphinina for those species he de- scribed. Bigot thought that Rhamphinina, a neo- tropical genus, and Rhynchiodexia, one of his Australian genera, could be distinguished from one another by the presence or absence of a facial carina. Wulp (1891) considered this to be an inconsistent character. Wulp (1891) emended the name to Rhyncho- dexia , which he used for what we now call Ptilo- dexia. He felt that Rhynchodexia and Rham- phinina were congeneric and mistakenly placed one Mexican species, contristans, in Hystri- chodexia. West (1924; 1925), a North American worker, thought that the species with hairy parafacials belonged to Ptilodexia while those with bare parafacials were Rhynchodexia. Austen (1907) shared this opinion. Later, West (1950) agreed with Curran (1934) that the two were probably one and Rhynchodexia was the proper name for the complex. Reinhard (1943) stated that the name Ptilodexia was available for American species. Examination of species of Rhynchodexia, now Senostoma (Crosskey 1973a), shows that this genus differs from Ptilodexia in having a pronounced facial carina as well as numerous other differences which will be discussed later. Rhamphinina dubia, the type-species of the ge- nus, is not a Ptilodexia. The name Estheria tibialis Robineau-Desvo- idy is frequently used for species of Ptilodexia (Townsend 1921; Aldrich 1905; Austen 1907). The type of this species is lost, so we cannot know if E. tibialis belongs to Ptilodexia. How- ever, since a characteristic of Estheria is the WILDER: NEARCTIC PTILODEXIA 17 presence in its members of a petiolate apical cell, and since this rarely occurs in Ptilodexia, I agree with previous workers who have chosen to reject the name E. tibialis. Key to the Nearctic Species of Ptilodexia la. Parafacial hairs present, although very small and pale in some individuals; legs of females varied in color 2 Ib. Parafacial hairs absent; legs of females pale in color 27 2a. Infrasquamal setulae present 5 2b. Infrasquamal setulae absent 3 3a. Parafacial hairs dark, long, present on most of parafacial (Fig. 66); length of plumosity on arista subequal to length of second antennal segment; face with traces or spots of brownish tomentum (northeastern U.S., trans-Canada, northern mountain states) canescens (Walker) 3b. Parafacial hairs pale and/or short, pres- ent only on upper anterior portion of parafacial; length of plumosity on arista at least 1.5 times length of second anten- nal segment (Fig. 5b); facial tomentum concolorous silvery gray or yellowish 4 4a. Femora of members of both sexes brown or black (eastern U.S. to about 100th meridian) incerta (West) (in part) 4b. Femora of members of both sexes or- ange with definite black or brown patches on flexor surfaces, coxae also with dark patches (Arizona and New Mexico) maculata n.sp. 5a(2a). Flies pale colored; thorax, abdomen, and femora pale brown to orange, or width of depression between oral vibris- sae less than distance between oral vi- brissae and oral margin (Fig. 4a) 6 5b. Flies dark; thorax, abdomen, or legs brown or darker in color; width of depression between oral vibrissae equal to or greater than distance between oral vibrissae and oral margin (Fig. 4b) .. 10 6a. Integument of tarsi pale, concolorous with legs ventrally halone (Walker) 6b. Integument of tarsi brown or black 7 7a. Width of the depression between oral vibrissae greater than distance from vi- brissae to oral margin 8 7b. Width of depression between oral vibris- sae less than or equal to distance from vibrissae to oral margin (Fig. 4a) prexaspes (Walker) 8a. Parafacial hairs minute, confined to up- per anterior parafacial; haustellum length 0.6 times head height (Fig. 6) (southern Florida) ponderosa (Curran) 8b. Parafacial hairs long, scattered on para- facial; haustellum length no more than 0.5 times head height (southwestern U.S. and Texas) 9 9a. Parafacial hairs dark, coarse and abun- dant (Fig. 61); femora or notum brown in many individuals; this color form un- common (Texas) ____ carolinensis Bauer and Bergenstamm (in part) 9b. Parafacial hairs pale or light brown, fine, sparse (Fig. 112); femora and notum pale orange-brown (southwestern U.S. into Colorado and Nebraska) flavotessellata (Walton) 10a(5b). Mediotergite polished immediately beneath postscutellum; parafacial hairs distant from eye; haustellum 0.5 times head height or less (Fig. 11); legs of fe- males yellow; species with members small, dark 11 lOb. Mediotergite with at least a fine dusting of tomentum; parafacial hairs and haus- tellum varied 12 lla. Abdomen and thorax black, strongly shining; abdomen with tomentum evenly distributed; parafacial hairs black, coarse (Fig. 16); squamae of males dark brown (northcentral and eastern U.S.) mathesoni (Curran) lib. Abdomen with tomentum in large irreg- ular patches; thorax with definite tomen- tose striping; parafacial hairs black, fine (Fig. 1 1); squamae of males white to pale brown (western U.S. and Texas) agilis Reinhard (in part) 12a(10b). Length of plumosity on arista less than or equal to 1.25 times length of sec- ond antennal segment (southwestern U.S. and Mexico) 13 12b. Length of plumosity on arista more than 1.25 times length of second antennal seg- ment (widespread) 15 13a. Abdomen and scutellum concolorous dark gray or brown, slightly lighter on ventral margins of tergites in a few in- 18 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 42, No. 1 dividuals: ninth tergite and parafacial hairs varied; haustellum length 0.7 to 0.8 times head height (Fig. 31) 14 13b. Abdomen and scutellum distinctly red- dish laterally; ninth tergite with several strong setae (Fig. 33); parafacial hairs long, dark, occurring along parafacial ventrally to level of oral vibrissae; haus- tellum length 0.6 times head height (Fig. 36) contristans (Wulp) 14a. Parafacial hairs long, coarse, numerous, occurring on entire parafacial (Fig. 31) (Arizona, New Mexico into Mexico) .. planifrons (Wulp) 14b. Parafacial hairs fine, sparse, occurring only on anterior parafacial (Fig. 41) (Texas, Oklahoma, Kansas) westi n.sp. (in part) 15a(12b). Parafacial hairs short, pale or oth- erwise inconspicuous 16 15b. Parafacial hairs long, dark, although sparsely placed in some individuals __ 20 16a. Mediotergite polished; legs of females yellowish; haustellum length no more than 0.5 times head height (Fig. 1 1); scu- tellum concolorous with rest of notum agilis Reinhard (in part) 16b. Mediotergite tomentose to subshining; legs of females dark; haustellum length varied; scutellum lighter in color than rest of notum in most individuals 17 17a. Parafacial hairs pale and, in most indi- viduals, small (Fig. 97) (west only to Ar- izona and Rocky Mountains) 18 17b. Parafacial hairs light brown to black, short or medium in length (Pacific coast states) 19 18a. Infrasquamal setulae absent or greatly reduced in number in most individuals; parafacial hairs only on upper anterior parafacial (Fig. 102); length of palpi equal to or slightly less than half length of haustellum, broadened at tip (central and eastern U.S.).... incerta (West) (in part) 18b. Infrasquamal setulae present in most in- dividuals; parafacial hairs extending ventrally on face to level of apex of an- tennae (Fig. 97); length of palpi greater than half length of haustellum, narrow (southwestern U.S. into Texas) major (Bigot) 19a(17b). Third antennal segment subequal to or shorter than second segment; smallest distance between eyes of male subequal to width of frontal vitta at antennal base; female with parafacial hairs confined to area near antennae; facial tomentum dull (southern California) sabroskyi n.sp. 19b. Third antennal segment longer than sec- ond segment; smallest distance between eyes of male less than width of frontal vitta at base of antennae; females with parafacial hairs scattered, often along center of parafacial; facial tomentum shining (throughout California) calif arnica n.sp. 20a. Haustellum length greater than 0.6 times head height, thin, narrowed apically 21 20b. Haustellum length less than or equal to 0.6 times head height, broad and linear in most individuals 24 2 la. Length of plumosity on arista more than twice length of second antennal segment; haustellum length 0.65 to 0.7 times head height, narrow (Fig. 81) (northern and eastern U.S.) harpasa (Walker) 21b. Length of plumosity on arista less than or equal to twice length of second anten- nal segment; haustellum length varied (western U.S.) 22 22a. Parafacial hairs long, dark, and abun- dant, uniformly covering parafacial (Fig. 26); haustellum length 0.6 to 0.8 times head height, narrowed apically in most individuals; abdomen reddish laterally __ conjuncta (Wulp) 22b. Parafacial hairs sparse (Fig. 41); haus- tellum length and abdomen varied 23 23a. Length of haustellum more than 0.7 times head height, strongly narrowed apically; parafacial hairs sparse, located only on anterior portion of parafacial (Fig. 41); male abdomen dark brown or gray with little if any reddish color; ninth tergite with several strong setae (Fig. 38) (Texas and Oklahoma) westi n.sp. (in part) 23b. Length of haustellum less than 0.65 times head height, broad; not narrowed apically in most individuals; parafacial hairs occurring along center of parafacial (Fig. 46); male abdomen distinctly red- dish with longitudinal stripe; ninth ter- WILDER: NEARCTIC PTILODEXIA 19 gite with hairs only (Pacific coast states) calif ornica n.sp. (in part) 24a(20b). Abdomen blackish with little if any orange coloration laterally; parafacial hairs inserted close to anterior margin of eye in most individuals; palpi very short, less than half length of haustellum (Fig. 21); wings of many males basally dark- ened obscura West 24b. Abdomen, especially of male, with at least some red or orange coloration lat- erally; parafacial hairs not inserted close to anterior eye margin in most individu- als; palpi longer than half length of haus- tellum (Fig. 51); wings of males not ba- sally darkened 25 25a. Femora of members of both sexes or- ange or marked with orange; abdominal tomentum concolorous (California) pacifica n.sp. 25b. Femora of members of both sexes brown or black (some specimens from Texas may have orange femora); abdominal to- mentum bicolored (widespread) 26 26a. Epistomal angle of head not prominent (Fig. 61); females with 6 to 8 dorsal and lateral marginal setae on abdominal seg- ment III; males with 1 pair of dorsal mar- ginal setae on abdominal segment II; haustellum length 0.4 to 0.5 times head height (central and eastern U.S.) carolinensis Brauer and Bergenstamm (in part) 26b. Epistomal angle of head prominent (Fig. 46); females with 10 to 12 dorsal and lat- eral marginal setae on abdominal seg- ment III; most males with 2 pairs of dor- sal marginal setae on abdominal segment II; haustellum length 0.5 to 0.65 times head height (western U.S.) calif ornica n.sp. (in part) 27a(lb). Abdomen without any reddish or or- ange coloration laterally in most males; tip of abdomen and genitalia generally reddish yellow; facial tomentum strongly shining; length of plumosity on arista more than twice length of second anten- nal segment; face appearing narrow (Fig. 87); infrasquamal setulae absent in many individuals; femora of many males yel- low (central and eastern U.S. into New Mexico and British Columbia) rufipennis (Macquart) 27b. Abdomen orange laterally in most males; tip of abdomen not noticeably lighter than rest of abdomen; facial tomentum rather dull; length of plumosity on arista at most equal to twice length of second antennal segment; face appearing broad (Fig. 92); infrasquamal setulae present; femora of males dark (Utah and southern Idaho south into Mexico) arida (Curran) The Nearctic Species of Ptilodexia agilis Group Ptilodexia agilis Reinhard (Figures 8-12) Ptilodexia agilis REINHARD, 1943:22. SABROSKY AND AR- NAUD (1965:988). [HOLOTYPE, male, deposited in CNCI, labeled, College Station, Texas, 8 Oct. 1933, H. J. Rein- hard.] TAXONOMIC NOTES. — Although I have not seen the holotype of P. agilis, members of this species are quite distinctive, and the original description is adequate to assure the identity of the specimens examined. Even though P. agilis was described from a disjunct population, the type-specimens are typical of the species. Type material is reported to be in excellent condition. DIAGNOSIS. — Ptilodexia agilis is a distinct species, members of which can be separated from their congeners by the following character states: size small; parafacial hairs fine, medium in length, inserted below apex of antennae or less than 0.25 mm from anterior margin of eye in only a few individuals; haustellum less than half head height; length of plumosity on the aris- ta 1.5 to 2.0 times length of second antennal segment; mediotergite polished; abdomen and scutellum entirely blackish; female with legs pale colored. MATERIAL EXAMINED. — Specimens exam- ined include 812 males and 488 females, data as listed by Wilder (1976). DISTRIBUTION. — Ptilodexia agilis ranges from Arizona and New Mexico north through Cali- fornia, the Great Basin, and the Rocky Moun- tains into Alberta and British Columbia. There are a few records from eastern Texas, where the topotypic population is found. BIOLOGICAL NOTES. — The flight period lasts from April to October, and adults can be col- 20 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 42, No. 1 10 FIGURES 8-12. Ptilodexia agilis Reinhard. Fig. 8. Genitalia of male, posterior view. Fig. 9. Ejaculatory apodeme. Fig. 10. Genitalia of male, lateral view. Fig. 11. Head of male, lateral view. Fig. 12. Distribution of P. agilis. lected at any time during this period. July and August are the most frequent months of collec- tion, but local variation is common. In Califor- nia, for example, P. agilis adults are collected more frequently in September and October along the coast and in the south; but in the Sier- ra, northern California, and Oregon, June and July are the main periods of activity. This species inhabits both mountains and low- lands. Adults have been collected at elevations up to 2,600 m in Arizona, 3,800 m in California, above the 3,000-m level in Colorado, and fre- quently above 3,000 m elsewhere. Ptilodexia agilis adults have frequently been taken at low elevations in such areas as the San Joaquin Val- ley, the Great Basin, and eastern Texas. Spec- imens have been collected by UV light trap, Malaise trap, wind vane trap, and by sweeping foliage. Flowers visited include the Compositae Achil- lea Millefolium, Chrysothamnus viscidiflorus , Baccharis glutinosa, B. pilularis, Eriogonum nudum, Solidago trinervata, and Lepidospartum squamatum. Unidentified Compositae visited were Achillea sp., Baccharis sp., Solidago sp., Eriogonum sp., and Haplopappus sp. Other flowers from which P. agilis adults have been collected are Allium (Liliaceae) and Salix (Sal- icaceae). At two localities in Colorado, speci- mens were collected on Dwarf Mistletoe (Ar- ceuthobium cyanocarpum)', and members of this species are believed to be pollinators of that plant. One specimen was collected in an emer- gence trap under a filbert tree in Oregon. There are no data on the life history of this insect. Its members are probably not host spe- cific, judging from the diverse assortment of habitats and wide geographical and temporal ranges of the species. Ptilodexia mathesoni (Curran) (Figures 13-17) Rhynchiodexia mathesoni CURRAN, 1931:93. WEST (1950:110); SABROSKY AND ARNAUD (1965:989). [LECTOTYPE (here des- ignated), male, deposited in CUIC, labeled, "Douglas Lake, Mich., 24-VII-22"/"Wing Slide, Cornell U., Lot. 919, Sub. 138, L. S. West"/"cJ Holotype Rhynchiodexia mathesoni WILDER: NEARCTIC PTILODEXIA 21 16 FIGURES 13-17. Ptilodexia mathesoni (Curran). Fig. 13. Genitalia of male, posterior view. Fig. 14. Ejaculatory apodeme. Fig. 15. Genitalia of male, lateral view. Fig. 16. Head of male, lateral view. Fig. 17. Distribution of P. mathesoni. Curran'T'Holotype Cornell U. No. 1938"/"Cornell U. Lot. 922, Sub. 40"/"Lectotype Rhynchiodexia mathesoni Cur- ran designated by D. Wilder, 1976."] TAXONOMIC NOTES. — Even though the label on the type-specimen reads "Holotype Rhyn- chiodexia mathesoni Curran," this specimen is not a holotype. The author of the paper vali- dating the species made no mention of type ma- terial or of type-locality. The type label with Curran as author was put on the specimen at a later date; it was West's "holotype," not Cur- ran's. The name P. mathesoni was proposed by West and validated by Curran in his 1931 key. I have designated West's "holotype" as the lec- totype. It is a large specimen in excellent con- dition. There is one other specimen which I believe Curran had before him while writing his key. This specimen, also from Douglas Lake, Mich- igan, is deposited in AMNH and was collected on the same date as the specimen West labeled as holotype. I have designated this specimen as a paralectotype. It might be argued that this sec- ond specimen, because it is labeled with Curran as the author, was later sent to AMNH and not seen by Curran. However, the type labels of all five of the species validated in Curran's key give Curran as the author. I believe that all these labels were changed at a later date, and since Curran makes no mention of material, I feel that these two identically labeled specimens are ac- tually syntypes. DIAGNOSIS. — Ptilodexia mathesoni is a dis- tinctive species, evidenced by the following combination of character states: body color black, shining; parafacial hairs long, black, in- serted ventral to apex of antennae or less than 0.25 mm from anterior margin of eye in only a few individuals; haustellum less than half head height in length; width of arista and plumosity 1.5 to 2 times length of second antennal seg- ment; mediotergite polished; wings dark in col- or; male abdomen with pollen evenly distribut- ed; female with pale-colored legs. MATERIAL EXAMINED. — Twenty-one males and eleven females were examined. 22 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 42, No. 1 21 FIGURES 18-22. Ptilodexia obscura West. Fig. 18. Genitalia of male, posterior view. Fig. 19. Ejaculatory apodeme. Fig. 20. Genitalia of male, lateral view. Fig. 21. Head of male, lateral view. Fig. 22. Geographical distribution. DISTRIBUTION. — The species, although rec- ords are few, seems to range through the north- eastern United States and eastern Canada. There is one record from Victoria Beach, Man- itoba. BIOLOGICAL NOTES. — The flight period is from April to October with the main period of activity between mid-June and mid-August. All records for months other than July and August are from Suffolk County, New York. There are no life-history data for P. mathe- soni. One male was collected on Solidago can- adensis. Ptilodexia obscura West (Figures 18-22) Ptilodexia obscura WEST, 1925:133. LEONARD (1928:822); CURRAN (1930:93); SABROSKY AND ARNAUD (1965:989). [HoLOTYPE, female, deposited in SIIS, labeled, "Wading River, L.I., June 29, 1917, W. T. Davis."] TAXONOMIC NOTES. — West described P. ob- scura from three female specimens, one of which he designated holotype. The two para- types are so labeled and are deposited in CUIC. They closely resemble the holotype. There is a male specimen from Victoria Beach, Manitoba, deposited in AMNH, which bears a handwritten label reading, "/?. obscura West." It appears that West recognized the male of the species at a later time, even though this particular male specimen differs considerably from the female type-series. West still used the generic name Ptilodexia for obscura in 1950, but because his concept of the genus changed, it is possible that he might have written "/?. obscura" instead of "P. obscura." It is doubtful that Curran labeled the specimen or ever looked at West's types, since in his 1931 key, he describes P. obscura specimens as being over 12.5 mm long. DIAGNOSIS. — Ptilodexia obscura is a fairly distinctive species. Its members may be sepa- rated from their congeners by the following combination of character states: face narrow, with parafacial hairs abundant, long, fine, dark, and inserted below lower edge of eye and close to its anterior edge; third antennal segment dark in most specimens, broadened apically; length of plumosity on arista 1.5 times length of second WILDER: NEARCTIC PT1LODEXIA 26 FIGURES 23-27. Ptilodexia conjuncta (Wulp). Fig. 23. Genitalia of male, posterior view. Fig. 24. Ejaculatory apodeme. Fig. 25. Genitalia of male, lateral view. Fig. 26. Head of male, lateral view. Fig. 27. Geographical distribution. segment; haustellum broad, length 0.55 to 0.6 times head height; palpi short, half haustellum length or less; notum covered with short, ap- pressed, grayish pollen, giving it a smooth, sub- shining appearance; legs of females brown; ab- dominal color, grossly appearing black, but actually rufescent laterally. MATERIAL EXAMINED. — One hundred forty- two males and 53 females of P. obscura were examined. DISTRIBUTION. — The range of P. obscura ex- tends from New Brunswick south into the Geor- gia Appalachians west into Arkansas, eastern Kansas and Nebraska, western South Dakota and Saskatchewan, and Manitoba. BIOLOGICAL NOTES. — The flight period lasts from April to September with June and July the most common months of collection, a later av- erage seen only in Manitoba. Most specimens have been collected at low elevations, the ex- ceptions coming from 800-1,700 m in the Ap- palachians in Tennessee, North Carolina, and Virginia. Collecting methods yielding specimens of P. obscura include sweeping and Malaise trap. Specimens have been collected on the flowers of Solidago canadensis (Compositae), and Aruncus sp. and Spiraea latifolia, both Rosa- ceae. conjuncta Group Ptilodexia conjuncta (Wulp) (Figures 23-27) Rhynchodexia conjuncta WULP, 1891:228. ALDRICH (1905:499); GUIMARAES (1971:33). [LECTOTYPE (here designated), male, deposited in BMNH, labeled "Lectotype"/"tJ'V "B.C. A. Dipt. II, Rhynchodexia conjuncta v.d.W. "/"Cen- tral America. Pres. by F. D. Godman. O. Salvin. 1903- 172"/"Ciudad, Mex., 8100 ft., Forrer'VLectotype Rhyn- chodexia conjuncta Wulp designated by D. Wilder, 1975."] Rhynchodexia simulans WULP, 1891:229. ALDRJCH (1905:499); GUIMARAES (1971:34). [LECTOTYPE (here designated), male, deposited in BMNH, labeled, "Lectotype'V'N. So- nora, Mexico. Morrison"/" d"/"B. C. A. Dipt. II, Rhyn- chodexia simulans, v.d.W. "/"Central America, pres. by F. D. Godman, O. Salvin. 1903-172"/"Lectotype Rhyncho- dexia simulans Wulp, designated by D. Wilder, 1975."] NEW SYNONYMY Ptilodexia tibialis (partim): ALDRICH (1905:504). TAXONOMIC NOTES. — Wulp described P. conjuncta from two male cotypes, one of which I have designated as lectotype. This specimen 24 PROCEEDINGS OF THE CALIFORNIA ACADEMY OF SCIENCES, Vol. 42, No. 1 is in good condition except for a few broken setae. Wulp described P. simulans and P. conjuncta in the same paper. The cotype (one of two) which I saw and designated lectotype is in good condition, although it has the dorsal setae and abdomen broken. Wulp realized that P. conjuncta and P. sim- ulans were very closely related. He separated members of each on the basis of size, curvature of the hind tibia, and some other minor char- acters. Size, of course, cannot be used effec- tively to separate these parasitic flies. A curved hind tibia is a character state which occurs fre- quently in members of many species of Ptilo- dexia, especially in those of P. conjuncta, where it constitutes part of normal intraspecific varia- tion. DIAGNOSIS. — Ptilodexia conjuncta is a vari- able species. Its members can be distinguished by the following: face wide; parafacial hairs long, fine, dark, abundant, inserted below level of lower edge of eye in some specimens and at least 0.12 mm from anterior edge of eye in all but a few specimens; haustellum long, ranging from 0.6 to 0.8 times head height, slender, nar- rowed apically in most specimens; carina fairly well developed; width of arista and plumosity 1.5 to 2 times length of second antennal seg- ment; oral margin distinctly protruding; scutel- lum and sides of abdomen reddish. MATERIAL EXAMINED. — Specimens exam- ined included 1,238 males and 629 females. DISTRIBUTION. — This species ranges from British Columbia and Alberta south through the Rocky Mountains into Mexico. There are scat- tered records from the plains states, the Great Basin, California, and Oregon. There are a few doubtful records from the eastern United States. BIOLOGICAL NOTES. — The flight period ranges from February to October. Most of the activity occurs in July and August, earlier in the north- ern areas and later further south. In Arizona (and possibly Texas), P. conjuncta appears to have two broods, one in March and April, and another in August and September. Ptilodexia conjuncta adults are generally found in mountainous areas at altitudes from 1,500 to 2,750 m and in some areas up to 3,660 m. They are also, although less commonly, col- lected at low elevations in coastal as well as in- land areas. Label data indicate that they have been collected in meadow sweeps, in a meadow in spruce-fir zone (2,750 m), in pine-spruce-as- pen zone, and above timberline. Productive col- lecting methods for P. conjuncta specimens are UV light and Malaise trap. Specimens of P. conjuncta have been col- lected from many flowers. These include: Se- necio salignus, Heliopsis parvifolia, Cacalia de- composita, Enceliafarinosa, Geraea canescens, Chrysothamnus greeni, Achillea Millefolium; and unidentified species of Ence I'm sp., Gutier- rezia sp., Solidago sp., Baccharis sp., Helian- thus sp., Senecio sp., Eriogonum sp., Achillea sp., Bigelovia sp., Aster sp., and Rudbeckia sp. (all Compositae). Other flowers visited include, Ceanothus fendleri and unidentified Ceanothus sp. (Rhamnaceae); Arctostaphylos sp. (Erica- ceae); Melilotus sp., and Dale a sp. (Fabaceae); Lippia wrightii (Verbenaceae); and Arceuthob- ium sp. (Loranthaceae). Ptilodexia conjuncta adults, along with those of P. agilis, have been observed pollinating Dwarf Mistletoe (Arceu- thobium cyanocarpum). Ptilodexia planifrons (Wulp) (Figures 28-32) Rhynchodexia planifrons WULP, 1891:234. ALDRICH (1905:499); GUIMARAES (1971:33). [HOLOTYPE, male, deposited in BMNH, labeled, "Holotype'VCiudad, Mexico, 8100 ft., Ferrer"/" d"/"B. C. A. Dipt. II, Rhynchodexia planifrons, v.d.W'VCentral America pres. by F. D. Godman, O. Sal- vin. 1903-172."] Dexia harpasa (partim): ALDRICH (1925:114). (Misidentifica- tion). TAXONOMIC NOTES. — The holotype of this species, deposited in BMNH, is in poor condi- tion, but is still recognizable. The facial band is obscured, the tibiae are quite light and distinctly curved (a frequently encountered anomaly in species of Ptilodexia), and the striping on the notum is more distinct. Labels on the specimen say only "Ciudad, Mexico, 8100 ft.," but in the description, the origin of this specimen is stated as "Ciudad in Durango, 8100 ft." I follow Wulp's original pub- lication in calling Durango the type-locality. DIAGNOSIS. — Ptilodexia planifrons is a dis- tinctive species and its members can be identi- fied by the following character combination: face wide; presence of a contrasting tomentose diagonal band extending from antennal base to eye margin; parafacial hairs dark, coarse, abun- dant, inserted lower than level of oral vibrissae only in a few specimens; antenna with plumosity WILDER: NEARCTIC PTILODEXIA 25 28 31 FIGURES 28-32. Ptilodexia planifrons (Wulp). Fig. 28. Genitalia of male, posterior view. Fig. 29. Ejaculatory apodeme. Fig. 30. Genitalia of male, lateral view. Fig. 31. Head of male, lateral view. Fig. 32. Geographical distribution. on arista less than or equal to length of second antennal segment; carina long, slightly keeled and prominent; two pairs of oral vibrissae in most specimens; frontal vitta very wide at an- tennal base; haustellum long, 0.7 to 0.85 times head height, narrowed apically; scutellum and abdomen entirely dark gray or black. MATERIAL EXAMINED. — One hundred sev- enty-three males and 70 females of P. planifrons were examined. DISTRIBUTION. — This species ranges from Flagstaff south through the mountains of central and southeastern Arizona, through the central mountainous region of New Mexico, the west- ern tip of Texas, and south into Durango, Mex- ico. There is one record from Colorado; how- ever, no exact locality is given. BIOLOGICAL NOTES. — The flight period lasts from August through October, with the majority of records from mid-August to mid-September. The earliest seasonal record is one specimen collected on 27 June (error?) from El Paso, Tex- as, in 1921, and the latest is a series of 24 females collected on 22 October 1964, in Cochise Coun- ty, Arizona. Ptilodexia planifrons adults generally occur in the mountains and have been collected at many elevations between 1,370 and 3,350 m. Flowers visited include Gutierrezia sarothrae (1,800- 2,440 m, Apache County, Arizona), Heliopsis parvifolia (2,590 m, Chiricahua Mts., Cochise County, Arizona), Solidago trinervata (Sierra Madre, 2,230 m), and Rudbeckia sp. in Chihua- hua, all Compositae. All except two of the flower-visiting flies were males (one female col- lected on Heliopsis sp. and another at Rudbeck- ia sp.). One specimen was collected at a light. Ptilodexia contristans (Wulp) (Figures 33-37) Hystrichodexia contristans WULP, 1891:221. [HOLOTYPE, male, deposited in BMNH, labeled, "Holotype'VOmil- teme, Guerrero, 8000 ft., July, H. H. Smith"/" d"/"B. C. A. Dipt. II., Hystrichodexia contristans, v.d.W. "/"Central America, Pres. by F. Godman, O. Salvin. 1903-172."] Rhynchodexia punctipennis WULP, 1891:233. ALDRICH (1905:499); GUIMARAES (1971:33). [LECTOTYPE (here des- ignated), male, deposited in BMNH, labeled, "Cotype"/ "Sierra de las Aguas Escondidas, Guerrero, 9500 ft., July. H. H. Smith"/"